Defrosting for air source heat pump : research, analysis and methods [First edition] 9780081025185, 0081025181, 9780081025178

Citation preview

Defrosting for Air Source Heat Pump

Defrosting for Air Source Heat Pump Research, Analysis and Methods

SONG Mengjie DENG Shiming

An imprint of Elsevier

Woodhead Publishing is an imprint of Elsevier The Officers’ Mess Business Centre, Royston Road, Duxford, CB22 4QH, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, OX5 1GB, United Kingdom © 2019 Elsevier Ltd. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-08-102517-8 (print) ISBN: 978-0-08-102518-5 (online) For information on all Woodhead publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Joe Hayton Acquisition Editor: Maria Convey Editorial Project Manager: John Leonard Production Project Manager: Joy Christel Neumarin Honest Thangiah Cover Designer: Victoria Pearson Typeset by SPi Global, India

Introduction 1.1

1

Background

Similar to a water pump that could move fluids (liquids or gases) or sometimes slurries against gravity, a heat pump can transfer heat from the heat source at a low temperature to a heat sink at a high temperature by consuming power, which is against the direction of heat transfer from high temperature to low temperature. In applications, heat pumps are widely used to transfer thermal energy in the opposite direction of spontaneous heat transfer by absorbing heat from a cold space and releasing it to a warmer one. The most common design of a heat pump includes four basic components: a condenser, an expansion valve, an evaporator, and a compressor. Refrigerant is the heat transfer medium being circulated through these components in a heat pump. There are two main types of heat pumps: absorption and compression. An absorption heat pump could be powered by oil, gas, or solar energy. The fuel utilization efficiency in an absorption heat pump is evaluated by the ratio of the energy supplied to that consumed. For example, when gas is used to power an absorption heat pump, the utilization efficiency could reach 1.0–1.5. Compression heat pumps, on the other hand, are most likely powered by electricity. When comparing the performance of a compression heat pump, the term “coefficient of performance (COP)” as the ratio of useful heat to work input is commonly used. For a compression-type heat pump, although it has the obvious advantage of higher energy performance than that of an absorption-type heat pump, its disadvantage of consuming high-quality electricity limits its application in waste energy recovery. Heat pumps might also be categorized by their heat sources, such as air, ground soil, and water, into air-source heat pumps (ASHPs), ground-source heat pumps (GSHPs), and water-source heat pumps (WSHPs). Clearly, air is everywhere, and is the most common, safe, stable, and cheapest thermal source. An ASHP extracts heat from the ambient air and transfers the heat to the indoor air as an air-air heat pump, or to hot water in a domestic hot water tank as an air-water heat pump. In practice, air-air heat pumps are widely found; science air conditioners just work following the same principle. When an air conditioner is used in summer, it removes thermal energy from the indoor air to the outdoor air. However, when it reversely operates in winter, thermal energy is moved from the outside air for heating the indoor air. ASHPs are the most common type of heat pumps with the lowest investment cost. Air-water heat pumps may also easily be found as water heaters, by transferring the extracted heat from the outside air for water heating all year round. In winter, the hot water provided by an air-water heat pump may also be partly or wholly used in a water-based spaceheating system, such as a floor radiant heating system, which is efficient and enables a comfortable indoor environment. ASHPs may also be used to extract thermal energy from exhaust air, and they are thus sometimes called exhaust air heat pumps. The exhaust air can be from buildings due to ventilation or from some industrial Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00001-1 © 2019 Elsevier Ltd. All rights reserved.

2

Defrosting for Air Source Heat Pump

production processes. It is easy to understand that the evaporator side heat exchanger might be specially made. The use of ASHPs is advantageous. Compared with other space heating methods, an ASHP unit does not need a plant room, but could be placed on a roof or ground at will to save floor area and reduce the construction and installation costs. As compared to a boiler-based space heating system, ASHPs are safe and reliable and do not produce environmental pollution. However, the performances of ASHP units would vary with the changes in outdoor weather conditions. When the unit extracts heat from lowtemperature air, frost may be formed on the surface of its outdoor coil. To maintain normal operation of the ASHP unit, defrosting will be necessary once adequate frost has been accumulated, which reduces the operating efficiency and output heating capacity of the ASHP unit. On the other hand, a GSHP unit converts low-grade shallow geothermal energy to high-grade thermal energy by consuming a small amount of electricity. A GSHP unit usually consumes 1 kWh of electrical energy to produce 4.4 kWh of thermal energy. In fact, the heat drawn from soil is in most cases the stored solar heat, and hence should not be confused with that in direct geothermal heating. In a geothermal heating system, a circulation water pump but not a heat pump is required because the soil temperature is higher than that in a space to be heated. Therefore, such technology relies only upon convective heat transfer. As an efficient measure against rising energy costs, GSHP technology has attracted worldwide attention since the 1980s and has been a hot topic in China since the late 1990s. Earlier investigations on GSHP technology focused on the performance tests for experimental GSHP systems, and technical and economic comparisons with traditional ASHP units. Later, the complex heat and mass transfer between ground heat exchangers used in a GSHP system and subsurface rock and soil was investigated in great detail, with a large number of heat transfer models developed and reported. Recently, novel hybrid GSHP units and the determination of soil thermal properties became hot research topics. Furthermore, a WSHP utilizes energy resources in shallow water on the Earth’s surface, such as the solar energy and geothermal energy absorbed by groundwater, rivers, streams, and lakes. However, river/seawater heat pumps and wastewater heat pump systems are the most widely used systems. When the river/seawater is used as a heat source or sink, it is always stable with an almost unlimited capacity. For example, several big data centers were built near rivers/seas, and released heat to the water using a heat pump or direct sea/river water cooling. However, in recent years, more and more environmental problems, such as the death of fish and water grass or the changes and migration of microorganisms, emerged. These further destroyed local ecological environments. On the other hand, wastewater heat pump systems take thermal energy from treated domestic wastewater, and would thus have no adverse impacts on local environments. Hybrid heat pump systems have also been widely studied and reported. A solarassisted heat pump integrates a heat pump and thermal solar panels in a single integrated system. Typically, the two technologies are used independently to produce hot water. However, in this integrated system, the solar thermal panel functions as a low-temperature heat source and the heat collected in the panel is fed to the heat

Introduction

3

pump’s evaporator. A hybrid system can produce thermal energy in a more efficient and less expensive way. A further typical hybrid system is an air/water-brine/water heat pump. Unlike other hybrid systems, this system usually utilizes both conventional and renewable energy sources. For example, it uses air and geothermal heat in a single compact device. There are two evaporators in an air/water-brine/water heat pump: an outdoor air evaporator and a brine evaporator. Both evaporators are connected to the heat pump cycle to allow the use of the most economical heating source according to the actual operating conditions, air, geothermal heat, or both.

1.2

Frosting and defrosting

As previously mentioned, when an ASHP unit works at heating mode in winter to take thermal energy from the ambient air, the surface temperature of its outdoor coil can be much lower than both the dewpoint of air and the freezing temperature of water. Therefore, frost would form and accumulate on an outdoor coil surface. In fact, the formation of frost on the surfaces of plant leaves is a well-known phenomenon in nature. This phenomenon was recorded in an ancient Chinese poem, The Reed. The second verse of this love poem reads, “Dew and frost gleam,” meaning that the dew on the surface of reed leaves was changed into frost. The fourth verse reads, “Beyond the stream,” meaning that the reed was near a river where the air humidity was high. The poem may describe the natural phenomenon on an autumn day at dusk, when air temperature was low. Clearly, the following three conditions–cold surface, high air humidity, and low air temperature–are the prerequisites for frost to be formed. Therefore, it can be easily understood that frosting is commonly observed in refrigeration fields because of operating conditions [1]. Frost deposition is inevitable once moist air is exposed to a cold surface having a temperature below the water triple point and the air dewpoint [2]. To clearly understand the frosting mechanism, the process of frost formation on a cold flat plate surface is shown in Fig. 1.1 [3]. It could be divided into four periods according to the growth timeline: (1) the droplet condensation period, (2) the solidified liquid tip growth period, (3) the frost layer growth period, and (4) the frost layer full growth period. During the droplet condensation period, the condensing droplets at a subcooling state are formed on the cold surface. The droplet nucleation occurs first, followed by the coalescence of the droplets. As the vapor-liquid and liquid-solid phase changes take place, the droplets merge and solidify; the diameters of the solid droplets increase significantly. h

Droplet nucleation Coalescence

Droplet

Crystal

Tip

Vertically growth

Horizontally growth

Tips Hollow frost

Ice layer

Frost layer

d

(1) Droplet condensation period

tc

(2) Solidified liquid tipgrowth period

(3) Frost layer growth period (4) Frost layer full growth

tt

period

Fig. 1.1 The four periods in a frost growth process on the surface of a cold plate.

t

4

Defrosting for Air Source Heat Pump

After reaching a critical time, tc, all coalescent droplets turn into ice particles or crystals. The length of this period is determined by the ambient and surface conditions, including the surrounding air temperature/relative humidity, the natural or forced convection affected by air velocity, the cold surface temperature, and its roughness. The second period, the solidified liquid tip-growth period, starts from the critical time to a transitional time, tt, when a relatively uniform porous layer of frost is formed. At this period, nonuniform tip growth occurs on individual droplets. The entire process of frost branch formation is referred to as the third period, the frost layer growth period, when frost branches would form at the top of ice crystals. These frost branches grow in three dimensions and connect to the neighboring frost branches, thus forming a flat frost layer. During the fourth period, the frost layer full growth period, the interface between the frost layer and the ambient air is at a temperature of 0°C, owing to the thermal resistance of the porous frost layer and the upper part of the frost layer being melted into water. The melted water penetrates the frost layer and freezes, thus forming a new, considerably thicker ice layer. When ice layers form inside the frost layer, the density of the frost layer increases, and the conduction thermal resistance decreases. As a result, the surface temperature of the frost layer is lower than 0°C, causing the frost layer to grow on the surface again. After ice crystals form on the surface, the water vapor on the surface that surrounds the humid air becomes frost, which branches on the top of the ice crystals and no further mass transfer take places between the surface and the humid air. In fact, all previously mentioned ambient and surface conditions may affect a frost formation process. Such has been extensively studied on the surfaces of plates and heat exchangers. In addition, frosting characteristics on hydrophobic and superhydrophobic surfaces are different from those on ordinary surfaces. The changes in surface properties affect the frosting behaviors at the early stage, from a dry surface to the formation of ice crystals. The surface properties are hardly possible to influence the growth of the frost layer, and the change in the surface properties is considered only prior to the crystal growth period or the droplet condensation period. Frosting has caused severe negative effects in various application fields. Therefore, numerous theoretical and experimental studies have been conducted, aiming at the mitigation and control of frost formation on various cold surfaces. For example, in the field of aerospace technology, the phenomenon of frost deposition is harmful. For the airplanes that fly at night, frost deposits may occur on their wings owing to low air temperature. The frost deposition increases the surface friction drag during take off or navigation, thus affecting safety [4]. In launching rockets, there are similar frosting problems. Similar to that on the surface of airplane wings, frost may deposit on the rocket surface, thus causing satellites to fail to enter a correct synchronous orbit [5]. For liquid-fueled rockets, frost may also deposit on the surface of cryogenic oxidizer tanks of very low temperature, as rockets fly through the atmosphere. The frost deposition may change both the shape and the weight of the oxidizer tanks, thus influencing the aviation performance of the rocket. In addition, in the field of LNG production and application, frost may deposit on the surfaces of LNG evaporators, and thus affect the so-called passive-evaporation technology that uses air as a heat source [6]. Harmful frosting phenomena mainly occur in various industry processes

Introduction

5

that occur under normal low-temperature conditions. Hence, most theoretical and experimental research on frost deposition has focused on normal low-temperature conditions. However, because of the increased practical and engineering applications, frost formation on the surfaces with a very low temperature has been receiving increasing attention. In addition, in the case of a cryogenic tank that is exposed to humid atmospheric air, frost formation may have positive effects because the frost that would form on the cold tank surface would act as insulation, thus reducing heat transfer into the tank [5]. For an ASHP unit, frost formed on the surface of its finned outdoor coil will behave as a layer of thermal resistance between the humid ambient air and the surface, which reduces the heat transfer rate. The frost layer also reduces airflow passages and hence increases the pressure drop on the air side, reducing the output heating capacity and performance operation of the ASHP unit. To understand the factors influencing the frosting of an ASHP unit, a large number of experimental investigations on the influence of inlet air temperature, relative humidity, and velocity of air passing through the outdoor coil have been carried out. Also, frosting models based on an ASHP unit, an outdoor coil, or a cold plate have been developed. Frost-suppression measures have also attracted growing research attention, including fin surface treatment, fin type adjustment, geometric structure optimization of an outdoor coil, dehumidifying and preheating inlet air, and adding external electric and magnetic fields, etc. Although these measures could efficiently delay frosting and lengthen a frosting operation of an ASHP unit, periodic defrosting is still necessary. Consequently, a series of defrosting methods has been reported. Based on the most widely used defrosting method, RCD, a large number of experimental and numerical studies have been conducted to improve the performances during both the frosting and defrosting processes.

1.3

Objectives and scopes

The outdoor coil in an ASHP unit is usually of a multicircuit structure in order to enhance its heat transfer and minimize its refrigerant pressure loss. It is easy to understand that it is hardly possible for frost to be evenly distributed on the surface of each circuit, and for the refrigerant to be evenly distributed into each circuit during defrosting. Therefore, for the subject of frosting and defrosting in an ASHP unit having a multicircuit outdoor coil, a series of investigations has been carried out and the study results will be given in this book. Hence, the objectives of the research work and the scopes presented in this book are as follows: 1. Investigation on the effect of downward-flowing melted frost due to gravity on the defrosting performance for a multicircuit outdoor coil in an ASHP unit. During RCD, the melted frost would flow downward over the surface of the multicircuit outdoor coil. It has previously been experimentally demonstrated that the melted frost might be the reason for uneven defrosting, and thus adversely impacting the defrosting performance by prolonging the defrosting duration of the entire coil. Therefore, a series of experimental studies has to be carried out to qualitatively and quantitatively investigate the effect of downward flowing

6

2.

3.

4.

5.

6.

Defrosting for Air Source Heat Pump

of melted frost on defrosting performance. In addition, when the circuit number in an outdoor coil is changed, the variations in the effects due to downward flowing have also been studied. Development, validation, and application of a defrosting model based on the multicircuit outdoor coil with the melted frost flowing downward over its surface in an ASHP unit. Although defrosting models are reported in the literature, most of them are based on an entire ASHP system, and thus cannot be directly applied to defrosting with downward-flowing melted frost. Based on a multicircuit outdoor coil, a new semiempirical defrosting model has been developed, taking into account the melted frost flowing downward over the surface of a multicircuit outdoor coil. This model is validated with experimental data, and thus applied to optimizing defrosting control. Alleviation of the uneven defrosting and the investigation of the influence of uneven frosting on frosting and defrosting in an ASHP unit having a multicircuit outdoor coil. When the defrosting process for different circuits is terminated at a different time, there would be some thermal energy waste in heating the ambient air during defrosting. Therefore, such an uneven defrosting should be avoided. An experimental study on the alleviation of the uneven defrosting has been carried out by changing the installation of the outdoor coil from vertically to horizontally. In addition, unevenly distributed frost on a different circuit’s surface leads to different thermal insulation levels during frosting and thermal loads during defrosting. Experimental studies on the influences of uneven frosting on both frosting and defrosting processes have been carried out. Investigation on the influences of uneven refrigerant distribution into each circuit on the RCD performance in an ASHP unit having a vertically installed multicircuit outdoor coil. Similar to the uneven distribution of frost on the surfaces of different circuits, the refrigerant would also be unevenly distributed into each circuit during frosting and defrosting. For a vertically installed multicircuit outdoor coil, this may be due to the gravity force effect at a refrigerant distributor and the tube internal resistance in each circuit. During frosting, the uneven refrigerant distribution can be coupled with that of the inlet air of an outdoor coil, thus affecting the frost distribution on each circuit. During defrosting, even and uneven refrigerant distribution statuses should be experimentally investigated, with the located defrosting performance analyzed. Investigation on the mechanism of heat transfer during RCD in an ASHP unit, and the evaluation of the effect of thermal energy stored in the metal of indoor and outdoor coils on defrosting performance. There are different types of heat supply and energy consumption during reverse cycle defrosting, and the reason for a long defrosting duration is the insufficient energy supply. Therefore, to optimize defrosting performance and shorten the defrosting duration, the mechanism of energy transfer through the refrigerant has been experimentally studied and analyzed. In addition, certain thermal energy is stored in the metal of indoor and outdoor coils during defrosting, and the sum of the two should vary during defrosting. Therefore, its influence on defrosting should be further experimentally investigated. Optimization of the defrosting initiation and termination control strategies during RCD for an ASHP unit, taking account of with and without the melted frost flowing downward along the surface of the multicircuit outdoor coil. The mal-defrost phenomenon appears during frosting and defrosting, resulting in low capacity and efficiency during frosting and a low defrosting efficiency and longer defrosting duration, respectively. Therefore, for an ASHP unit having a multicircuit outdoor coil, the optimization for the defrosting initiation and termination control strategies should be carried out. When the melted frost flowing downward along the surface of an outdoor coil is considered, the strategies could be more complicated.

Introduction

7

7. Technoeconomic performance analysis on the optimization adjustments of the ASHP unit, based on the previous frosting and defrosting experimental results. After a series of optimization adjustments is taken in the previous experimental study, including adjusting the refrigerant distribution by installing valves and local drainage of melted frost with water collecting trays, their practical applications are considered, and thus their economic performances during frosting and defrosting are quantitatively analyzed. The outcomes are also expected to be useful in the pricing or governmental subsidy policy for ASHP units.

1.4

Book outline

This book consists of 11 chapters and 6 appendices, as follows. Chapter 1 introduces the background, the classifications and advantages of heat pump technology, and frosting and defrosting issues for an ASHP unit; states the motivation to develop the new methods or technologies to improve the frosting and defrosting performances of an ASHP unit; and elaborates the objectives and scopes of the research work presented, followed by the outline of this book. Chapter 2 reviews the published research works on frost-suppression measures, defrosting methods, improvements for reverse cycle defrosting, and the defrosting initiation and termination control strategies for ASHP units, respectively. Chapter 3 presents experimental studies of the uneven defrosting on the outdoor coil in an ASHP unit, having both two working circuits and three working circuits. In each set of studies, both the conditions of with and without water collecting trays installed to locally drain the melted frost away during defrosting were considered. The effects of melted frost downward flowing due to gravity during defrosting along the multicircuit outdoor coil on the operating performance of the ASHP unit are qualitatively and quantitatively investigated. Chapter 4 develops two sets of semiempirical mathematical models for RCD for an ASHP unit having a multicircuit outdoor coil, with and without the consideration of melted frost downward flowing along the surface of the outdoor coil. A defrosting process is divided into four stages according to the defrosting physical timeline, and the models are validated with the experimental results reported in Chapter 3. Using the validated models, the heat supply and energy consumption during defrosting could be analyzed, and some parameters that are difficult or even hardly possible to be measured are predicted. The two models are further applied to alleviating uneven defrosting for ASHP units with different proposed control strategies. Chapter 5 reports the investigations of the effects of the elimination and retention of melted frost due to surface tension on uneven defrosting in an ASHP unit having a horizontally installed multicircuit outdoor coil. For a horizontally installed multicircuit outdoor coil, the flow path and direction of melted frost are changed, so that uneven defrosting is expected to be alleviated and the defrosting performance thus improved. In addition, some melted frost can also be retained on the downside of the outdoor coil surface due to surface tension during defrosting, which would consume some thermal energy.

8

Defrosting for Air Source Heat Pump

Chapter 6 introduces the results of experimental studies on unevenly distributed frost on the surface of each circuit in a multicircuit outdoor coil during both frosting and defrosting. First, the even and uneven frosting statuses are comparatively investigated, and thus the effect of uneven frosting on COP demonstrated. Then, the investigations on defrosting performance are carried out with even and uneven frosting and the considerations of whether the melted frost is locally drained. The investigation results are used to evaluate the effects of uneven frosting on defrosting performance. Chapter 7 investigates the influence of refrigerant distribution on defrosting performance in an ASHP unit having a vertically installed multicircuit outdoor coil. Similar to that in Chapter 6, whether the melted frost is locally drained away is considered. The results of defrosting durations and energy analysis are also used to evaluate the uneven refrigerant distribution effect on defrosting performance. Chapter 8 analyzes the energy transfer mechanism during RCD in an ASHP unit, where all types of heat supply and energy consumption are quantitatively compared. While the roles of indoor and outdoor coils are changed during defrosting, the thermal energy stored in the metal of the two coils is also changed. The net energy transfer in two coils is analyzed, with the influence on defrosting performance evaluated. Chapter 9 reports the optimization of defrosting initiation and termination control strategies for an ASHP unit having a multicircuit outdoor coil. First, a time-based initiation of defrosting control strategy is experimentally investigated, with different total frost accumulations on the outdoor coil surface. Then, the condition of melted frost locally drained during defrosting is considered, and thus the corresponding experiments carried out. Third, a defrosting termination strategy is optimized with a more suitable defrosting termination temperature suggested, based on extensive experimental results. In this chapter, the defrosting initiation and termination control strategies are evaluated based on defrosting performance. Chapter 10 reports the technoeconomic performances of optimized ASHP units by installing valves to adjust the refrigerant distribution and water collecting trays to locally drain away the melted frost during defrosting. Both frosting and defrosting are considered, and their operational conditions at three typical seasons analyzed. Finally, the effects of installing valves and trays on economic performance, the variation of total running costs, and a payback analysis for the additional initial costs are given. Chapter 11 draws conclusions from the research work presented in this book and recommends possible future work on the development of defrosting technologies for ASHP units. In addition, Appendices A and B are provided to calculate the defrosting efficiency and its error, and the metal energy storage effect on defrosting performance. Appendices D and E list the calculation methods for frosting evenness coefficient and defrosting evenness coefficient, respectively. Appendix F presents the program listing of Model 1 in Section 4.2.

Introduction

9

References [1] Liu ZL, Dong YW, Li YX. An experimental study of frost formation on cryogenic surfaces under natural convection conditions. Int J Heat Mass Transfer 2016;97:569–77. [2] Lee YB, Ro ST. Frost formation on a vertical plate in simultaneously developing flow. Exp Therm Fluid Sci 2002;26(8):939–45. [3] Hayashi Y, Aoki A, Adachi S, Hori K. Study of frost properties correlating with frost formation types. J Heat Transfer 1977;99SerC(2):239–45. [4] Bragg MB, Heinrich DC, Valarezo WO. Effect of under-wing frost on a transport aircraft airfoil at flight Reynolds number. J Aircr 1994;31(6):1372–9. [5] Dietenberger MA. A frost formation model and its validation under various experimental conditions, In: NASA contract report 3595, NASA scientific and technical information branch; 1982. [6] Chen SP, Lai JL, Yin JS. Experiment on frost characteristic of airwarmed cryogenic finnedtube vaporizer. J Refrig 2010;31(4):26–30 in Chinese.

Previous related work: A review 2.1

2

Introduction

A heat pump unit is an environmentally friendly and reliable means to maintain an appropriate thermal comfort level in an indoor space, and can be used for both space heating and cooling at a high operating efficiency. During a cooling season, it transfers heat from an indoor space to a heat sink, in the same way an air conditioner does. During a heating season, it extracts thermal energy from a heat source, and delivers the extracted thermal energy to a heated indoor space. From a global point of view, 90% of the worldwide population resides in regions where heat pump units can be suitably used for indoor environmental control [1–3]. Compared with traditional space heating and/or heat generation methods using coal or electricity, studies have shown the potential of using heat pump units to help drastically reduce greenhouse gases, in particular CO2 emissions. With the rising cost of energy at the forefront of world attention, there has been a growing interest in using heat pump technology as an energy-saving means. A number of heat sources are available for space heating heat pump units such as air, underground water, and soil. Among these heat sources, air and water are the most common ones for space heating heat pump units. Therefore, air-to-air heat pump units, air-to-water heat pump units, water-to-air heat pump units, and water-to-water heat pump units are commonly found in buildings or industry. Among them, ASHP units are relatively easy and inexpensive to install, and have therefore been the most widely used types of heat pump units for many years. An ASHP unit actually consists of vapor compression refrigeration, a condenser or an indoor coil, and an evaporator or outdoor coil to transfer heat from one place to the other. Heat pump technology is based on the Carnot cycle, which was presented by Carnot in 1824. With nearly 200 years of history, this technology is very mature. However, when an ASHP unit works at the heating mode in winter, at an ambient air temperature of
7oC to 5oC and relative humidity (RH) higher than 65%, frost is likely deposited and a frost layer is formed on the surface of its outdoor coil [4]. Frosting increases both heat transfer resistance and air flow passage resistance during the heating operation, and thus adversely degrades the system performance, or even results in an undesired shutdown. Therefore, extensive experimental and theoretical investigations have been carried out to study ASHP units’ operating performances under frosting and/or defrosting conditions. Related studies have become very hot in recent years. In addition, the server air pollution in winter and the corresponding coal to electricity policy in China have also motivated related interests and research work in the applications of ASHP units. Review articles have been published on the topics of domestic heat pumps [5, 6], air-to-air heat exchangers [7–9], frost-suppression methods [10], defrosting methods [11, 12], etc. To provide an overview of available studies for researchers, product developers, and policy makers, a review and a comparative analysis of the available literature from 2000 to 2017 on frosting/defrosting studies for ASHP units are Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00002-3 © 2019 Elsevier Ltd. All rights reserved.

12

Defrosting for Air Source Heat Pump

presented in this chapter. First, a review of the studies on frost-suppression measures for ASHP units is reported, covering changing ambient air parameters at the outdoor coil inlet, optimizing the structure of an outdoor coil, and other frost-suppression measures. This is followed by reporting various defrosting methods for ASHP units, with emphasis on the studies for RCD optimization using experimental and theoretical approaches. Finally, control strategies to initiate and end a defrosting operation for an ASHP unit are summarized.

2.2

Frost-suppression measures for ASHP units

As previously discussed, frost formation and accumulation on the surface of the outdoor coil in an ASHP unit is an undesirable phenomenon. Frosting duration accounts for more than 80% of the operational period in a frosting-defrosting cycle, and thus frost suppression plays an important role in the optimization of ASHP units. In order to improve the operating performance of ASHP units, frost-suppression measures have attracted increasing research attention in recent decades. Previous studies on developing frost-suppression measures are classified and summarized. There are basically two types of frost-suppression measures: the external type shown in Fig. 2.1A and the internal type shown in Fig. 2.1B, respectively.

2.2.1 Changing ambient air parameters at outdoor coil inlet It is easy to understand that frost formation on the surface of the outdoor coil of an ASHP unit is closely related to ambient air conditions at which the ASHP unit is operated, such as air temperature, relative humidity, airflow rate, etc. Consequently, studies on analyzing ambient air parameters at the outdoor coil inlet of an ASHP unit have been carried out, with a quantitative investigation of the influence of these parameters

Fig. 2.1 Frost-suppression measures for ASHP units.

Previous related work: A review

13

on frost suppression. These studies are meaningful for accurately controlling the operating performances of an ASHP unit at a fixed climate environment.

2.2.1.1 Reducing inlet air humidity Because water is the source of frost, the relationship between frosting and reduced inlet air humidity [4] was investigated, and the leading role of inlet air humidity among other parameters affecting frosting was demonstrated. Also, the use of a desiccant to significantly reduce the frosting rate was experimentally and comparatively investigated, with and without placing a solid desiccant upstream of the inlet air [13]. Furthermore, using a solid desiccant, a novel frost-free ASHP unit was presented, and its frost-suppression characteristics were experimentally and numerically investigated [14–16]. Using this method, not only was inlet air humidity reduced, but also the air temperature increased by absorbing the heat from the solid desiccant, so that less frost was formed on the outdoor coil’s surface at a given frosting duration. Thereafter, there have been increasing research interest in developing alternative dehumidifying methods and medium. For example, liquid desiccant was used for dehumidifying inlet air, due to a number of advantages including low air pressure drop, air cleaning effects, and lower regeneration temperature. Table 2.1 lists 10 typical studies on solid and liquid desiccants for frost suppression for ASHP units from 2000 to 2017. Clearly, related studies are still ongoing using both experimental and numerical approaches. For an ASHP unit, reducing inlet air humidity is the most fundamental measure to suppress frosting. However, its initial cost is high, and the floor space of the additional desiccant facility will be increased. In addition, desiccant regeneration costs thermal energy, thereby increasing the running cost. These economic issues limit the practical application of this frost-suppression measure.

2.2.1.2 Preheating inlet air Preheating the inlet air to an outdoor coil is a simple but effective technique to reduce or prevent frosting for an ASHP unit. Certain heating elements can be placed in the air duct for the inlet air, so that when the outdoor air temperature drops below the frosting point, the heating elements are activated. To prevent frosting, the air temperature at the inlet of an outdoor coil must always be higher than the frosting point [7]. This heatrecovery technique was first coupled with this frost-suppression method by Liu et al. [23]. The exhausted indoor air and ambient air were mixed before entering the outdoor coil, and thus the frosting duration was prolonged and the rate of frost growth reduced. Then, to experimentally investigate the frost-suppression effect, an electric heater was fixed upstream of the outdoor coil in an ASHP unit [24, 25]. It was shown that the ASHP unit’s heating capacity was increased by 38.0% and the COP by 57.0% when the electric heater was turned on for heating the inlet air at the outdoor air condition of below 2/1°C (dry-bulb/wet-bulb temperature). However, the disadvantage of preheating the inlet air in very cold regions was the high energy consumption [26]. A comparison of different frost-suppression measures suggested that preheating the

14

Table 2.1 Studies on using solid and liquid desiccants for frost suppression for ASHP units (2000–2017) Year

Author

Typea

Desiccant used

Results

1

2017

Wang et al. [15]

N

Solid (silica gel)

2

2017

Su and Zhang [17]

N

Liquid desiccant (membrane-based, lithium chloride solution)

3

2017

Wang et al. [16]

N

Solid (silica gel)

4

2015

Wang et al. [14]

E

Solid (silica gel)

5

2015

Zhang et al. [18]

E&N

Liquid (lithium chloride solution)

Frost-free durations when using R22, R407C, and R134a were 29, 34, and 35 min, respectively. At the given ambient temperature of
10°C and RH of 85%, the average coefficient of performance (COP) when using R134a was 3.3, 8.6% higher than using the other refrigerants The COPsen and the COPtot using the novel defrosting method were at least 37.7% and 64.3% higher than that of the COPreverse of using the conventional defrosting method in the variation ranges of the analyzed parameters, respectively. Air humidification by a regenerator can meet the needs of indoor thermal comfort when the ambient air was above 0°C and 70% RH At RH of 80%, average COP increased by 56.2% when ambient temperature was increased from
10°C to 0°C. The COP was decreased by 6.7% when RH was increased from 75% to 85% at 0° C It can keep the evaporator frost-free for 32, 34, 36 min at the heating mode at the ambient temperatures of
3°C, 0°C, and 3°C, with an RH of 85% The air velocity, temperature, and moisture content of the desiccant affects the average overall mass-transfer coefficient

Defrosting for Air Source Heat Pump

No.

2014

Wang et al. [19]

E

Solid (silica gel)

7

2014

E

Liquid (glycerol)

8

2012

N

Solid (not given)

9

2010

Jiang et al.[20] Zhang et al.[21] Zhang et al.[22]

E

Liquid (lithium chloride solution)

10

2005

Wang et al. [13]

E

Solid (zeolite plates and active carbon)

Inlet air RH can be reduced to 52% after dehumidification and the outdoor heat exchanger can be kept frost-free for 34 min at a temperature of 0oC and RH of 80% Frost would never form on the outdoor coil’s surface if a solution spray subsystem was operated continuously COP of the frost-free ASHP water heater could be increased by 5%–30%, as compared to that of an ASHP water heater Frosting could be suppressed, and the COP of the hybrid system was improved by approximately 20% and 100% in summer and winter, respectively Frosting problem was resolved, and the performance of the heat pump unit was improved in winter

Previous related work: A review

6

a

N, numerical study; E, experimental study.

15

16

Defrosting for Air Source Heat Pump

inlet air was not economical in regions with long periods of very low outdoor air temperatures, at
54°C to 10°C [27]. Therefore, to improve the economy of the ASHP units, the energy used in preheating the inlet air should come from waste heat, such as the heat in the exhausted indoor air or waste hot water.

2.2.1.3 Increasing the inlet airflow rate to an outdoor coil Increasing the inlet airflow rate to an outdoor coil is also an external frost-suppression measure. An experimental study on frost formation on a finned-tube heat evaporator considering fan characteristics was conducted by Da Silva et al. [28]. The study results demonstrated that airflow rate reduction was a dominant factor for the drop in the evaporator’s capacity. It was further suggested that the fan evaporator should be treated as a coupled system under frosting conditions. To predict the performance of an outdoor coil considering airflow reduction due to frost growth, a numerical model was developed and validated by Ye and Lee [29]. In this study, the frost layer was assumed to be evenly distributed on the surface of the heat exchanger. Results showed that the simulated heat transfer rates and the accumulated frost mass agreed well with the experimental data by 7% and 9%, respectively. In practice, for better frost suppression, the above air parameters can be changed altogether. The changes in different ambient air parameters would influence the rate of frost suppression. However, an experimental investigation on the adverse effect of frost formation on a microchannel evaporator was undertaken by Moallem et al. [30], suggesting that the air face velocity of the evaporator impacted less significantly on the rate of frost growth. In addition, increasing the fan power input and the noise level may be hardly avoided.

2.2.2 Removing frost with additional equipment A further external frost-suppression measure is to destroy frost with additional equipment, such as ultrasonic vibration or air jet techniques. Unlike the aforementioned frost-suppression measures, no heat but only mechanical energy is used to destroy the frost formation and growth.

2.2.2.1 Ultrasonic vibration technique The ultrasonic technique was first used for frost suppression by Yan et al. [31] and Li et al. [32]. As reported, the frost formation process on a flat surface was remarkably restrained due to the effect of the ultrasound. Droplets with ultrasound are smaller than those without ultrasound. After quantitative analysis of the sizes, the frost coverage was all less than 52% of the coil surface with ultrasound, as compared to more than 65% without ultrasound. Frost crystals and frost branches on the ice layer could be fractured and removed more effectively [33]. Using intermittent ultrasonic vibrations, an experimental study on the defrosting performance of a finned-tube evaporator was carried out by Tan et al. [34]. The study results indicated that the intermittent ultrasonic vibrations could effectively remove the frost accumulated on the fin surface. The energy consumption for defrosting the ASHP unit decreased by about 3.14%–5.46%,

Previous related work: A review

17

whereas the heating capacity was increased by 2.2%–9.03% and the COP by 6.51%– 15.33%. As reported, the indoor thermal comfort level was improved. However, the effect of using intermittent ultrasonic vibrations for frost suppression is limited because a basic ice layer on the surface could not be removed with ultrasonic vibrations. It was believed that the mechanism of ultrasonic frost suppression was mainly attributed to high-frequency ultrasonic mechanical vibrations that could break up frost crystals and frost layers, then frost would fall off by gravity, but not due to the ultrasonic cavitation effect or heat effect. These continued studies promoted the use of the ultrasonic vibration technique for frost suppression, but its application is limited in practice due to its high initial cost and complex control system.

2.2.2.2 Air jet technique The air jet technique may also be an effective frost-suppression measure, and no thermal energy is needed to melt the frost [27]. It was first applied to a horizontal single-row array of cooled tubes immersed in a gas-solid fluidized bed. The heat transfer and defrosting characteristics of the cooled tubes were experimentally investigated, and the fluidized bed produced gas-solid particle impinging jets that effectively removed frost layers on the tube surface. It had been verified that frost-free running of the cooled tubes was possible under an operating condition of inlet air temperature of
7°C, inlet air RH of 80%, and a tube surface temperature of
17°C. Fei and Mao [35] experimentally investigated the use of compressed air for frost suppression, and indicated that this measure could remove frost in a timely manner. Hence, it could be applied where compressed air was available. Furthermore, the measure of frost suppression on heat exchangers using solid particles accelerated by an air jet impinging on the heat exchanger surfaces was studied by Sonobe et al. [36]. The study was motivated by the development of a cryogenic heat exchanger for a hypersonic aircraft engine. However, as with the ultrasonic vibration technique, not much literature about frost suppression using air jet techniques has been identified. In addition to the disadvantages of high initial and running costs and the inconvenience of use, the technique’s effect on frost suppression for an ASHP unit is not well understood, which limits its practical applications.

2.2.3 Optimizing outdoor coil structure Apart from the external frost-suppression measures described in Sections 2.2.1 and 2.2.2, a number of internal measures to suppress frosting have been developed through optimizing the structure of an outdoor coil so as to alleviate the negative impact of frosting on the operating performance of an ASHP unit.

2.2.3.1 Adjusting fin and tube geometry The use of an outdoor coil having a wider fin space was first recommended to slow down frost growth by Young and Watters et al. [37]. Then, it was experimentally investigated by Yan et al. [38] and Sommers and Jacobi [39]. As reported, at air-side Reynolds numbers between 500 and 1300, the air-side thermal resistance was reduced

18

Defrosting for Air Source Heat Pump

by 35%–42% when vortex generation was used by way of adjusting the fin structure [39]. Yang et al. [40] proposed the optimal values of design parameters for a fin-tube heat exchanger of a household refrigerator under frosting condition to improve the thermal performance and extend the operating time. After optimizing the fin and tube geometry, the average heat transfer rate and operating time were increased by up to 6.3% and 12.9%, respectively. Lee et al. [41] measured and analyzed the air-side heat transfer characteristics of flat finned-tube heat exchangers at different fin pitches, numbers of tube rows, and tube alignment under frosting conditions. It was shown that the fin pitch and staggered tube alignment had greater affects on airflow reduction, and thus affects frost suppression. Recently, the frosting behaviors and thermal performance of louvered fins with an unequal louver pitch were studied by Park et al. [42]. The study results demonstrated that the blocking of the spaces between louvers at the front side by frost was delayed and the thermal performance was improved by 21% when an unequal louver pitch design was used. Frost accumulation on the equal louver pitch was more than that on the unequal lover pitch. Further, the design in which the louver pitch was successively decreased from the air inlet region to the redirection region provided more uniform frost growth and improved thermal performance. Due to the aforementioned frost-suppression effect, optimizing the fin and tube geometry has attracted more and more attention from researchers.

2.2.3.2 Adjusting fin type The fin-type adjustment was used as a frost-suppression measure. Yan et al. [43] experimentally investigated the operating performances of frosted finned-tube heat exchangers with flat-plate fins, one-sided louver fins, and redirection louver fins. When other conditions were the same, the amount of frost formation was the largest for the heat exchanger with redirection louver fins. Dong et al. [44] experimentally compared the effects of periodic frosting-defrosting performance by using three fin types in an outdoor coil of a residential ASHP unit. The outdoor coil with a flat fin demonstrated the best thermal performance in the periodic frosting/defrosting cycles of the ASHP unit, followed by that with wavy and louver fins, respectively. Zhang and Hrnjak [45] experimentally studied three types of heat exchangers with louver fin geometry under dry, wet, and frost conditions: (1) a parallel flow serpentine fin with extruded flat tubes, (2) a parallel flow parallel fin with extruded flat tubes, and (3) a round tube wave plate fin. As indicated, at the frosting condition, the heat exchanger with the round tube wave plate fin can be used for the longest time due to its largest surface area. The increase in air-side pressure drop for the heat exchanger of the parallel flow parallel fin with extruded flat tubes was the lowest. Although certain types of fins could be used to suppress frosting, the total number of fin types and thus their effects on frost suppression are limited.

2.2.3.3 Coating treatment on the fin surface There have been reported studies on the influence of fin surface coating treatment of outdoor coils on frosting and defrosting performances. Okoroafor and Newborough [46] found that frost growth on cold surfaces exposed to warm humid air streams could

Previous related work: A review

19

be reduced significantly by means of cross-linked hydrophilic polymeric coatings. The frost thickness was decreased in the range of 10%–30% when compared to using an uncoated metallic surface. Wu and Webb [47] investigated both frosting and defrosting processes on hydrophilic and hydrophobic surfaces, showing that a hydrophilic coating was preferable, with less frost and retained water on it. Cai et al. [48] experimentally studied the frosting conditions on a normal copper surface, a hydrophobic coating (car wax coating) surface, and a hygroscopic coating (glycerol coating) surface. Frost growth could be restrained by using both a hydrophobic coating and a hygroscopic coating at the initial stage of its formation, and the thickness of the hydrophilic coating was directly proportional to the frost-suppression effect. Similar results have also been reported by Jhee et al. [49] and Liu et al. [50]. It was quantitatively reported that the use of surface hydrophilic polymer paint could suppress frost formation by up to 3 h and reduce frost thickness by at least 40%, and the frost layer formed on the coated surface was loose and could be easily removed [50]. In addition, there have been studies in an attempt to understand the mechanism of surface treatment on frosting/defrosting. For example, Chen et al. [51] reported a hierarchical surface that allowed interdroplet freezing wave propagation suppression and efficient frost removal. It was demonstrated that the enhanced performances were mainly because of the activation of the microscale edge effect on the hierarchical surface, which increased the energy barrier for ice bridging as well as engendering the liquid lubrication during defrosting. It was believed that the concept of harnessing the surface morphology to achieve superior performances in two opposite phase transition processes might shed new light on the development of novel materials for various applications. As summarized in Table 2.2, related studies on the mechanism of surface treatment on frost suppression at regular/nanoscales were widely published in 2000–2017. Although optimizing the structure of outdoor coils by way of adjusting the fin space and alignment could effectively suppress frost, changing the fin types and coating treatment on the fin surface would make the design and manufacture of outdoor coils more difficult, and increase the initial cost of an ASHP unit. Hence, more and more related research work is being carried out.

2.2.4 System adjustment and optimization Adjusting and optimizing the structure of ASHP systems may also be viewed as external frost-suppression measures for ASHP units.

2.2.4.1 Vapor-injection technique The vapor-injection technique has been marketed for use in room air conditioners since 1979, but its application to ASHP units only received more attention recently, as it can help suppress frosting in cold climates [69]. Zhnder et al. [70] tested an airwater vapor-injection heat pump unit at an inlet air temperature of
7°C and reported an increase in heat output of 28% and a COP improvement of 15%, respectively, as compared to a unit without injection. Also, at an ambient temperature of
7°C,

Table 2.2 Studies on the mechanism of surface treatment on frost suppression (2000–2017) No.

Year

Author

Country

Scalea

Results

1

2017

Zuo et al. [52]

China

N-P

2

2017

da Silva et al. [53]

Brazil

R-HEX

3

2017

Wu et al. [54]

China

R-F

4

2016

Japan

N-P

5

2016

Moriya et al. [55] Sommers et al. [56]

The United States

R-F

6

2015

Zhao et al. [57]

China

N-P

7

2015

Kim et al. [58]

South Korea

R-F

8

2015

Liang et al. [59]

China

R-F

9

2015

China

R-HEX

10

2014

Wang et al. [60] Bharathidasan et al. [61]

India

R-F

Frost formation on an as-prepared superhydrophobic ZnO surface was effectively delayed for more than 140 min at
10°C The test conducted for an evaporator operated at
5°C ambient resulted in a denser frost structure, which was quite different from that observed when the evaporator was maintained at operated
10°C temperature, when dendritic and needle-shaped ice crystals were presented The surface with crossed grooves has the least frost accumulation, and that with parallel grooves the best melted frost drainage performance The use of fluorocarbon-based coatings delayed frost formation compared with the use of an uncoated surface The frost density on the hydrophilic surface was 20%–26% higher than a baseline surface. Reductions in frost density of 37%–41% were observed on hydrophobic surfaces A type of aluminum-based condensate microdrop self-propelling functional film, based on the controllable fabrication of anodic alumina rodcapped nanopores, was reported, with self-cleaning, antifrosting, and antidewing functions When the refrigerant temperature was at
10°C or
12°C, the effect of frost suppression was increased remarkably with superhydrophobic surfaces; when the refrigerant temperature was at
8°C, the effect of superhydrophobicity diminished at water contact angles greater than 150 degrees The time required for frost melting on the surfaces of the hydrophilic fin, the bare fin, the hydrophobic fin, and the super hydrophobic fin was 36, 25, 23, and 22 s, respectively The frost thickness and mass on a superhydrophobic heat exchanger were 17.1% and 28.8% less than those on a bare one Hydrophilic surface coatings displayed higher ice-adhesion strength than hydrophobic silicone coatings. Superhydrophobic coatings showed the best performance

11

2014

Li et al. [62]

China

R-HEX

12

2013

Chen et al. [51]

Hong Kong

N-P

13

2013

Miljkovic et al. [63]

The United States

N-P

14

2013

Kim et al. [64]

South Korea

R-HEX

15

2012

Moallem et al. [65]

The United States

R-HEX

16

2009

Canada

R-F

17

2009

China

R-HEX

18

2009 2002

20

2000

The United States South Korea The United Kingdom

N-P

19

Kulinich and Farzaneh [66] Huang et al. [67] Boreyko et al. [68] Jhee et al. [49]

a

Okoroafor et al. [46]

R-HEX R-HEX

Water retention on the coil surface has three stages: water, water and ice, and then mainly ice. In the third stage, a “permafrost area” appears, and takes 20% of the surface area A hierarchical surface that allowed for interdroplet freezing wave propagation suppression and efficient frost removal was reported. The enhanced performances were mainly because of the activation of the microscale edge effect on the hierarchical surface Condensation heat transfer was significantly enhanced, and a low cost and scalable approach to increase efficiency for applications, such as atmospheric water harvesting and dehumidification, was promised The leading-edge effect was not observed for a hydrophobic heat exchanger, and a hydrophobic surface showed the highest overall heat transfer rate during the repeated frosting and defrosting experimental cycles, due to frost suppression There were visible differences in the type, appearance, and patterns of frost. Hydrophobic and hydrophilic coatings on microchannel coils affected the heat transfer capacity at frosting conditions by up to 15% On super-hydrophobic surfaces with low wetting hysteresis, ice adhesion strength was observed to be up to  5.7 times lower than on a bare polished aluminum surface The coated hydrophilic fins were free of frost deposition during the whole test while the uncoated fins were completely covered by a dense and thick frost layer Continuous dropwise condensation spontaneously occurring on a superhydrophobic surface without any external forces was reported Hydrophilic treatment mainly influenced the behaviors of frosting while hydrophobic treatment influenced the behaviors of defrosting Cross-linked hydrophilic polymer coating is a potential approach in minimizing frost growth on cold surfaces exposed to humid air

N-P, plate at nanoscale; R-F, fin at regular scale; R-HEX, heat exchanger at regular scale.

22

Defrosting for Air Source Heat Pump

Nguyen et al. [71] evaluated the thermal performances of a flash tank vapor-injection cycle and that of a subcooler vapor-injection cycle using R-407C and reported their heating COPs as 24% and 10% higher than those of a single-stage cycle, respectively. Then, with outdoor temperatures of
20°C to 15°C, Shao et al. [72] concluded that a vapor-injection heat pump unit could provide enough heating capacity. At an ambient temperature of
25°C, Ma and Zhao [73] experimentally investigated the operating performance of an ASHP unit with a flash tank coupled with a scroll compressor and demonstrated that the ASHP unit was more efficient than that with a subcooler at
25°C to
7°C.

2.2.4.2 Two-stage technique For the two-stage technique, Wang et al. experimentally investigated a double-stage heat pump heating system, which coupled an ASHP unit and a water source heat pump unit [74]. Results indicated that it offered an average energy efficiency ratio up to 3.2 with a minimum of 2.5, and the average indoor temperature of 19.5°C with a minimum of 18°C during tests. Compared with conventional ASHP systems, the operating characteristics of the coupled system were greatly improved, so that it had considerable application potential in cold regions. Li et al. [75] proposed and experimentally tested a new frost-free ASHP system, concluding that the system could operate more efficiently than a conventional ASHP unit in winter, and there was no need to periodically defrost. At an ambient temperature of
15°C, the COP and heating capacity of a twostage vapor injection cycle were enhanced by 10% and 25%, respectively, as reported by Heo et al. [76]. At an ambient temperature of
17.8°C, Wang et al. [77] found that a maximum COP improvement of 23% for a two-stage heat pump system was achieved. Bertsch and Groll [78] tested a specially designed R410A two-stage ASHP unit, and a heating COP of 2.1 was observed at an ambient temperature of
30°C. Although this technique may be employed for frost suppression in cold climates, it makes the system more complicated than vapor injection.

2.2.4.3 Applying an external heating source It is easy to understand that applying an external heat source could improve the system operating performance at frosting conditions for an ASHP unit. Masaji qualitatively demonstrated that the performances of an ASHP unit with a-kerosene fired burner placed either near its indoor coil or under its outdoor coil could be improved at low ambient temperatures [27]. Using the same method, Mei et al. [79] reported that the heating capacity of an ASHP unit could be increased and the frost accumulation on its outdoor coil suppressed by heating the liquid refrigerant in its accumulator. By heating the liquid refrigerant, it was shown that the frequency of defrosting cycles was reduced by a factor of 5 in Knoxville, Tennessee, in the United States, and the indoor supply air temperature increased by 2–3°C because of the increased compressor suction pressure. Different from preheating the inlet air of the outdoor coil, the measure of applying an external heating source was to the heat refrigerant. However, they both consumed a lot of energy. Therefore, to improve the economy of the ASHP

Previous related work: A review

23

units, the heating source should come from waste heat, such as the heat recovered from exhausted indoor air or waste hot water. However, the application of this frost-suppression measure is also limited due to the disadvantages of high running cost and inconvenience. Although the energy consumed for frost suppression is carried over by the refrigerant circulated in an ASHP system, it is actually input into the system. This measure is essentially different from external measures such as increasing the inlet air temperature or decreasing the inlet air RH, where the energy is carried away by the inlet air. As listed in Table 2.3, the evaluation results of 11 classical frost-suppression measures are given. Nearly all the measures would increase the initial cost and/or the running cost, except the two measures of adjusting the fin and tube geometry and the fin type. Adjustment and optimization would increase the system complexity and decrease system stability. Additional thermal energy is needed for the measures of preheating the inlet air and employing an external heat source. Four measures require floor space for additional equipment, such as a desiccant bed for reducing inlet air humidity and an air jet or ultrasonic vibration facility to destroy frost formation or growth. Among all the measures, reducing the inlet air humidity and preheating the inlet air have the best frost-suppression effect. Among all measures listed, preheating the inlet air with waste heat and putting a coating treatment on the fin surface with new coating materials are highly recommended, due to both having the highest evaluation index in Table 2.3.

2.3

Defrosting methods for ASHP units

As discussed earlier, the presence of frost on the surface of the outdoor coil in an ASHP unit would deteriorate its operating performance, energy efficiency, reliability, and lifespan. While the use of frost-suppression measures can delay or reduce frost formation or growth, these measures can be expensive or consume additional energy, and there will still be frost to be removed. Therefore, periodic defrosting becomes necessary for guaranteeing the satisfactory operation of ASHP units. To distinguish frost suppression from defrosting, the differences between the two are summarized in Table 2.4. In this section, various defrosting methods are reviewed based on the assumption of normal frosting operation, where no frost-suppression measures are implemented during heating/frosting. Generally speaking, there are five types of defrosting methods: (1) compressor shutdown defrosting, (2) electric heating defrosting, (3) hot water spraying defrosting, (4) hot gas bypass defrosting, and (5) reverse cycle defrosting. In order to clearly show the differences for the five defrosting methods.

2.3.1 Compressor shutdown For the compressor shutdown defrosting (CSDD) method, ambient air is used as the heat source for defrosting. Therefore, it is normally applied to where the ambient air temperature is not lower than 1°C. When defrosting is needed, the compressor is shut down but the outdoor coil air fan continues to move the ambient air at >1°C to pass

Table 2.3 Evaluation results of 11 classical frost-suppression measures

Frost-suppression measure

Initial costa

Running costa

System complexitya

Installation floor spacea

Additional thermal energyb

System stabilitya

Defrosting effectc

Comprehensive evaluation indexc

External

"

"

!

"



3

3

2

!

"

!

!

√

3

3

3

"

"

!

!



2

2

2

"

"

!

"



2

2

2

" !

" !

! !

" !

 

1 3

2 1

2 1

!

!

!

!



3

1

1

"

!

!

!



3

2

3

"

!

"

!



1

1

1

"

"

"

"



1

2

2

"

"

"

!

√

2

2

1

1 2 3 4

Internal

5 6

7 8

9 10 11

Reducing inlet air humidity Preheating inlet air Increasing inlet airflow rate Ultrasonic vibration technique Air jet technique Adjusting fin and tube geometry Adjusting fin type Coating treatment on fin surface Vapor-injection technique Two-stage technique Adding outside heat source

", increased; !, unchanged. √, in need; , no need. c 3, the best; 1, the worst. a

b

Previous related work: A review

25

Table 2.4 The differences between frost suppression and defrosting No.

Aspects

1

Operation initiation

2 3

Operation termination During operation

4

Operation effects

5

Evaluation index

6

Duration fraction in a frostingdefrosting cycle

Frost suppression When frost appears Never Heating mode continued Not all frost removed COP More than 80%

Defrosting After frost accumulates Periodic Heating mode stopped All frost removed Defrosting efficiency; defrosting duration Less than 20%

through the outdoor coil to melt the frost. Ameen et al. experimentally investigated the defrosting performance of an ASHP unit using warm air under controlled conditions in an air-conditioned wind tunnel [12]. The effectiveness of the compressor shutdown defrosting method was also demonstrated by Shang [80], by experimentally investigating the effect to prestart an outdoor coil fan of an ASHP unit on defrosting performance. Its advantages would include low initial cost, no reconstruction work, and easy control, making it widely applied. However, because the energy for defrosting comes from ambient air, it would take a long time for a defrosting operation to complete, during which the indoor thermal comfort can be degraded.

2.3.2 Electric heating Electric heating defrosting (EHD) usually involves electrically heating the surface of an outdoor coil to melt off frost. Kim et al. [81], Bansal et al. [82], and Ozkan et al. [83] conducted comparative studies using different types of defrosting heaters, but no quantitative defrosting results or frosting conditions were provided. Melo et al. [84] carried out a series of experiments using a purposely built testing apparatus. Among three types of heaters, the highest heating efficiency of approximately 48% was obtained with a glass tube heater. A calrod heater seemed to be the mostly appropriate not only because of its efficiency, which was compatible with that of the other heaters, but also due to its low cost and easy installation. Using air bypass circulation and electric heaters, Yin et al. [85] comparatively and experimentally studied a new cold storage method with different defrosting heaters and air circulation modes. When using the new method, the defrosting duration was shortened by 62.1% and the defrosting energy consumption reduced by 61.0%. The defrosting efficiency was increased to 77.6%, which was 2.93 times that using a conventional electric heating defrosting method. However, the additional electrical energy required to melt frost is high-quality energy. Meanwhile, an ASHP unit is out of operation during defrosting,

26

Defrosting for Air Source Heat Pump

which results in the interruption of indoor air heating. Therefore, this method is limited in the civilian application domain such as household air conditioning and residual ASHP units.

2.3.3 Hot water spraying The hot water spraying defrosting (HWSD) method can be applied where hot water for defrosting is available. During defrosting, the indoor and outdoor fans are turned off, with the hot water spraying onto the outdoor coil. Thus, frost could be melted and the melted frost flows away with the water. However, only limited reported studies can be identified, including a patent from Tanker and Abdel-Wahed’s experimental study on applying the hot water spraying defrosting method to a horizontal flat-plate surface [12]. Obviously, the availability of hot water limits the application, especially for low cost and continuous hot water supply. In addition, at the termination of the hot water spraying defrosting, some water may be retained on the coil surface due to surface tension. The retained water would degrade the operating performance of an ASHP unit when operated at the heating mode. Hence, this defrosting method is not widely applied.

2.3.4 Hot gas bypass Hot gas bypass defrosting (HGBD) is mainly applied to industrial ASHP units. The superheated refrigerant vapor discharged from the compressor is directed into an evaporator, or outdoor coil, bypassing a condenser and an expansion device. The latent heat of the condensation of refrigerant vapor is used as the heat source; however, the sensible heat of highly superheated refrigerant vapor may also be used [86]. On the basis of hot gas bypass defrosting, Fu et al. [87] divided an outdoor coil into two parts, a front part and a rear part, which were used as an evaporator and a condenser, respectively, during defrosting. It was indicated that energy was used more efficiently, and thus the defrosting duration was shorter and the defrosting loss less than those using reverse cycle defrosting. A novel dual hot gas bypass defrosting method was also developed to remove frost from the outdoor coil of an ASHP unit [88], showing that the proposed method could overcome the main disadvantages for RCD and hot gas bypass defrosting. However, the defrosting duration is always very long because the energy use for defrosting comes from the power input to the compressor. In addition, it is easy for the compressor to suck in liquid during a hot gas bypass defrosting process due to insufficient energy supply, which impacts badly on the safety of the compressor [27]. Finally, the hot gas bypass defrosting method is mostly used in industry units.

2.3.5 Reverse cycle When an ASHP unit is operated at RCD mode, its outdoor coil acts as a condenser and its indoor coil as an evaporator [75, 76]. During defrosting, the normal operation cycle during heating for an ASHP unit is reversed by using a four-way valve, and hot gas is

Previous related work: A review

27

pumped into the outdoor coil to melt the frost. When the frost is melted and drained away from the coil, the ASHP unit returns to heating operation. Apart from requiring a four-way valve, the RCD does not need anything else. That means the system is simple and easily installed [77, 78]. The energy used for RCD mainly comes from four sources: (1) the thermal energy of indoor air, (2) the energy stored in the indoor coil’s metal, (3) the electricity input to the indoor air fan, and (4) the electricity input to the compressor. The defrosting energy is for (1) heating the outdoor coil’s metal, (2) melting the frost, (3) heating the melted frost, (4) vaporizing the retained water, and (5) heating the ambient air. With sufficient energy supply, the duration of an RCD operation can be much shorter than that of hot gas bypass defrosting. In fact, RCD has been the most widely used standard defrosting method for ASHP units for many years. To clearly distinguish the five defrosting methods, their operation differences and evaluation results are summarized in Tables 2.5 and 2.6, respectively. To shorten a defrosting duration, turning on an indoor air fan is only required when using reverse cycle defrosting; when using compress shutdown defrosting, the outdoor air fan has to be operated. When using hot gas bypass defrosting and reverse cycle defrosting, the compressor should be turned on to supply enough defrosting energy. Although the stability for the compress or shutdown defrosting and electric heating defrosting is highly rated, the former results in a poor defrosting effect and the later costs much highquality energy. Consequently, the two defrosting methods have the lowest

Table 2.5 Operation differences for the five defrosting methods No.

Methods

Indoor fan

Outdoor fan

Compressor

Thermal source

1 2 3 4 5

CSDD EHD HWSD HGBD RCD

Off Off Off Off On

On Off Off Off Off

Off Off Off On On

Ambient air Electricity Hot water Electricity Electricity

Table 2.6 Evaluation results for the five defrosting methods

No.

Methods

System complexitya

1 2 3 4 5

CSDD EHD HWSD HGBD RCD

! " " " "

", increased; !, unchanged. 3, the best; 1, the worst.

a

b

System stabilityb

Defrosting effectb

Defrosting energy used

Valuation indexb

3 3 2 2 1

1 2 2 2 3

2 1 2 3 3

1 1 2 2 3

28

Defrosting for Air Source Heat Pump

comprehensive evaluation index and are not widely applied. Hot water spraying defrosting is limited by its inconvenience, discontinuity, and high cost for the hot water supply. Hot gas bypass defrosting has a good operational stability and defrosting effect. However, more electric energy is needed than that for reverse cycle defrosting. Overall, the comprehensive evaluation index for RCD is the highest.

2.4

Improvements for reverse cycle defrosting

Currently, the most widely used standard defrosting method is reverse cycle defrosting. The RCD system is simple and can be easily controlled, but an RCD operation is a complex process involving spatial and time variations of the temperatures of the refrigerant, metal, and air as well as many other indeterminate factors resulting from transient cycling, which may last for only a few minutes [27]. Also, an energy balance on the airside of an outdoor coil is complex due to the fact that the energy extracted from the hot refrigerant gas is utilized in five different ways. As previously discussed, defrosting an ASHP unit consumes energy and causes undesirable fluctuations of the indoor air temperature and other operational problems, such as lowpressure cut off or wet compression. Therefore, using both experimental and numerical approaches, extensive research work has been carried out to improve the operating performance of ASHP units during reverse cycle defrosting.

2.4.1 Experimental studies 2.4.1.1 Basic component optimization Due to the operational characteristics of reverse cycle defrosting, investigations on the optimization of the basic component in a refrigerant cycle have been conducted. First, a thermal expansion valve (TEV) was used to experimentally investigate the transient RCD performance for a nominal 3-ton residential ASHP unit. It was found that the accumulator of the ASHP unit and the TEV impacted significantly on the unit’s dynamic responses. With either a scroll or a reciprocating compressor, the cycle performances during RCD for an ASHP unit were further experimentally compared [12]. Second, the effects of an accumulator in the suction line on the frosting/defrosting performance were investigated, showing that the removal of the accumulator produced a 10% reduction in defrosting duration but a 25% reduction in the integrated cyclic COP [27]. Also, using a refrigerant charge compensator instead of an accumulator led to an increase in refrigerant flow rate and higher suction and discharge pressures of the compressor in an ASHP unit during its defrosting [89]. The defrosting effect was improved by the addition of a compensator with an increased circulation. However, the effect of basic component optimization is limited, and other enhancement methods such as using a thermal energy storage (TES) system in an ASHP unit are being studied.

Previous related work: A review

29

2.4.1.2 PCM-TES-based RCD During reverse cycle defrosting, the indoor air fan in an ASHP unit is normally switched off to avoid blowing cold air directly to a heated indoor space, affecting the thermal comfort of the occupants [90]. A defrosting operation may also occur at night when the ambient air temperature is lower than that during the day; sleeping thermal comfort may also be negatively affected [91–93]. The energy available from the indoor coil is basically that stored in the coil metal, but there is an insignificant amount of energy available from the indoor air because of a negligibly small airside convective heat coefficient resulting from a deenergized indoor air fan during defrosting. Consequently, low-pressure cut off or wet compression may take place, which may cause the ASHP unit to shut down and possibly damage the compressor. To avoid these aforementioned problems, the technologies of TES and phase change materials (PCM) may be applied due to the advantage of high-density energy storage [94]. First, DX40 was used as a thermal storage material [95] in a heat source tank for defrosting, and a higher defrosting efficiency was reached after using PCM-TES. Then, inorganic PCM such as CaCl2  6H2O was used in an ASHP unit [96, 97]. As shown, this PCM-TES-based defrosting method could help achieve improved indoor thermal comfort with a shorter defrosting period and a higher indoor supply air temperature during reverse cycle defrosting. The same conclusions were given in similar studies [98, 99]. As summarized in Table 2.7, PCM-TES-based RCD was widely studied in 2000–2017.

2.4.1.3 Airflow and refrigerant distribution adjustment Mal-distribution of refrigerant or airflow might result in uneven defrosting, thus degrading the defrosting performance. Aganda et al. [103] compared the predicted and experimental heat transfer performances for a finned tube outdoor coil, and found that airflow mal-distribution reduced the performances of an evaporator circuit [104]. With refrigerant flow controlled by a TEV, the worst-performing circuit affected the performance of the entire outdoor coil by as much as 35%. Kim et al. [105, 106] experimentally and numerically investigated a hybrid-individual degree of superheat control method for refrigerant flow balancing in a multicircuit evaporator: upstream vs downstream flow control. The results showed that the upstream refrigerant flow control consistently outperformed the downstream refrigerant flow control, and recovered most of the loss in cooling capacity and COP due to nonuniform airflow distribution. Based on these conclusions, they utilized the model to further evaluate the effects of uneven air and refrigerant flow distributions and the benefits of upstream hybrid control during defrosting for an ASHP unit [106]. Hence, adjusting airflow and/or refrigerant distribution could improve defrosting performance.

2.4.1.4 Sensible heat defrosting method To avoid adverse shock and “oil rush,” which were commonly seen in conventional RCD operations, a sensible heat-defrosting method was proposed and numerically investigated by Liang et al. [86], by using a self-organizing fuzzy controller in an

30

Table 2.7 Studies of PCM-TES based RCD (2000–2017)

Year

Author

PCM type

TES type

1

2017

Qu et al. [99]

Water

Shell and tube

Refrigerant

2

2015

Dong et al. [98]

Finned tube

Refrigerant

3

2015

Wang et al [14]

65 mol% capric acid with 35 mol% lauric acid CaCl26H2O

Finned tube

Refrigerant

4

2014

Zhang et al. [100]

Finned tube

Refrigerant

5

2014

Daikin [27]

Finned tube

6

2012

Qu et al. [96]

Polyethylene glycol Refrigerant

65 mol% capric acid with 35 mol% lauric acid Polyethylene glycol CaCl26H2O

Shell and tube

Conclusions Defrosting duration was shortened by 71.4%– 80.5%, and defrosting energy consumption reduced by 65.1%–85.2% Defrosting duration was reduced by 60%, compressor suction pressure and temperature increased by 0.28 MPa and 13.6°C, respectively, and COP increased by about 1.8 The regeneration duration was 4 min longer at
3°C than at 3°C, but the regeneration efficiency was reduced by 15% due to more heat stored in the PCM-TES during heating mode Compared to the traditional method, the defrosting duration was shortened by 65% and the total energy consumption was less than 27.9% The heating capacity was improved by about 10% and the COP by 50% Defrosting duration was reduced by 36.4%

Defrosting for Air Source Heat Pump

No.

Heat transfer fluid

2012

Wu et al. [101]

75 wt.% paraffin and 25 wt.% expanded graphite

Tube and fin

Refrigerant

8

2011

Hu et al. [96]

CaCl26H2O

Shell and tube

Refrigerant

9

2011

Dong et al. [102]

Shell and tube

Refrigerant

10

2006

Chen et al. [95]

CaCl26H2O with SrCl26H2O DX40

Plate table

Air

When the outlet temperature drops to the minimum temperature for usage (40°C), the efficiency of a water tank with PCM is 94.23% while the efficiency of a water tank without PCM is 97.85%, meaning a water tank with PCM has a lower energetic efficiency than that without PCM Defrosting duration was reduced by  3 min or 38%, compressor suction pressure increased by about 200 kPa, and the mean indoor coil surface temperature increased about 25°C Defrosting duration was reduced by 60% and energy consumption during defrosting reduced by 48.1% Indoor air temperature was kept stable, and the indoor thermal comfort level improved

Previous related work: A review

7

31

32

Defrosting for Air Source Heat Pump

Table 2.8 Evaluation results of four defrosting enhancement methods Methods Basic component optimization PCM-TESbased RCD Airflow and refrigerant distribution adjustment Sensible heat defrosting method

Initial costa

Running costa

System complexityb

Defrosting effectb

Evaluation indexb

"

!

!

2

2

"

!

"

3

3

"

"

!

2

2

"

!

"

1

1

", increased; !, unchanged. 3, the best; 1, the worst.

a

b

ASHP unit. For this defrosting method, the four-way valve did not act when the ASHP unit changed its operation mode from heating to defrosting. Therefore, it is similar to the hot gas bypass defrosting method. Refrigerant was discharged from the compressor and passed through a hot gas solenoid valve and a discharge conduit. After being throttled by the expansion valve, the refrigerant with high temperature and low pressure flowed into the outdoor coil and exchanged heat with the outer frost layer. After that, the refrigerant passed through the accumulator and was sent back to the compressor for recompression. In this way, the cycle continued. However, in order to ensure the normal operation of the unit, it should be guaranteed that there would not be any condensation of the refrigerant. Moreover, the defrosting duration would be prolonged because the only energy supply for defrosting is from the electricity input to the compressor, and thus the indoor thermal comfort is adversely affected [90]. Consequently, it is more suitable for industry applications. The evaluation results of four improvement methods are listed in Table 2.8 by considering and comparing the initial cost, running cost, complexity, and defrosting effects. As seen, the PCM-TES-based RCD is the most recommended method to enhance defrosting for an ASHP unit. The sensible heat defrosting method, on the contrary, is the lowest rated.

2.4.2 Theoretical studies As a nonlinear, moving-boundary, variable-density, multidimensional, phasecontinuously changing, dynamic heat- and mass-transfer process, defrosting is a common and complicated physical phenomenon. During a defrosting operation, the frost

Previous related work: A review

33

on the surface of an outdoor coil will not necessarily be melted uniformly throughout the surface. The frost over certain parts of the coil surface remains attached to the coil surfaces until it is completely melted and sublimated while the frost over other parts may be partially melted. It would then detach from the coil surface, falling down due to gravity to the coil surfaces at lower levels or to a drainage tray. Two types of models were developed, one system-based and the other the outdoor coil only.

2.4.2.1 Defrosting model system based From the open literature available, modeling a defrosting process has attracted lots of research attention. The early modeling work focused mainly on the outdoor coils of simple geometry, such as finite slabs, a horizontal flat plate, or a flat-plate cooler. Then, on a cylindrical coil cooler, a semiempirical model for electric defrosting was presented, and an analytical model was developed to predict the evaporation, sublimation, and melting rates during defrosting. A moving boundary technique was used and a defrosting process was divided into two stages, premelting and melting [12]. Later, Alebrahim and Sherif [107] reported an electric defrosting model for a finned-tube outdoor coil using the enthalpy method to predict the defrosting duration and frost surface temperature profiles. Thereafter, a number of studies on modeling a defrosting process in ASHP units were carried out. Noticeably, an RCD model for an outdoor coil was developed where the process of frost melting on the surface of an outdoor coil was subdivided into four stages: preheating, melting, vaporizing, and dry heating [12], as illustrated in Fig. 2.2. A number of heat and mass transfer parameters required for simulating defrosting performance, e.g., the maximum mass of surface water, the free-convection air film conductance, the air/water film conductance, and the surface water vaporization coefficient, were however experimentally determined. On the basis of the aforementioned model, a validated defrosting model for an ASHP unit using a capillary tube was developed by Liu et al. [108]. Distributed modeling was used for both the evaporator and condenser because of their importance during reverse cycle defrosting.

Fig. 2.2 Schematic diagrams illustrating a defrosting process of an outdoor coil cell.

34

Defrosting for Air Source Heat Pump

2.4.2.2 Defrosting model outdoor coil only Although the aforementioned defrosting models [12, 107, 108] were developed and used in studying defrosting performance, none of them considered the negative effects of the downward flowing of melted frost due to gravity along the surface of an outdoor coil on defrosting performance, by either assuming a stable water layer or no water retention on the coil surface. In 2012, Qu et al. [109] reported on a modeling analysis where a semiempirical model for the defrosting on the airside of a four-circuit outdoor coil in an ASHP unit was developed. The negative effects of melted frost on defrosting performance were considered and quantitatively studied using this model, which was different from the aforementioned defrosting modeling studies. It was further predicted that if the melted frost could be drained away locally, the defrosting efficiency for the ASHP unit could be increased by up to 18.3%. A similar energy consumption ratio for defrosting was also reported by Dong et al. [110] in a study on the energy consumption analysis for vaporizing the melted frost and heating the ambient air during RCD in an ASHP unit. To clearly understand the modeling study reported by Qu et al., the main equations used in the two models are listed in Table 2.9. Being the most popular defrosting method for ASHP units, RCD has attracted much research attention. A series of related experimental investigations was undertaken, covering optimizing the original component, the use of PCM-TES-based reverse cycle defrosting, adjusting the air and refrigerant distribution, and the sensible heat defrosting method. Among all the experimental studies on RCD optimization, the PCM-TES-based RCD was the most strongly recommended, with its advantages of easy and low-cost installation and a better defrosting effect. However, both frosting and defrosting performances should be examined when an ASHP unit is optimized with any defrosting enhancement methods. On the other hand, numerical studies on RCD were conducted, with defrosting models developed. For the system-based defrosting model, component optimization and the effects on defrosting efficiency were the hot issues. However, in outdoor coil only defrosting models, the defrosting process was always described in greater detail. For example, a defrosting process was divided into only two stages, premelting and melting [12], in a system-based defrosting model while it was divided into three stages in Qu’s defrosting model [109].

Table 2.9 Main energy equations used in two defrosting modeling studies [109] Stage

Description

Main equation for energy balance

First stage

Frost melting without water flow Frost melting with water flow Water layer vaporizing

qj ¼ Lsf mf , j + cp

Second stage Third stage

dðMw, j Tw, j Þ dt

dT w, j qj + cp mw, j
1 Tw, j
1 ¼ Lsf mf , j + cp Mw, max
dt + cp mw, j Tw, j + hc, w Tw, j
Ta Af
a

dðMw, j Tw, j Þ + hc, w Tw, j
Ta Aw
a + hc, d Tr, j
Ta Ad
a + mv, j Lv q j ¼ cp dt

Previous related work: A review

2.5

35

Defrosting control strategy

Frosting would seriously affect the operating performance of an ASHP unit by reducing its COP and output heating capacity. Hence, periodic defrosting is necessary. Defrosting initiation and termination control strategies impact system reliability and energy efficiency [111]. Studies on selecting suitable control parameters for defrosting initiation and termination were therefore carried out.

2.5.1 Defrosting initiation There are two types of control strategies to start a defrosting operation: time-based and demand-based. For the former, the defrosting start is controlled by a preset timer. Due to the advantages of simplicity and low cost, many of the earlier ASHP units employed the time-based defrosting start method. Usually, for every 60–90 min of frosting operation, a defrosting operation would be initiated [12, 96, 110]. However, the performances of these earlier ASHP units could well suffer from some unnecessary defrosting operations, resulting in a degraded operational efficiency. A time-based defrosting start results in two typical mal-defrosting problems: unnecessary defrosting when no or little frost has accumulated on the surface of the outdoor coil, and no defrosting when frost presents. As reported from a field test experiment, no maldefrosting was initiated when more than 60% of the outdoor coil surface area was frosted after 5 days of operation [112]. During this 5-day frosting period, the system COP was significantly degraded at only 2.3 under an environmental temperature of 7.9°C. Comparing the test data before and after frosting, it was found that such a mal-defrosting phenomenon decreased the system COP by up to 40.4% and the heating capacity by up to 43.4%. To avoid mal-defrosting, and thus to improve the energy efficiency of the system, the temperature and/or pressure parameters were considered in the time-based defrosting start method. When the parameters reach preset values, the timing starts. Currently, these time-based defrosting start methods are widely used in applications. A further control strategy applied to initiate a defrosting cycle was first proposed by Eckman, and called the demand defrosting start [27]. As the name suggests, it could start a defrosting operation only when needed. A demand defrosting start control strategy defrosts the display cabinet when sufficient frost is formed, adversely affecting the operating performance. The use of the demand-based defrosting start method would lead to: (1) better temperature control, (2) increased product quality and life, (3) reduced product losses, and (4) significant energy savings. Therefore, when applying this strategy, an ASHP unit would start defrosting only when an adequate frost buildup was detected. Thus, it was important to accurately detect the presence and growth of frost. A number of frost-detecting techniques have been developed over the years, including: (1) measuring the thermal conductivity of ice, (2) calculating the air pressure differential across an evaporator, (3) calculating the degree of refrigerant superheat [113], (4) sensing the temperature difference between the air and evaporator surface, and (5) sensing the outdoor air fan power [12, 27].

36

Defrosting for Air Source Heat Pump

Based on the aforementioned direct and indirect types of frost accumulation sensing technologies, recent defrosting start control strategies include: (1) measuring the ice thickness using a holographic interferometry technique [114], (2) measuring the frost surface temperature by an infrared thermometer [115], (3) sensing refrigerant flow instability [116], (4) sensing frost using a photocoupler, photooptical systems, or fiberoptic sensors [12], (5) modeling the amount of frost on the coil surface by applying neural networks [27], and (6) calculating the effective mass-flow fraction by the fin surface temperature [117]. Notably, the principle of photoelectric technology for frost detection using a photocoupler was systematically investigated by Wang et al. using experimental, numerical, and theoretical approaches [118–121]. Based on a frosting map for ASHP units, a novel temperature-humidity-time defrosting control strategy was further proposed by them [122]. In their given frosting map, there are three typical regions: (1) the frosting region, (2) the condensing region, and (3) the nonfrosting region. The frosting region was further divided into three frosting zones: (1) the server frosting zone, (2) the moderate frosting zone, and (3) the mild frosting zone. In addition, each zone established two subzones, I and II. The six subzones were divided and surrounded with Curves A to E and Line 1. Therefore, the working condition of an ASHP unit will be clearly presented in this frosting map, and thus the corresponding control strategy could be suitably made. This frosting map is fundamental and meaningful for the intelligent control design work of the ASHP units.

2.5.2 Defrosting termination Mal-defrosting was found and reported in many studies, which was described by Wang et cal. [123] as follows: a defrosting operation is carried out either a long time after a “critical” level of frost has been reached or when it is not necessary. Clearly, this description focuses on the start of defrosting, but the conditions where a defrosting operation is terminated earlier or later than a “critical termination time” are not considered. With respect to defrosting termination, related research is much less seen. It should be noted that for defrosting an ASHP unit, a complete defrosting process covers both melting the frost and drying the coil surface. Otherwise, once the ASHP unit returns to heating operation, the melted frost retained on the outdoor coil surface would become ice. This may change the structure of a frost layer, increasing the density and enhancing the thermal conductivity of the frost layer [124]. During reverse cycle defrosting, not only is a great deal of energy for melting the frost and vaporizing the melted frost off the outdoor coil surface consumed, but also the occupants’ thermal comfort may be adversely affected [125]. Therefore, shortening a defrosting period is always one of the defrosting control purposes for ASHP units. For example, Chinese Standard GB/T 7725-2004 specifies that the defrosting duration for an ASHP unit should not exceed 20% of its total working hours. Therefore, it is meaningful to accurately terminate a defrosting operation. In practical applications, an RCD operation can be terminated based on the tube or fin surface temperature of an outdoor coil, the refrigerant pressure difference across an

Previous related work: A review

37

outdoor coil, or the defrosting operation time duration [12]. Although the control strategy of a time-based defrosting start is widely used, a defrosting operation is not always terminated based on time duration. Currently, the most-used method for terminating a defrosting operation is based on the tube surface temperature of an outdoor coil. A temperature sensor is usually placed on the tube surface at the exit of the lowest liquid-line circuit of a vertically installed multicircuit outdoor coil [96]. A defrosting operation will be terminated once a preset temperature is reached. It is obvious that when the preset temperature is higher or lower, a defrosting duration would be prolonged or more residual water is left, respectively. Both result in potential energy waste for an ASHP unit, or even degrading the indoor thermal comfort. However, there is no standard defrosting termination temperature (DTT) or even a fixed range given in the application, due to the diversity of equipment and operating climates. Different DTT settings were used and reported in 2000–2017, from 10°C [126] to 50°C [127], as summarized in Table 2.10. Obviously, the temperature range covering 40°C is too big. However, in the open literature, no relative study on defrosting termination temperature was reported, or was this fundamental problem even pointed out. Different strategies to start and end defrosting were studied to improve the ASHP system operating performance at an entire frosting-defrosting cycle. Currently, developed time-based control strategies are widely used to start a defrosting cycle, due to their advantages of simplicity and reliability, although the accuracy remains

Table 2.10 DTT settings for ASHP units (2000–2017) Item

DTT (°C)

Circuit number

Capacity (kW)

1 2 3

10 12 15

12 2 4

4 5 6 7 8 9 10 11 12

18 20 20 22 24 24 25 26 30

4 / 12 12 12 4 2 2 4

13 14 15

33 35 50

2 2 2

55 (Cooling) 8.82 (Cooling) 55–350 (Heating) 6.8 (Heating) / 16 (Cooling) 50 (Cooling) 55 (Cooling) 6.8 (Heating) 37.5 (Heating) 4.8 (Heating) 0.88 (Compressor) 2.8 (Cooling) 2.5 (Heating) 2.5 (Heating)

Year

Author

2009 2004 2013

Huang et al. [126] Ding et al. [128] Wang et al. [129]

2010 2005 2011 2004 2007 2012 2015 2012 2003

Qu et al. [90] Cho et al. [130] Choi et al. [88] Huang et al. [131] Huang et al. [132] Qu et al. [96] Dong et al. [98] Dong et al. [110] Liu et al. [23]

2012 2011 2011

Dong et al. [133] Dong et al. [102] Hu WJ [127]

38

Defrosting for Air Source Heat Pump

questionable. For an ASHP unit with a multicircuit outdoor coil, as summarized, terminating an RCD operation based on the tube surface temperature at the exit of the lowest circuit is the most widely used. Also, it is still questionable in practical applications due to the randomness of the DTT setting at a wide temperature range. More defrosting start and termination control strategies should be further studied.

2.6

Concluding remarks

Recent studies on frost/defrosting for ASHP units are reviewed in this chapter based on papers published in 2000–2017. The current research status is summarized, and the finished work, unsolved problems, potential practical applications, and their barriers identified. It is demonstrated that the 11 listed frost-suppression measures, preheating the inlet air with waste heat, and a fin surface coating treatment with new coating materials are highly recommended. To resolve the frosting problem, five typical defrosting methods are classified and discussed. Being the most widely used defrosting method, RCD is extensively evaluated and its four enhancement methods are thereby listed. A defrosting operation starts based on time duration, at 60–90 min, and one or two other parameters. It is terminated when the tube surface temperature at the exit of the lowest circuit in a multicircuit outdoor coil reaches a preset value within the range of 10–50°C.

References [1] Nishimura T. “Heat pumps-status and trends” in Asia and the Pacific. Int J Refrig 2002;25 (4):405–13. [2] Wang Q, He W, Liu YQ, Liang GF, Li JR, Han XH, Chen GM. Vapor compression multifunctional heat pumps in China: a review of configurations and operational modes. Renew Sustain Energy Rev 2012;16(9):6522–38. [3] Mohanraj M, Jayaraj S, Muraleedharan C. Applications of artificial neural networks for refrigeration, air-conditioning and heat pump systems—a review. Renew Sustain Energy Rev 2012;16:1340–58. [4] Yao Y, Ma ZL. Modeling and analyzing of airside heat exchanger under frosting in air source heat pump water chiller/heater. Int J HIT 2003;35:781–3. [5] Neal E. Heat pumps—application for heating conservation and heat recovery. Prog Energy Combust Sci 1983;9:179–97. [6] Staffell I, Brett D, Brandon N, Hawkes A. A review of domestic heat pumps. Energy Environ Sci 2012;5:9291–306. [7] Nasr MR, Fauchoux M, Besant RW, Simonson CJ. A review of frosting in air-to-air energy exchangers. Renew Sustain Energy Rev 2014;30:538–54. [8] Zeng C, Liu S, Shukla A. A review on the air-to-air heat and mass exchanger technologies for building applications. Renew Sustain Energy Rev 2017;75:753–74. [9] Patil MS, Seo JH, Lee MY. Heat transfer characteristics of the heat exchangers for refrigeration, air conditioning and heat pump systems under frosting, defrosting and dry/wet conditions—a review. Appl Therm Eng 2017;113:1071–87. [10] Sheng W, Liu PP, Dang CB, Liu GX. Review of restraint frost method on cold surface. Renew Sustain Energy Rev 2017;79:806–13.

Previous related work: A review

39

[11] Amer M, Wang CC. Review of defrosting methods. Renew Sustain Energy Rev 2017;73:53–74. [12] Song MJ, Dong JK, Wu CL, Jiang YQ, Qu ML. Studies on improving frosting and defrosting operation performances for air source heat pump: review and outlooks. HKIE Trans 2017;24(2):88–98. [13] Wang SW, Liu ZY. A new method for preventing heat pump from frosting. Renew Energy 2005;30:753–61. [14] Wang FH, Wang ZH, Zheng YX, Lin Z, Hao PF, Huan C, Wang T. Performance investigation of a novel frost-free air-source heat pump water heater combined with energy storage and dehumidification. Appl Energy 2015;139:212–9. [15] Wang ZH, Wang FH, Ma ZJ, Song MJ. Numerical study on the operating performances of a novel frost-free air-source heat pump unit using three different types of refrigerant. Appl Therm Eng 2017;112:248–58. [16] Wang ZH, Wang FH, Wang XK, Ma ZJ, Wu XZ, Song MJ. Dynamic character investigation and optimization of a novel air-source heat pump system. Appl Therm Eng 2017;111:122–33. [17] Su W, Zhang XS. Performance analysis of a novel frost-free air-source heat pump with integrated membrane-based liquid desiccant dehumidification and humidification. Energy Build 2017;145:293–303. [18] Zhang L, Saikawa M, Fujinawa T. Experimental study on heat and mass transfer characteristics for a desiccant-coated fin-tube heat exchanger. Int J Heat Mass Transf 2015;89:641–51. [19] Wang ZH, Zheng YX, Wang FH, Wang XK, Lin Z, Li JC, Huan C. Experimental analysis on a novel frost-free air-source heat pump water heater system. Appl Therm Eng 2014;70 (1):808–16. [20] Jiang YQ, Fu HY, Yao Y, Yan L, Gao Q. Experimental study on concentration change of spray solution used for a novel non-frosting air source heat pump system. Energy Build 2014;68:707–12. [21] Zhang L, Fujinawa T, Saikawa M. A new method for preventing air-source heat pump water heaters from frosting. In J Refrig 2012;35(5):1327–34. [22] Zhang L, Dang CB, Hihara E. Performance analysis of a no-frost hybrid air conditioning system with integrated liquid desiccant dehumidification. Int J Refrig 2010;33(1):116–24. [23] Liu D, Zhao FY, Tang GF. Frosting of heat pump with heat recovery facility. Renew Energy 2007;32(6):1228–42. [24] Kwak K, Bai C. A study on the performance enhancement of heat pump using electric heater under the frosting condition: heat pump under frosting condition. Appl Therm Eng 2010;30:539–43. [25] Tang JC, Gong GC, Su H, Wu FH, Herman C. Performance evaluation of a novel method of frost prevention and retardation for air source heat pumps using the orthogonal experiment design method. Appl Energy 2016;169:696–708. [26] Kragh J, Jørgen R, Svend S. Mechanical ventilation with heat recovery in cold climates. In: Proceedings of the seventh Nordic symposium on building physics in Nordic countries Technical University of Denmark; 2005. p. 1–8. [27] Song MJ. An experimental and numerical study on improving defrosting performances for an air source heat pump unit having a multi-circuit outdoor coil. The Hong Kong Polytechnic University; 2014 [PhD thesis]. [28] Da Silva DL, Hermes CJ, Melo C. Experimental study of frost accumulation on fansupplied tube-fin evaporators. Appl Therm Eng 2011;31:1013–20.

40

Defrosting for Air Source Heat Pump

[29] Ye HY, Lee KS. Performance prediction of a fin-and-tube heat exchanger considering airflow reduction due to the frost accumulation. Int J Heat Mass Transf 2013;67:225–33. [30] Moallem E, Hong T, Cremaschi L, Fisher DE. Experimental investigation of adverse effect of frost formation on microchannel evaporators, Part 1: effect of fin geometry and environmental effects. Int J Refrig 2013;36:1762–75. [31] Yan QL, Zhu L, Zhang ME, Yan NX, Hui XY. Study on ultrasonic defrost technology of refrigeration fan. Trans Chin Soc Agric Mach 2003;34:74–85. [32] Li D, Chen Z, Shi M. Effect of ultrasound on frost formation on a cold flat surface in atmospheric air flow. Exp Therm Fluid Sci 2010;34:1247–52. [33] Wang D, Tao T, Xu G, Luo A, Kang S. Experimental study on frosting suppression for a finned-tube evaporator using ultrasonic vibration. Exp Therm Fluid Sci 2012;36:1–11. [34] Tan HH, Xu GH, Tao TF, Sun XQ, Yao WD. Experimental investigation on the defrosting performance of a finned-tube evaporator using intermittent ultrasonic vibration. Appl Energy 2015;158:220–32. [35] Fei Q, Mao YB. Research on defrost of air cooler by use of compressed air jet sweeping. Mech Electr Equipm 2000;5:8–12. [36] Sonobe N, Fukiba K, Sato S, Yoshimura Y. Method for defrosting heat exchangers using an air-particle jet. Int J Refrig 2015;60:261–9. [37] Watters RJ, O’Neal DL, Yang JX. Frost/defrost performance of a three-row fin staged heat pump evaporator. ASHRAE Trans 2002;108:318–29. [38] Yan WM, Li HY, Wu YJ, Lin JY, Chang WR. Performance of finned tube heat exchangers operating under frosting conditions. Int J Heat Mass Transf 2003;46:871–7. [39] Sommers AD, Jacobi AM. Air-side heat transfer enhancement of a refrigerator evaporator using vortex generation. Int J Refrig 2005;28:1006–17. [40] Yang DK, Lee KS, Song S. Fin spacing optimization of a fin-tube heat exchanger under frosting conditions. Int J Heat Mass Transf 2006;49:2619–25. [41] Lee M, Kim Y, Lee H, Kim Y. Air-side heat transfer characteristics of flat plate finnedtube heat exchangers with large fin pitches under frosting conditions. Int J Heat Mass Transf 2010;53:2655–61. [42] Park JS, Kim DR, Lee KS. Frosting behaviors and thermal performance of louvered fins with unequal louver pitch. Int J Heat Mass Transf 2016;95:499–505. [43] Yan WM, Li HY, Tsay YL. Thermofluid characteristics of frosted finned-tube heat exchangers. Int J Heat Mass Transf 2005;48:3073–80. [44] Dong H, Zhao RJ, Yun L, Yi DB. Effect of fin types of outdoor fan-supplied finned-tube heat exchanger on periodic frosting and defrosting performance of a residential air-source heat pump. Appl Therm Eng 2014;69:251–60. [45] Zhang P, Hrnjak PS. Air-side performance evaluation of three types of heat exchangers in dry, wet and periodic frosting conditions. Int J Refrig 2009;32:911–21. [46] Okoroafor EU, Newborough M. Minimizing frost growth on cold surfaces exposed to humid air by means of crosslinked hydrophilic polymeric coatings. Appl Therm Eng 2000;20:737–58. [47] Wu XM, Webb RL. Investigation of the possibility of frost release from a cold surface. Exp Therm Fluid Sci 2001;24:151–6. [48] Cai L, Wang RH, Hou PX, Zhang XS. Study on restraining frost growth at initial stage by hydrophobic coating and hygroscopic coating. Energy Build 2011;43:1159–63. [49] Jhee S, Lee KS, Kim WS. Effect of surface treatments on the frosting/defrosting behavior of a fin-tube heat exchanger. Int J Refrig 2002;25:1047–53. [50] Liu L, Wang HY, Zhang XH, Meng S, Ma CF. An experimental study on minimizing frost deposition on a cold surface under natural convection conditions by use of a novel anti-

Previous related work: A review

[51]

[52]

[53] [54] [55]

[56]

[57]

[58]

[59] [60] [61]

[62]

[63]

[64]

[65]

[66] [67]

41

frosting paint, Part I: anti-frosting performance and comparison with the uncoated metallic surface. Int J Refrig 2006;29:229–36. Chen X, Ma R, Zhou H, Zhou X, Che L, Yao S, Wang Z. Activating the microscale edge effect in a hierarchical surface for frosting suppression and defrosting promotion. Sci Rep 2013;3:2515. Zuo ZP, Liao RJ, Zhao XT, Song XY, Qiao ZW, Guo C, Zhuang AY, Yuan Y. Antifrosting performance of superhydrophobic surface with ZnO nanorods. Appl Therm Eng 2017;110:39–48. da Silva DL, Melo C, Hermes CJL. Effect of frost morphology on the thermal-hydraulic performance of fan-supplied tube-fin evaporators. Appl Therm Eng 2017;111:1060–8. Wu XM, Chu FQ, Ma Q, Zhu B. Frost formation and frost meltwater drainage characteristics on aluminum surfaces with grooved structures. Appl Therm Eng 2017;118:448–54. Moriya T, Manabe K, Tenjimbayashi M, Suwabe K, Tsuchiya H, Matsubayashi T, Navarrini W, Shiratori S. A superrepellent coating with dynamic fluorine chains for frosting suppression: effects of polarity, coalescence and ice nucleation free energy barrier. RSC Adv 2016;6:95. Sommers AD, Truster NL, Napora AC, Riechman AC, Caraballo EJ. Densification of frost on hydrophilic and hydrophobic substrates—examining the effect of surface wettability. Exp Therm Fluid Sci 2016;75:25–34. Zhao Y, Luo YT, Li J, Yin F, Zhu J, Gao XF. Condensate microdrop self-propelling aluminum surfaces based on controllable fabrication of alumina rod-capped nanopores. ACS Appl Mater 2015;7:11079–82. Kim D, Kim H, Kim SW, Kim DR, Lee KS. Experimental investigation of frost retardation for superhydrophobic surface using a luminance meter. Int J Heat Mass Transf 2015;87:491–6. Liang CH, Wang F, L€u Y, Wu CX, Zhang XS, Zhang YF. Experimental study of the effects of fin surface characteristics on defrosting behavior. Appl Therm Eng 2015;75:86–92. Wang F, Liang CH, Yang MT, Fan C, Zhang XS. Effects of surface characteristic on frosting and defrosting behaviors of fin-tube heat exchangers. Appl Therm Eng 2015;75:1126–32. Bharathidasan T, Vijay Kumar S, Bobji MS, Chakradhar RPS, Basu BJ. Effect of wettability and surface roughness on ice-adhesion strength of hydrophilic, hydrophobic and superhydrophobic surfaces. Appl Surf Sci 2014;314:241–50. Li LT, Wang W, Sun YY, Feng YC, Lu WP, Zhu JH, Ge YJ. Investigation of defrosting water retention on the surface of evaporator impacting the performance of air source heat pump during periodic frosting-defrosting cycles. Appl Energy 2014;135:98–107. Miljkovic N, Enright R, Nam Y, Lopez K, Dou N, Sack J, Wang EN. Jumping-dropletenhanced condensation on scalable superhydrophobic nanostructured surfaces. Nano Lett 2013;13:179–87. Kim K, Lee KS. Frosting and defrosting characteristics of surface-treated louvered-fin heat exchangers: effects of fin pitch and experimental conditions. Int J Heat Mass Transf 2013;60:505–11. Moallem E, Cremaschi L, Fisher DE, Padhmanabhan S. Experimental measurements of the surface coating and water retention effects on frosting performance of microchannel heat exchangers for heat pump systems. Exp Therm Fluid Sci 2012;39:176–88. Kulinich SA, Farzaneh M. Ice adhesion on super-hydrophobic surfaces. Appl Surf Sci 2009;255(18):8153–7. Huang LY, Liu ZL, Liu YM, Gou YJ, Wang JT. Experimental study on frost release on finand-tube heat exchangers by use of a novel anti-frosting paint. Exp Therm Fluid Sci 2009;33(7):1049–54.

42

Defrosting for Air Source Heat Pump

[68] Boreyko B, Chen CH. Self-propelled dropwise condensate on superhydrophobic surfaces. Phys Rev Lett 2009;103:184501. [69] Winandy EL, Lebrun J. Scroll compressor using gas and liquid injection: experimental analysis and modeling. Int J Refrig 2002;25:1143–56. [70] Zhnder M, Favrat D, Hohl H, Olivier C, Perevozchikov M. High performance air-water heat pump with extended application range for residential heating. In: 7th IEA Heat Pump Conference, Beijing, China; 2002. [71] Nguyen M, Hewitt N, Huang M. Performance evaluation of an air source heat pump using economized vapor injection compressor and flash tank coupled with capillary tubes. In: Proceedings of International Congress of Refrigeration, Beijing, China. ICR07-E2-1110; 2007. [72] Shao S, Wang B, Xia J, Ma G, Yan Q. Study on heat pump cycle for cold regions with twostage compression in one scroll compressor, In: 7th IEA Heat Pump Conference, Beijing, China; 2002. [73] Ma G, Zhao H. Experimental study of a heat pump system with flash-tank coupled with scroll compressor. Energy Build 2008;40:697–701. [74] Wang W, Ma ZL, Jiang YQ, Yao Y, Xu S, Yang Z. Field test investigation of a doublestage coupled heat pumps heating system for cold regions. Int J Refrig 2005;28(5):672–9. [75] Li Y, Chen G, Tang L, Liu L. Analysis on performance of a novel frost-free air-source heat pump system. Build Environ 2011;46:2051–9. [76] Heo J, Jeong MW, Kim Y. Effects of flash tank vapor injection on the heating performance of an inverter-driven heat pump for cold regions. Int J Refrig 2010;33:848–55. [77] Wang X, Hwang Y, Radermacher R. Two-stage heat pump with vapor-injected scroll compressor using R410A as a refrigerant. Int J Refrig 2009;32:1442–51. [78] Bertsch SS, Groll EA. Two-stage air-source heat pump for residential heating and cooling applications in northern U.S. climates. Int J Refrig 2008;31:1282–92. [79] Mei VC, Gao Z, Tomlinson JJ. Frost-less heat pump. ASHRAE Trans 2002;108:452–9. [80] Shang PJ. Effect of pre-start fans on defrosting characteristics of air source heat pump. Refrig Air Cond 2009;9(6):54–9. [81] Kim Y, Tikhonov A, Shin Y, Lee J. Experimental study on high performance defrosting heater for household refrigerator, In: 13th International Heat Conference, Sydney, Australia; 2006. [82] Bansal P, Fothergill D, Fernandes R. Thermal analysis of the defrost cycle in a domestic freezer. Int J Refrig 2010;33:589–99. [83] Ozkan DB, Ozil P, Inan C. Experimental investigation of the defrosting process on domestic refrigerator finned tube evaporators. Heat Transf Eng 2012;33:548–57. [84] Melo C, Knabben FT, Pereira PV. An experimental study on defrost heaters applied to frost-free household refrigerators. Appl Therm Eng 2013;51:239–45. [85] Yin HJ, Yang Z, Chen AQ, Zhang N. Experimental research on a novel cold storage defrost method based on air bypass circulation and electric heater. Energy 2012;37:623–31. [86] Liang CH, Zhang XS, Li XW, Chen ZQ. Control strategy and experimental study on a novel defrosting method for air-source heat pump. Appl Therm Eng 2010;30:892–9. [87] Fu WC, Guo XM, Tao XC. Experiment on a novel defrost cycle for air source heat pump. J Refrig 2009;30:16–24. [88] Choi HJ, Kim BS, Kang D, Kim KC. Defrosting method adopting dual hot gas bypass for an air-to-air heat pump. Appl Energy 2011;88:4544–55. [89] Wang ZY, Wang XX, Dong ZM. Defrost improvement by heat pump refrigerant charge compensating. Appl Energy 2008;85:1050–9.

Previous related work: A review

43

[90] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse-cycle defrosting method for air source heat pumps. Build Environ 2010;45:2354–61. [91] Mao N, Deng SM. Application of TOPSIS method in evaluating the effects of supply vane angle of a task/ambient air conditioning system on energy utilization and thermal comfort. Appl Energy 2016;180:536–45. [92] Mao N, Deng SM, Pan DM, Chen SJ. Experimental and numerical study on air flow and moisture transport in sleeping environments with a task/ambient air conditioning (TAC) system. Energy Build 2016;133:596–604. [93] Mao N, Pan DM, Deng SM. Computational fluid dynamics analysis of convective heat transfer coefficients for a sleeping human body. Appl Therm Eng 2017;117:385–96. [94] Niu FX, Ni L, Yao Y, Yu YB, Li HR. Performance and thermal charging/discharging features of a phase change material assisted heat pump system in heating mode. Appl Therm Eng 2013;58(1):536–41. [95] Chen C, Ouyang J, Wang XL, Liu M. New method of heat compensation for defrosting of air source heat pump system in winter. J Refrig 2006;27:37–40. [96] Qu ML, Xia L, Deng SM, Jiang YQ. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part I: experiments. Appl Energy 2012;91:122–9. [97] Hu WJ, Jiang YQ, Qu ML, Yao Y, Deng SM. An experimental study on the operating performance of a novel reverse-cycle hot gas defrosting method for air source heat pumps. Appl Therm Eng 2011;31:363–9. [98] Dong JK, Li S, Yao Y, Jiang YQ, Tian Y, Tian H. Defrosting performances of a multi-split air source heat pump with phase change thermal storage. Int J Refrig 2015;55:49–59. [99] Qu ML, Li TR, Deng SM, Fan YN, Li Z. Improving defrosting performance of cascade air source heat pump using thermal energy storage based reverse cycle defrosting method. Appl Therm Eng 2017;121:728–36. [100] Zhang L, Dong JK, Jiang YQ, Yao Y. A novel defrosting method using heat energy dissipated by the compressor of an air source heat pump. Appl Energy 2014;133:101–11. [101] Wu JH, Yang ZG, Wu QH, Zhu YJ. Transient behavior and dynamic performance of cascade heat pump water heater with thermal storage system. Appl Energy 2012;91 (1):187–96. [102] Dong JK, Jiang YQ, Yao Y, Zhang XD. Operating performance of novel reverse-cycle defrosting method based on thermal energy storage for air source heat pump. J Cent South Univ Technol 2011;18:2163–9. [103] Aganda AA, Coney JER, Farrant PE, Sheppard CGW, Wongwuttanasatian T. A comparison of the predicted and experimental heat transfer performance of a finned tube evaporator. Appl Therm Eng 2000;20:499–513. [104] Aganda AA, Coney JER, Sheppard CGW. Airflow maldistribution and the performance of a packaged air conditioning unit evaporator. Appl Therm Eng 2000;20:515–28. [105] Kim JH, Braun JE, Groll EA. A hybrid method for refrigerant flow balancing in multicircuit evaporators: upstream versus downstream flow control. Int J Refrig 2009;32:1271–82. [106] Kim JH, Braun JE, Groll EA. Evaluation of a hybrid method for refrigerant flow balancing in multi-circuit evaporators. Int J Refrig 2009;32:1283–92. [107] Alebrahim AM, Sherif SA. Electrical defrosting analysis of a finned tube evaporator coil using the enthalpy method. Proc Inst Mech Eng C: J Mech Eng Sci 2002;216:655–73. [108] Liu ZQ, Tang GF, Zhao FY. Dynamic simulation of air-source heat pump during hot-gas defrost. Appl Therm Eng 2003;23:675–85.

44

Defrosting for Air Source Heat Pump

[109] Qu ML, Pan DM, Xia L, Deng SM, Jiang YQ. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part II: modeling analysis. Appl Energy 2012;91:274–80. [110] Dong JK, Deng SM, Jiang YQ, Xia L, Yao Y. An experimental study on defrosting heat supplies and energy consumptions during a reverse cycle defrost operation for an air source heat pump. Appl Therm Eng 2012;37:380–7. [111] Hewitt N, Huang MJ. Defrost cycle performance for a circular shape evaporator air source heat pump. Int J Refrig 2008;31:444–52. [112] Wang W, Feng YC, Zhu JH, Li LT, Guo QC, Lu WP. Performances of air source heat pump system for a kind of mal-defrost phenomenon appearing in moderate climate conditions. Appl Energy 2013;112:1138–45. [113] Jiang YQ, Dong JK, Qu ML, Deng SM, Yao Y. A novel defrosting control method based on the degree of refrigerant superheat for air source heat pumps. Int J Refrig 2013;36 (8):2278–88. [114] Hao YL, Iragorry J, Castro D, Tao YX, Jia S. Microscopic characterization of frost surface during liquid-ice phase change period. Proceedings of the 2002 ASME international mechanical engineering congress & exposition, New Orleans, LA. No. IMECE200232797. 2002. [115] Iragorry J, Tao YX. Frost temperature relations for defrosting sensing system. J Heat Transf 2005;127:344–9. [116] Lawrence JMW, Evans JA. Refrigerant flow instability as a means to predict the need for defrosting the evaporator in a retail display freezer cabinet. Int J Refrig 2008;31:107–12. [117] Kim MH, Lee KS. Determination method of defrosting start-time based on temperature measurements. Appl Energy 2015;146:263–9. [118] Byun JS, Jeon CD, Jung JH, Lee J. The application of photo-coupler for frost detecting in an air-source heat pump. Int J Refrig 2006;29:191–8. [119] Xiao J, Wang W, Zhao YH, Zhang FR. An analysis of the feasibility and characteristics of photoelectric technique applied in defrost-control. Int J Refrig 2009;32:1350–7. [120] Xiao J, Wang W, Guo QC, Zhao YH. An experimental study of the correlation for predicting the frost height in applying the photoelectric technology. Int J Refrig 2010;33:1006–14. [121] Ge YJ, Sun YY, Wang W, Zhu JH, Li LT, Liu JD. Field test study of a novel defrosting control method for air-source heat pumps by applying tube encircled photoelectric sensors. Int J Refrig 2016;66:133–44. [122] Zhu JH, Sun YY, Wang W, Ge YJ, Li LT, Liu JD. A novel temperature-humidity-time defrosting control method based on a frosting map for air-source heat pumps. Int J Refrig 2015;54:45–54. [123] Wang W, Xiao J, Guo QC, Lu WP, Feng YC. Field test investigation of the characteristics for the air source heat pump under two typical mal-defrost phenomena. Appl Energy 2011;88:4470–80. [124] Lee YB, Ro ST. An experimental study of frost formation on a horizontal cylinder under cross flow. Int J Refrig 2001;24:468–74. [125] Mao N, Chan MY, Pan DM, Deng SM. Computational fluid dynamics (CFD) modelling of air flow field, mean age of air and CO2 distributions inside a bedroom with different heights of conditioned air supply outlet. Appl Energy 2016;164:906–15. [126] Huang D, Li Q, Yuan X. Comparison between hot-gas bypass defrosting and reversecycle defrosting methods on an air-to-water heat pump. Appl Energy 2009;86:1697–703. [127] Hu WJ. Study on the dynamic characteristic of a novel PCM based defrosting for air source heat pump. (PhD thesis) Harbin Institute of Technology (in Chinese); 2010.

Previous related work: A review

45

[128] Ding YJ, Chai QH, Ma GY, Jiang Y. Experimental study of an improved air source heat pump. Energ Conver Manage 2004;45:2393–403. [129] Wang W, Xiao J, Feng YC, Guo QC, Wang LC. Characteristics of an air source heat pump with novel photoelectric sensors during periodic frost-defrost cycles. Appl Therm Eng 2013;50:177–86. [130] Cho H, Kim Y, Jang I. Performance of a showcase refrigeration system with multievaporator during on-off cycling and hot-gas bypass defrost. Energy 2005;30:1915–30. [131] Huang D, Yuan XL, Zhang XQ. Effects of fan-starting methods on the reverse-cycle defrost performance of an air-to-water heat pump. Int J Refrig 2004;27:869–75. [132] Huang D, He ZL, Yuan XL. Dynamic characteristics of an air-to-water heat pump under frosting/defrosting conditions. Appl Therm Eng 2007;27:1996–2002. [133] Dong JK, Jiang YQ, Deng SM, Yao Y, Qu ML. Improving reverse cycle defrosting performance of air source heat pumps using thermal storage-based refrigerant sub-cooling energy. Building Serv Eng Res Technol 2012;33:223–36.

Uneven defrosting on the outdoor coil in an ASHP 3.1

3

Introduction

Among several defrosting methods currently used for ASHP units such as EHD, HESD, and HGBD, RCD is the most widely used. During a standard and complete RCD process for an ASHP unit, both frost melting and coil surface drying are included. Because a drying coil surface plays an important role in the RCD process, the effects of the downward-flowing melted frost due to gravity on the surface of the outdoor coil in an ASHP unit on its operating performance should be considered. For a vertically installed multicircuit outdoor coil of an ASHP unit, the defrosting process on a different circuit’s surface is always uneven, as the tube surface temperatures at the exits of each circuit reach the preset temperature at different times. It was reported that when the top circuit ends defrosting, the bottom ones may still be covered with frost. When the tube surface at the exit of the lowest circuit reached 24°C, the temperature of the top circuit was nearly 42°C [1, 2]. One important reason for this is the existence of melted frost flowing from top to bottom along the outdoor coil surface due to gravity. A few studies on the effects of the downward flow of the melted frost over a multicircuit outdoor coil may be identified in the open literature. However, in a previous related study [3], it was suggested that the downward flow of the melted frost over a multicircuit outdoor coil of an ASHP unit during RCD could affect the defrosting performance by using more energy for defrosting and prolonging the defrosting process. This is because the downward flow of melted frost helps form or reinforce a water layer between the frost and the coil surface, which introduces a thermal resistance [4] and thus reduces the heat transfer between them. However, no detailed quantitative analysis of these negative effects was carried out and reported.

3.2

An experimental study on an outdoor coil with two refrigerant circuits

This section reports on an experimental work on the effects of downward flowing of the melted frost over the surface of a five-circuit outdoor coil in an experimental ASHP unit during RCD. The detailed description of the experimental ASHP unit is first presented. This is followed by presenting various experimental conditions and experimental results. Third, a quantitatively analysis on the impacts of the melted frost flowing downward due to gravity on the heat and mass transfer process is presented. Finally, a brief conclusion is given.

Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00003-5 © 2019 Elsevier Ltd. All rights reserved.

48

Defrosting for Air Source Heat Pump

3.2.1 Experimental setup 3.2.1.1 The ASHP unit An experimental ASHP unit was established specifically for undertaking the experimental work presented in this work. The unit was built based on a commercially available 6.5 kW heating-capacity variable speed (VS) ASHP unit. The experimental unit was placed in an existing environmental testing chamber when there were both a simulated outdoor frosting space and a simulated heated indoor space. The sizes of both spaces were 3.8 m (L)
3.8 m (W)
2.8 m (H). Fig. 3.1 shows the schematic diagrams of the ASHP unit installed in the environmental testing chamber. It was a split-type unit made of a swing-type compressor, an accumulator, an electronic expansion valve (EEV), a four-way valve, an outdoor coil, and an indoor coil. The outdoor coil was specially designed for this study, as shown in Fig. 3.2. There were five individual and parallel refrigerant circuits and the airside surface area for each circuit was equal. The outdoor coil was vertically installed and in each circuit a solenoid valve (SV) and a manual stop valve (MV) were used. Five water-collecting trays made of polyvinyl chloride (PVC), which can be placed under each circuit of the outdoor coil when needed, were added to the outdoor unit. In this way, the flowing of melted frost would be restricted within a circuit and could not flow to the surfaces of other circuits underneath. Furthermore, five water-collecting cylinders were connected to these trays so that the melted frost from each circuit during defrosting could be collected and weighed. The specifications for the five-parallel refrigerant circuit outdoor coil are summarized in Table 3.1.

Fig. 3.1 Schematic diagrams of the experimental ASHP unit.

Uneven defrosting on the outdoor coil in an ASHP

49

Fig. 3.2 Details of the five-circuit outdoor coil and the locations of valves. Table 3.1 Specifications for the outdoor coil of the experimental ASHP unit Item

Parameters

Values

Units

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Height Width Fin height Fin width Fin pitch Fin thickness External diameter of tube Tube thickness Tube spacing Number of circuits Number of tube rows Circuit pitch (mm) Material of tube Material of fin Fin type Water-collecting tray material

850 590 152.4 44 2.1 0.115 10 0.5 20 5 2 22 Copper Aluminum Plate PVC

mm mm mm mm mm mm mm mm mm / / / / / / /

50

Defrosting for Air Source Heat Pump

Fig. 3.3 The schematic diagram of the environmental chamber.

As shown in Fig. 3.3, there was a separate direct expansion air-conditioning (DX A/C) system in the environmental chamber, and sensible and latent loadgenerating units (LGUs) that were used to simulate the thermal load so that suitable experimental conditions in both the indoor and outdoor spaces could be maintained. During normal heating (frosting) operation, a frosting environment in the outdoor space was maintained by running the experimental ASHP unit and LGUs together while an indoor heated environment by the experimental ASHP unit and the existing A/C system. Fig. 3.4 shows the details of the airside of the outdoor coil in the experimental ASHP unit placed in the frosting outdoor space. On the windward side, air dry bulb temperatures were measured at 10 points using thermocouples (Type K, of 0.1°C accuracy) and air wet bulb temperatures at 5 points using temperature sensors (PT100, class A). The average values from these measurements were taken as the inlet air dry bulb temperature and wet bulb temperature in the follow-up calculation. On the other hand, the air parameters of temperature and humidity exiting the outdoor coil were measured using a hygrosensor (0.2°C and 1.0% RH accuracy) located inside an air duct 900 mm away from the outlet of the outdoor coil. To ensure the best possible accuracy of measurement, the air wet bulb temperature sensors positioned on the windward side of the outdoor coil were calibrated using the hygrosensor. Moreover, the outdoor coil air flow rate was measured by using a flow hood (of 3% accuracy) having a 16-point velocity grid located at the center of a 400
400 mm air duct 600 mm long, as shown in Fig. 3.4. The temperatures of the tube/coil and fin surfaces of the outdoor coil were measured using K-type thermocouples. Ten were for measuring the refrigerant tube surface temperatures at both the entrance and exit of the five refrigerant circuits. Five were affixed on the fin surface at the center of each circuit. Furthermore, five more thermocouples were placed inside the cylinders for measuring the temperature of

Uneven defrosting on the outdoor coil in an ASHP

51

Fig. 3.4 Airside details of the outdoor coil in the experimental ASHP unit.

the melted frost collected. In addition, pressure transmitters having an accuracy of 0.3% of the full scale reading were used to measure refrigerant pressures. The mass flow rate of refrigerant was measured by a variable-area flow meter with an accuracy of 1.6% of the full scale reading. All sensors/measuring devices can output a direct current signal of 4–20 mA or 1–5 V. In addition, during defrosting, photos were taken at an interval of 10 s to visually record the conditions of frost melting on the surface of the outdoor coil.

3.2.1.2 Experimental conditions and procedures Prior to the defrosting operation, the experimental ASHP unit was in operation at frosting (heating) mode for 1 h, at the frosting ambient temperature of 0.5  0.2°C (dry-bulb temperature) and 90%  3% RH. This was achieved by the joint use of both the LGUs and the experimental ASHP unit placed in the outdoor frosting space. Before beginning defrosting, the compressor was turned off first. Then, after 1 min, the four-way valve was changed to defrost mode. Four seconds later, the compressor was switched on again, and thus a defrost operation was commenced. The defrost operation was ended manually when the temperature of the tube surface at the exit of the lowest refrigerant circuit of the outdoor coil arrived at 24°C [5–10]. The outdoor air fan was switched off during defrosting, but the indoor air fan remained in operation at a lower speed. During frosting, the air temperature in the heated indoor space was kept at 20°C. This was achieved by the joint use of both the existing A/C system and the experimental ASHP unit. Table 3.2 shows the experimental conditions.

3.2.2 Experimental cases A series of experimental works using the experimental ASHP unit have been carried out to examine the effects of the downward flow of the melted frost over the outdoor coil surface due to gravity. Various combinations of refrigerant circuits, from two

52

Defrosting for Air Source Heat Pump

Table 3.2 Experimental conditions Item

Parameters

Value

Units

1 2 3 4 5 6 7

Air temperature in indoor heated space Air temperature in outdoor frosting space Air relative humidity in outdoor frosting space Outdoor coil face velocity Face velocity of indoor coil at defrosting mode Face velocity of indoor coil at frosting (heating) mode Frosting (heating) operation duration

20 0.5  0.2 90  3 1.2a 2.31 3.68 60

°C °C % m/s m/s m/s min

a

At the start of a frosting (heating) operation. During frosting (heating), the face velocity decreased due to frost growth.

Table 3.3 Two experimental cases Number of water-collecting trays installed

Positions of watercollecting trays

Case 1

1

Under Circuit 2

Case 2

2

Under Circuit 1 and Circuit 2

Results shown in Figs. 3.5, 3.6, 3.7, 3.10 Figs. 3.5, 3.8–3.10

circuits to five circuits, were used in the experimental work. The experimental results based on two circuits (the first and second circuit from the top) are reported in this section and the results of the other circuit combinations will be separately reported elsewhere. In order to obtain meaningful experimental results, it was necessary to ensure that the frost that accumulated on the surface of the two circuits was even. This was done by adjusting the manual stop valves so that the amount of frost accumulation on the two circuits was close to each other (difference < 10%). Experiments were then carried out at two experimental cases, as detailed in Table 3.3, so that the effects of the downward flow of the melted frost over the surface of a two-circuit outdoor coil were comparatively and quantitatively studied.

3.2.3 Results and analysis Fig. 3.5 presents eight photographs showing the defrost process on the airside of the top two circuits of the outdoor coil up to 120 s into the defrost operation with photos (a1) to (a4) for Case 1 and photos (b1) to (b4) for Case 2. It then took a further 15 s in Case 2 and 30 s in Case 1, respectively, for the tube surface temperature of Circuit 2 to arrive at 24°C, when the defrost was ended. As observed from Fig. 3.5A1 and B1, the frost covers the whole surface of the outdoor coil at the start of defrosting in the two cases, and from Fig. 3.5A2 and B2, the frost melting started at 60 s into defrosting, and no melted frost flowing downward

Uneven defrosting on the outdoor coil in an ASHP

53

Fig. 3.5 Airside conditions on the outdoor coil surface during defrosting. (A1) 0 s; (A2) 60 s; (A3) 100 s; (A4) 120 s; (B1) 0 s; (B2) 60 s; (B3) 100 s; (B4) 120 s.

was observed at this point in time. Furthermore, as seen from Fig. 3.5A3, in Case 1, the melting of frost on the Circuit 1 surface was slightly faster than that on the surface of Circuit 2, as melted frost flowed downward from Circuit 1 to Circuit 2. However, as seen from Fig. 3.5B3 in Case 2, the state of frost melting on the two circuits was basically the same. The melted frost from Circuit 1 was taken away by the collecting tray before it reached Circuit 2. Finally, the airside conditions of the outdoor coil at 120 s into the defrost operation are shown in Fig. 3.5A4 and B4, respectively. It can be seen from Fig. 3.5A4 that in Case 1, when there was no frost on the airside of Circuit 1, still there was frost on the surface of the lower part of Circuit 2 waiting to be melted. However, as seen from Fig. 3.5B4, with two collecting trays installed in Case 2, the frost on both circuits all disappeared. Therefore, defrosting was quicker with a tray installed between the two circuits. From the eight photos, the negative effects of the downward flow of the melted frost over a two-circuit outdoor coil during reverse cycle defrost can be visually observed. Figs. 3.6–3.10 show the measured operating performances for the experimental ASHP unit during defrost. In Figs. 3.6–3.9, for the time (horizontal) axis, 50 s is the chosen starting time in order to clearly show the temperature rise during defrosting. Figs. 3.6 and 3.8 present the measured temperatures of the tube surface at the exits of the two refrigerant circuits during defrost. Figs. 3.7 and 3.9 show the measured temperatures of the fin at the center point of the two circuits. In fact, the variation trends for these temperatures were very similar to those reported by O’Neal et al. [11]. The measured temperatures of the melted frost in the water-collecting cylinders in the two cases from 110 s into the defrosting operation to the end of defrosting at 155 s are further shown in Fig. 3.10. It is seen from Fig. 3.6 that the temperatures remained around 0°C during the first 60 s, and started to rise steadily thereafter. As already shown in Fig. 3.5A2 and B2,

54

Defrosting for Air Source Heat Pump

Fig. 3.6 Measured tube surface temperature at the exits of refrigerant circuits in Case 1.

Fig. 3.7 Measured fin surface temperatures at the center of refrigerant circuits in Case 1.

frost melting began at 60 s into the defrosting operation, and afterward the tube surface temperatures increased steadily from 0°C to 24°C. The tube temperatures of the two circuits reached 24°C at 138 and 152 s, respectively. However, in Fig. 3.7 with two collecting trays installed in Case 2, the temperatures of the tube surface at the exits of the two circuits reached 24°C almost at the same time, at 136 s. The time difference of 14 s for the two circuits to reach 24°C in Case 1 clearly showed that it would take

Uneven defrosting on the outdoor coil in an ASHP

55

Fig. 3.8 Measured tube surface temperature at the exits of refrigerant circuits in Case 2.

Fig. 3.9 Measured fin surface temperatures at the center of refrigerant circuits in Case 2.

longer for the lower circuit to reach 24°C, so that a defrosting operating was prolonged and more energy was consumed. Fig. 3.6 shows the variations of the measured surface temperatures of the fin at the center point of each circuit. Unlike tube surface temperatures, fin surface temperatures remained at 0°C during the first 90 s into defrosting. The rise in fin temperature was later than that in tube surface temperature for 30 s. This was because the tube was in

56

Defrosting for Air Source Heat Pump

Fig. 3.10 Measured temperature of the melted frost in the water-collecting cylinders.

direct contact with the hot refrigerant, but the fin was indirectly in contact with the refrigerant via the tube. As a result, the fin temperature increased later than the tube temperature. Then, it took 62 and 65 s for the fin temperatures to reach 24°C in the two circuits, respectively. Again, it required a longer time for the fin temperature in Circuit 2 to arrive at 24°C, which could be also attributed to the downward flow of the melted frost from Circuit 1. In Fig. 3.9, the variations of the measured temperatures of the fin at the center of the two refrigerant circuits during defrost in Case 2 are shown. Unlike those shown in Fig. 3.6, the fin temperatures at both circuits reached 24°C almost at the same time, at 148 s into defrosting, as the downward flowing of melted frost from Circuit 1 to Circuit 2 was stopped by the presence of the collecting tray. Fig. 3.10 presents the variations of the measured melted frost temperature in the water-collecting cylinders in two cases. The temperature of melted frost for Case 1 rose from 0.68°C at 110 s to 1.87°C at 155 s. However, for Case 2, the temperatures of the melted frost in the two water-collecting cylinders both rose slowly from about 0°C to less than 0.4°C at 155 s into the defrosting operation. The temperature of melted frost for Case 1 was always higher than that in the water-collecting cylinders for Case 2 because the melted frost flowing path was shortened and the heat transfer between the fins and melted frost decreased in Case 2. This, therefore, further confirmed the negative effects of the downward flow of the melted frost due to gravity during RCD on the defrosting efficiency. Moreover, the total amount of melted frost collected in Case 1 was 620 g, 35 g less than the 655 g collected in Case 2. The net difference of 35 g was considered to have been evaporated, as some of the melted frost flowed down from Circuit 1 to the upper

Uneven defrosting on the outdoor coil in an ASHP

57

part of Circuit 2, which was warmer due to the hot refrigerant flowing from top to bottom within the circuit during defrosting. Furthermore, the downward flowing of melted frost not only delayed the defrosting process, but it also led to energy waste. The energy used for RCD comes from three sources: the input power to the compressor, the input power to the indoor air fan, and the thermal energy from the indoor air. In the experimental study presented in this paper, the total energy used was 580 kJ for Case 1, but 517 kJ for Case 2, or 10.3% less. Therefore, allowing melted frost to freely flow downward due to gravity would lead to more energy use for RCD. There can be two reasons for more energy use: (a) evaporating some of the melted frost as mentioned earlier, or (b) when a defrosting process was prolonged, the surface of the fins in the upper part of Circuit 1 could be dry already while that of the fins in the lower part of Circuit 2 was still wet or covered with frost. Therefore, thermal energy would be used for just heating the ambient cold air, which was highly undesirable. In conclusion, the experimental work reported in this section quantitatively demonstrated the effects of allowing melted frost to downward flow from up to down over the airside surface of an outdoor coil in an ASHP unit during RCD. The time duration for defrosting was prolonged and more energy was consumed during defrosting. Therefore, allowing melted frost to flow downward due to gravity would impact negatively on the operating performance of a reverse cycle defrost operation for ASHP units. However, in this section, only two circuits were used and tested. The influence of downward-flowing melted frost on defrosting performance for a multicircuit outdoor coil should therefore be studied. Additionally, measures to mitigate the negative impacts should be further considered.

3.3

Three-circuit experimental study

As reported in Section 3.2, the downward flowing of melted frost over a vertical twocircuit outdoor coil during RCD could adversely affect the defrosting performance of an ASHP unit by using more energy for defrosting and prolonging the defrosting process. This was because the downward flowing of the melted frost helped form or reinforce a water layer between the frost and coil surface, introducing a thermal resistance, and thus reducing the heat transfer between the two. More residual water could be left on the surface of the bottom circuits and thus more energy would be needed to dry the residual water on the bottom circuits. Moreover, during a defrosting process, not only a great deal of energy for melting frost and vaporizing the retained melted frost off the outdoor coil surface is consumed, but also the occupants’ thermal comfort may be adversely affected because no heating is provided during defrosting [1]. Therefore, shortening a defrosting period should be one of the control purposes for ASHP units. For example, Chinese Standard GB/T 7725-2004 specifies that the defrosting duration for an ASHP unit should not exceed 20% of its total working hours. However, for an outdoor coil having more than two circuits, the melted frost

58

Defrosting for Air Source Heat Pump

effect on defrosting performance should also be further studied, and thus its negative effects could be finally fixed for an ASHP unit. This section reports on an experimental study on the effects of the downward flowing of melted frost due to gravity over a vertical three-circuit outdoor coil surface on the defrosting performance of an experimental ASHP unit during RCD. A detailed description of the experimental ASHP unit is first presented. This is followed by reporting various experimental conditions and experimental results. Finally, based on the experimental results, a quantitative analysis on the impacts of the downward flowing of the melted frost due to gravity on the defrosting performance is reported.

3.3.1 Experimental setup The experimental ASHP unit used in this section is also placed in the environmental chamber, which is the same as the previous section. The difference is the outdoor coil, here a three-circuit outdoor coil, as shown in Fig. 3.11. In fact, we designed a fivecircuit outdoor coil at first, but only two circuits were used in the previous section. But here, we used the top three circuits, and thus the outdoor coil was correspondingly adjusted. That means that, compared with the previous two-circuit outdoor coil, the three-circuit one has a total heat transfer area that is 1.5 times larger. The specifications of the tailor-made three-circuit outdoor coil are summarized in Table 3.4. Table 3.5 summarizes the measuring accuracy for various sensors/instruments used in the experimental ASHP unit, and the calculated relative standard errors for the three calculated parameters of total energy supply for defrosting, total energy consumption during defrosting, and defrosting efficiency, respectively.

Fig. 3.11 Location of valves on the three-circuit outdoor coil.

Uneven defrosting on the outdoor coil in an ASHP

59

Table 3.4 Specifications of the tailor-made three-circuit outdoor coil Item

Parameter

Value

Unit

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Height of the outdoor coil Width of the outdoor coil Thickness of the outdoor coil Fin height Fin width Fin thickness Fin pitch Fin type Tube external diameter Tube thickness Tube spacing Circuit pitch Number of tube rows Number of circuits Number of water-collecting trays Number of water-collecting cylinders Number of wind boards Material of tube Material of fin Material of wind board Material of water-collecting tray Material of water-collecting cylinder Volume of Cylinder A, B, and C Volume of Cylinder D

500 590 44 152.4 44 0.115 2.1 Plate 10 0.5 20 22 2 3 3 3 2 Copper Aluminum Wood PVC PVC 500 2000

mm mm mm mm mm mm mm – mm mm mm mm – – – – – – – – mL mL

Table 3.5 Measurement/calculation errors of system parameters Item

Parameter

1

6

Air dry-bulb temperature at upstream of the outdoor coil Air wet-bulb temperature at upstream of the outdoor coil Air temperature at downstream of the outdoor coil Air relative humidity at downstream of the outdoor coil Air flow rate passing through the outdoor coil Temperatures of tube/coil and fin surfaces

7

Refrigerant pressure

2 3 4 5

Measurement/ calculation error

Unit

1 (K-type thermocouple) 0.1 (PT100, class A)

°C °C

0.2 (hygrosensor)

°C

1.0% (hygrosensor)

–

3% (flow hood)

–

1 (K-type thermocouple) 0.3% (pressure transmitters)

°C – Continued

60

Defrosting for Air Source Heat Pump

Table 3.5 Continued Measurement/ calculation error

Item

Parameter

8

Refrigerant volumetric flow rate

9 10 11

Mass of melted frost collected Mass of residual water collected Temperature of melted frost collected

12 13

Total energy supply for defrosting Total energy consumption during defrosting Defrosting efficiency

14

Unit

1.6% (variable area flow meter) 0.1 (weighing scale) 0.1 (weighing scale) 1 (K-type thermocouple) 1% (calculated) 1% (calculated)

–

kJ kJ

0.15% (calculated)

–

g g °C

3.3.2 Experimental cases To study the effects of the downward flowing of the melted frost over the three-circuit outdoor coil surface due to gravity on defrosting performance, a series of experimental works using the experimental ASHP unit have been carried out. In order to obtain meaningful experimental results, it was necessary to ensure that the frost that accumulated on the surface of the three circuits was even first. Therefore, a series of trial-and-error manual adjustments of the degree of opening of the stop valves, to adjust the refrigerant flow into each circuit so that a set of fixed degree of valve openings was obtained, such that the amount of frost accumulation on the three circuits was close to each other (difference < 10%). Experimental work was then carried out at three experimental cases, as detailed in Table 3.6, so that the effects of the downward flowing of melted frost on the defrosting performance can be comparatively and quantitatively analyzed.

3.3.3 Results and analysis Fig. 3.12 presents 12 photographs showing the defrosting process on the airside of the three circuits of the experimental outdoor coil, with photos (A1)–(A4) for Case 1, photos (B1)–(B4) for Case 2, and photos (C1)–(C4) for Case 3, respectively. Table 3.6 Water trays at frosting and defrosting stages in three experimental cases Trays

Case 1

Case 2

Case 3

Number of water-collecting trays installed at frosting stage Number of water-collecting trays installed at defrosting stage Positions of water-collecting trays at defrosting stage

0

0

0

1 (Tray C)

2 (Trays A, C)

Under Circuit 3

Under Circuits 1 and 3

3 (Trays A, B, C) Under each circuit

Uneven defrosting on the outdoor coil in an ASHP

61

Fig. 3.12 Airside surface conditions of the outdoor coil during defrosting in three cases. (A1) 0 s; (A2) 100 s; (A3) 130 s; (A4) 150 s; (B1) 0 s; (B2) 100 s; (B3) 130 s; (B4) 150 s; (C1) 0 s; (C2) 100 s; (C3) 130 s; (C4) 150 s.

As observed from Fig. 3.12A1, B1, and C1, the surface conditions at the start of defrosting for the three cases were virtually the same. From Fig. 3.12A2, B2, and C2, it can be seen that frost melting started 100 s into defrosting, and no melted frost flowing downward may be observed at this stage. Furthermore, as seen from Fig. 3.12A3, at 130 s into the defrosting operation in Case 1, the frost melting on the surface of Circuit 1 was quicker than that on the surface of Circuit 2 while the melted frost flowed downward from Circuit 1 to Circuit 2. On the other hand, the frost melting on the surface of Circuit 3 was later than that on the surface of Circuit 2 because the melting frost flowing downward to Circuit 3 was more than that to Circuit 2. However, as seen from Fig. 3.12B3 for Case 2 and C3 for Case 3, the state

62

Defrosting for Air Source Heat Pump

of frost melting on the two up-circuits was basically the same. The melted frost from Circuit 1 was taken away by the collecting tray before it reached Circuit 2. Finally, the airside conditions of the outdoor coil at 150 s into the defrosting operation at the three experimental cases are shown in Fig. 3.12A4, B4, and C4, respectively. It can be seen from Fig. 3.12A4 that with only one tray, while the airside of Circuit 1 was already free of frost, there was still frost on the surface of the lower part of Circuit 2 and Circuit 3 waiting to be melted. Also as seen from Fig. 3.12B4 with two collecting trays installed under Circuit 1 and Circuit 3, there was nearly no frost left on the two up-circuits while clearly there was some frost left on Circuit 3. However, as seen from Fig. 3.12C4, with three collecting trays installed, the frost on the three circuits all disappeared. Therefore, defrosting was quicker and more even with trays installed. From the above 12 photos, the negative effects of the downward flowing of melted frost over the experimental three-circuit vertical outdoor coil due to gravity on the defrosting performance during RCD can be visually observed. The measured operating performances of the experimental ASHP unit during defrosting, corresponding to the three experimental cases, are presented in Figs. 3.13–3.18. In all these figures, for their time (horizontal) axis, 80 s is the chosen starting time in order to clearly show the temperature rise during defrosting. Figs. 3.13–3.15 present the measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting. Figs. 3.16–3.18 show the measured fin temperatures at the center point of the three circuits. It is noted that the variation trends of these temperatures are similar to those reported by O’Neal et al. [11]. It can be seen from Fig. 3.13 that in Case 1, the temperatures remained around 0°C during the first 100 s, and started to rise steadily thereafter. As already shown in Fig. 3.12A2, B2, and C2, the frost on the surface of Circuit 3 began melting at

Fig. 3.13 Measured tube surface temperatures at the exits of refrigerant circuits in Case 1.

Uneven defrosting on the outdoor coil in an ASHP

63

Fig. 3.14 Measured tube surface temperatures at the exits of refrigerant circuits in Case 2.

Fig. 3.15 Measured tube surface temperatures at the exits of refrigerant circuits in Case 3.

100 s into the defrosting operation, and thereafter, the tube surface temperatures increased steadily from 0°C to 24°C. The tube surface temperatures of the three circuits reached 24°C at 172, 182, and 186, respectively. In Case 2, as shown in Fig. 3.14, the time for the surface temperature at the exit of the two up-circuits to reach 24°C was nearly the same at 174 s, and that for Circuit 3 was a little longer at 180 s. However, as shown in Fig. 3.15, in Case 3, the surface temperatures at the exits of the three circuits

64

Defrosting for Air Source Heat Pump

Fig. 3.16 Measured fin temperatures at the center of the three refrigerant circuits during defrosting in Case 1.

Fig. 3.17 Measured fin temperatures at the center of refrigerant circuits in Case 2.

reached 24°C almost at the same time, at 167, 166, and 168 s, respectively. The time differences of 6 and 18 s in Case 2 and Case 3 compared to that in Case 1 clearly showed that it would take longer for the bottom circuit to reach 24°C when fewer trays were installed, so that a defrosting operating was prolonged and more energy consumed.

Uneven defrosting on the outdoor coil in an ASHP

65

Fig. 3.18 Measured fin temperatures at the center of refrigerant circuits in Case 3.

Figs. 3.16–3.18 show the variations of the measured fin surface temperatures at the center point of each circuit in the three experimental cases. Unlike the tube surface temperatures, the fin surface temperatures remained at 0°C at the first 110 s into defrosting. The rise in fin temperature was later than that in the tube surface temperature. This was because the tube was in direct contact with the hot refrigerant, but the fin was indirectly in contact with the refrigerant via the tube. As seen in Fig. 3.16, in Case 1, it took 75, 80, and 85 s for the fin temperatures to reach 24°C in the three circuits, respectively. Again, it took a longer time for the fin temperature in the bottom circuit to reach 24°C, which could be also attributed to the downward flowing of melted frost from the up-circuits. In Fig. 3.17, the variations of measured fin temperatures in Case 2 are shown. Unlike those shown in Fig. 3.16, the fin temperatures at the two up-circuits reached 24°C almost at the same time at 185 s, as the downward flowing of melted frost from Circuit 1 to Circuit 2 was stopped by the presence of the collecting trays, but the fin temperature of Circuit 3 reached 24°C at about 7 s later, at 192 s. When the melted frost from each circuit was taken away by the collecting trays before downward flowing to the lower circuits in Case 3, as shown in Fig. 3.18, the time durations for the fin temperatures at the three circuits to reach 24°C were very close to each other, at 177, 179, and 175 s, respectively. Furthermore, as more water-collecting trays were installed, the duration for the fin temperature in Circuit 3 to reach 24°C gradually became smaller. Therefore, the negative effects of the downward flowing of melted frost over a vertical multicircuit outdoor coil on the defrosting performance of an ASHP unit during RCD are further shown. Fig. 3.19 presents the variations of measured melted frost temperatures in the water-collecting cylinders in the three cases. The temperature of the melted frost in

66

Defrosting for Air Source Heat Pump

Fig. 3.19 Temperatures of melted frost collected in the water-collecting cylinders.

Cylinder C, in Case 1, rose from 0.3°C at 130 s to about 1.6°C at 195 s. In Case 2, the temperature of the melted frost in Cylinder A, collected from Circuit 1, was around 0°C at 130 s to 150 s into the defrosting operation and rose to about 0.15°C at 195 s. The temperature of melted frost from Circuit 3 in Cylinder C increased from 0.4°C at 130 s to 1.2°C at 195 s. However, in Case 3, the temperatures of the melted frost in the three cylinders, that is, A, B, and C, all rose slowly from about 0°C to less than 0.2°C at 195 s. The temperature of the melted frost in Case 1 was always higher than that in Case 3 because the melted frost flowing path was shortened so that the heat transfer between the fins and the melted frost was decreased. On the other hand, in Case 2, the flowing path of the melted frost collected in Cylinder A was half that in Cylinder C, so that its temperature was always lower. These, therefore, further confirmed the negative effects of the downward flowing of melted frost due to gravity during RCD on defrosting performance. Moreover, the total amount of melted frost collected in Case 1 was 921 g, 22 g less than that collected in Case 2 and 48 g less than that collected in Case 3. The net differences of 22 g and 48 g were considered to have been evaporated. This was because the melted frost flowing down was heated by the upper part of the down circuits, which was warmer as the hot refrigerant flowed from top to bottom within the circuit during defrosting. This explained why less melted frost was collected with fewer collecting trays. Furthermore, downward flowing of the melted frost not only delayed the defrosting process, but also led to energy waste. The energy used for RCD comes from three sources: the power input to the compressor, the power input to the indoor air fan, and the thermal energy from the indoor air. The total energy used for defrosting was 727 kJ in Case 1, but 652 kJ in Case 3, or 10.4% less. In Case 2, the total energy consumed was 683 kJ, or 6.05% less than that in Case 1.

Uneven defrosting on the outdoor coil in an ASHP

67

Defrosting efficiency can be used to evaluate the performance of a defrosting operation. It is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation, as follows: ηd ¼

E m + Ev Qcom + Qfan + Qair

(3.1)

where Em and Ev are the total heat used for melting frost and vaporizing the retained water, respectively, and they are evaluated by: Em ¼ Mf Lsf

(3.2)

Ev ¼ Mv Lv

(3.3)

where Mf and Mv are the total mass of the frost formed on the outdoor coil and the mass of vaporized melted frost, respectively, and Lsf and Lv the latent heat of frost melting and latent heat of evaporation of water, respectively. Also in Eq. (3.1), Qcom, Qfan, and Qair are the energy consumptions by the compressor and supply fan, and the thermal energy from the indoor air during defrosting, respectively. In this section, the defrosting efficiencies calculated for the three cases were 43.5%, 50.6%, and 56.7%, respectively. Similar to the previous section, it was also demonstrated that allowing melted frost to freely flow down due to gravity would lead to more energy consumption during defrosting. There can be two reasons for more energy use: (a) evaporating some of the melted frost as mentioned earlier; (b) when a defrosting process was prolonged, the fin surface in an up-circuit could be dry already while that in a down-circuit was still wet, if not still covered with frost. Therefore, thermal energy would be used for just heating the ambient cold air, which was highly undesirable. Furthermore, the study results also suggested that the use of water-collecting trays was effective in mitigating the negative effects.

3.3.4 Comparison analysis The experimental results by using two-circuit and three-circuit outdoor coils are comparatively analyzed, with their differences summarized in Table 3.7. As seen, after the outdoor coil was changed from a two circuit to a three circuit, the defrosting duration shorted after the trays used was increased from 16 to 18 s, with the rate decreased from 15.8% to 7.5%. From the view of indoor thermal comfort, the three-circuit outdoor coil improvement is more obvious. Although the differences of the melted frost collected in two series are similar, at 5.3% for two circuit and 5.0% for three circuit, the energy consumed savings are nearly the same, at 10.3% and 10.4%, respectively. Finally, after the defrosting efficiency was compared, the improved values for them are 10.5% for a two-circuit outdoor coil and 13.2% for a three-circuit outdoor coil, respectively. The energy performance also demonstrated that the elimination of the melted frost effects is more obvious after the working circuit number increased from two to three.

68

Defrosting for Air Source Heat Pump

Table 3.7 Differences between two-circuit and three-circuit outdoor coil experiments Item

Parameter

1 2 3 4

Total case number Defrosting duration without trays Defrosting duration with trays Reduction in defrosting duration after using trays Total melted frost collected without trays Total melted frost collected with trays Melted frost collected difference Total energy consumed without trays Total energy consumed with trays Total energy saved after trays used Defrosting efficiency with trays Defrosting efficiency without trays Defrosting efficiency improved after trays used

5 6 7 8 9 10 11 12 13

3.4

Two-circuit outdoor coil

Three-circuit outdoor coil

2 152 s 136 s 16 s (15.8%)

3 186 s 168 s 18 s (7.5%)

620 g

921 g

655 g

969 g

35 g (5.3%) 576 kJ

48 g (5.0%) 727 kJ

517 kJ 59 kJ (10.3%) 52.5% 63.0% 10.5%

652 kJ 75 kJ (10.4%) 43.5% 56.7% 13.2%

Concluding remarks

The experimental results and corresponding quantitative analysis reported in this chapter demonstrated the negative effects of allowing melted frost to flow downward due to gravity over the airside surface of experimental vertical two-circuit and threecircuit outdoor coils in an ASHP unit on defrosting performance during RCD: a longer defrosting duration and more energy consumption. Furthermore, the study results also suggested that the use of water-collecting trays was effective in mitigating the negative effects. For two-circuit and three-circuit outdoor coils, the increases in defrosting efficiency values are different, at 10.5% and 13.2%, respectively. That means the energy performance could be improved more obviously for a three-circuit outdoor coil after using the water-collecting trays. Additionally, for a three-circuit outdoor coil, a greater reduction in defrosting duration could be achieved. This further demonstrates that a better thermal comfort performance may be achieved when using a three-circuit outdoor coil. However, the circuit number in an outdoor coil is always limited by its physical dimensions. Also, when a larger number of circuits is used, or when the heat exchanger area of an outdoor coil is fixed and the circuit number increased, the effects of melted frost flowing downward should be further studied. On the other hand, further mathematical modeling studies on the heat and mass transfer mechanisms when the melted frost from the upper circuits flows over the frosted surface of the lower circuits in a vertical multicircuit outdoor coil have also been carried out, and the study results will be reported in the next chapter separately.

Uneven defrosting on the outdoor coil in an ASHP

69

References [1] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse cycle defrosting method for air source heat pumps. Build Environ 2010;45(11):2354–61. [2] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-part I: experiments. Appl Energy 2012;91:122–9. [3] Wang ZY, Wang XM, Dong ZM. Defrost improvement by heat pump refrigerant charge compensating. Appl Energy 2008;85(11):1050–9. [4] Payne V, O’Neal DL. Defrost cycle performance for an air-source heat pump with a scroll and a reciprocating compressor. Int J Refrig 1995;18(2):107–12. [5] Ashiqur Rahman M, Jacobi AM. Drainage of frost melt water from vertical brass surfaces with parallel microgrooves. Int J Heat Mass Transf 2012;55(5–6):1596–605. [6] Wang W, Xiao J, Feng YC, Guo QC, Wang LC. Characteristics of an air source heat pump with novel photoelectric sensors during periodic frost-defrost cycles. Appl Therm Eng 2013;50(1):177–86. [7] Cho H, Kim Y, Jang I. Performance of a showcase refrigeration system with multievaporator during on-off cycling and hot-gas bypass defrost. Energy 2005;30(10): 1915–30. [8] Huang D, Li QX, Yuan XL. Comparison between hot-gas bypass defrosting and reversecycle defrosting methods on an air-to-water heat pump. Appl Energy 2009;86(9): 1697–703. [9] Hu WJ, Jiang YQ, Qu ML, Yao Y, Deng SM. An experimental study on the operating performance of a novel reverse-cycle hot gas defrosting method for air source heat pumps. Appl Therm Eng 2011;31(2–3):363–9. [10] Wang DY, Tao TF, Xu GH, Luo AL, Kang SY. Experimental study on frosting suppression for a finned-tube evaporator using ultrasonic vibration. Exp Therm Fluid Sci 2012;36:1–11. [11] O’Neal DL, Peterson K, Anand NK, Schliesing JS. Refrigeration system dynamics during the reverse cycle defrost. ASHRAE Trans 1989;95(2):689–98.

Modeling study on uneven defrosting 4.1

4

Introduction

When the surface temperature of the outdoor coil in an ASHP unit is below both the air dewpoint and the freezing point of water, frost can form and accumulate over the surface of the outdoor coil, which is usually of a multicircuit structure in order to minimize its refrigerant pressure loss and enhance the heat transfer between the refrigerant and outdoor air. However, frosting adversely affects the operational performance and hence the energy efficiency of an ASHP unit; therefore, periodic defrosting is necessary. Currently, the most widely used standard defrosting method for ASHP units is RCD. During RCD, a space heating ASHP unit actually cools a space, degrading the indoor thermal comfort while consuming electrical energy for melting frost. Therefore, a defrosting period should be controlled to be as short as possible. In order to better improve the defrosting performance for an ASHP unit, a number of previous experimental studies were carried out to examine various ways for better defrosting performances. These included optimizing the structure of an outdoor coil [1,2], fin space adjustment [3], fin surface treatment [4], heating up the liquid refrigerant in an accumulator [5], and providing heat for defrosting using PCM [6,7]. On the other hand, uneven defrosting was reported in limited previous experimental studies. O’Neal et al. [8] and Qu et al. [9] both investigated the transient defrosting performances of ASHP units, each with a vertically installed four-parallel refrigerant circuit outdoor coil. It was reported that when a defrosting process was terminated, the surface temperature at the exit of the lowest circuit was much lower than that at the exit of the highest circuit. In the study by Wang et al. [10], it was shown that for a seven-circuit outdoor coil, at 6 min into defrosting, the surface of the down refrigerant circuits, which accounted for almost 1/4 of the entire surface area, was still covered by frost while that of the up-circuits was already free of frost. An efficient alternative to experimentally studying the defrosting performance in an ASHP unit is via a numerical approach and therefore, the last two decades saw a growing number of modeling studies [11–21] on defrosting performance, although most of them were on defrosting methods [17–23] rather than RCD. Krakow et al. [13,14] developed an RCD model for an outdoor coil. In this model, a frost-melting process on the outdoor coil surface was divided into four stages: preheating, melting, vaporizing, and dry heating. Krakow et al. [15,16] later presented an idealized RCD model for an ASHP unit with a receiver. Two parameters—system performance coefficient and coil efficiency—were defined and used when evaluating defrosting performance. Dopazo et al. [23] developed a detailed transient simulation model for hot-gas bypass defrosting in an air-cooled evaporator. As compared to the model by Krakow et al. [13,14], six stages were used in this model: preheating, tube frost melting, fin Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00004-7 © 2019 Elsevier Ltd. All rights reserved.

72

Defrosting for Air Source Heat Pump

frost melting, air presence, retained water vaporizing, and dry heating. However, although the above-mentioned defrosting models were developed and used in studying defrosting performance, the effects of the downward flowing of melted frost due to gravity along the surface of an outdoor coil on the defrosting performance were neglected, by assuming either no water retention on the coil surface or a stable water layer. In 2012, Qu et al. [22] reported on a modeling analysis where a semiempirical model for the defrosting on the airside of a four-circuit outdoor coil in an ASHP unit was developed. In this model, the negative effects of melted frost on defrosting performance were quantitatively studied. It was further predicted that if the melted frost could be drained away locally, the defrosting efficiency for the ASHP unit could be increased by up to 18.3%. Similar results were also reported by Dong et al. [23] in a study on the energy consumption for vaporizing the melted frost and heating the ambient air during RCD for an ASHP unit. To remove the melted frost as soon as possible, it was further suggested that a vertically installed outdoor coil be installed horizontally, as the flow path for the melted frost was shortened [23]. However, in Qu’s model [22], the energy used to heat the metal in an outdoor coil during defrosting was neglected. In fact, this part of energy consumption accounted for as much as 16.5% of the total defrosting energy used during RCD [23]. As a follow-up, a series of experimental studies was carried out on the defrosting performance when the melted frost was drained away locally during RCD in an ASHP unit with a two-circuit and a three-circuit outdoor coil; this was separately reported in Chapter 3. In these studies, a tailor-made experimental multicircuit outdoor coil was used by placing water-collecting trays under each circuit. Comparative experiments with and without the use of water-collecting trays between circuits were carried out and the experimental results suggested that the use of the water-collecting trays helped shorten the defrosting duration by 9.2% and reduce the defrosting energy use by 10.4%. To enable further quantitative analysis on the effects of locally draining away the melted frost on RCD performance in an ASHP unit, a related mathematical modeling study on defrosting performance when using water-collecting trays was considered necessary. Therefore, a modeling study of the defrosting process taking place in the tailor-made three-circuit experimental outdoor coil, at two experimental settings of with and without the use of water-collecting trays between circuits, was carried out and is presented in this chapter. Two semiempirical mathematical models, corresponding to the two settings, were developed. In this chapter, first the detailed development of the two semiempirical models is presented. This is followed by reporting the experimental validations of the two models using the experimental data previously introduced. Then, some parameters that are hard to measure could be numerically predicted. Finally, detailed discussions on the potential uses of the two models developed and the limitations of the modeling work reported are included. Moreover, the models will be used to predict some control methods, and thus the defrosting performance for an ASHP unit with a multicircuit outdoor coil is expected to be fundamentally and efficiently improved.

Modeling study on uneven defrosting

4.2

73

Semiempirical modeling study

4.2.1 Development of two semiempirical mathematical models As shown in Fig. 4.1 and Table 4.1, two semiempirical models were developed based on whether water-collecting trays were installed between circuits. Model 1 was developed for the condition when only one water-collecting tray (Tray C) was installed under the lowest circuit, Circuit 3, and connected to a water-collecting cylinder (Cylinder C), that is, no water trays were installed between circuits. However, Model 2 was for the condition when three water-collecting trays (A, B, and C) were each placed under Circuits 1–3, and connected to three cylinders (A, B, and C), respectively. When developing the two mathematical models, the entire airside of the Outdoor coil (air side)

Refrigerant in mw,j –1Tw,j –1

mw,jTw,j

Circuit 1 mf,j

mw,j+1Tw,j+1

Circuit 2

Circuit 3

Water collecting tray C Refrigerant out

Water collecting cylinder C

C

Model 1 Refrigerant in

mw,j –1Tw,j –1

A

B

mw,jTw,j

mw,j+1Tw,j+1

mf,j

Circuit 1

Water collecting tray A

Circuit 2

Water collecting tray B

Circuit 3

Water collecting tray C Refrigerant out

C

Water collecting cylinder A, B, C Model 2 Thermal-couple

Fig. 4.1 Conceptual models for the airside of a three-circuit outdoor coil.

74

Defrosting for Air Source Heat Pump

Table 4.1 Two semiempirical models Item

Parameter

Model 1

Model 2

1

1 (Tray C)

3 (Trays A, B, C)

2

Number of water-collecting trays installed Positions of water-collecting trays

Under Circuit 3

3

Results shown in

Under each circuit Figs. 4.8–4.11

Figs. 4.6 and 4.7

tailor-made three-circuit outdoor coil surface was divided into three control volumes, respectively, corresponding to the three circuits shown in Fig. 4.1. For each control volume, a lumped parameter modeling approach was applied. In this section, unlike the models developed in previous studies [11,12,22], a defrosting process on the airside of the outdoor coil was chronologically divided into four stages: (1) preheating, (2) frost melting without water flowing away from a circuit, (3) frost melting with water flowing away from a circuit, and (4) water layer vaporizing. Such a way of staging a defrosting process could enable a proper account for the flow of the melted frost into, or away from, a control volume according to the use of water-collecting trays in the experimental three-circuit outdoor coil. Under the assumption of four stages of defrosting, a defrosting process began with preheating. In this stage, all the melted frost could be held on the finned coil surface due to surface tension. When the frost in direct contact with the surface of the tubes and fins was melted, a thin water layer was formed. At the end of the first stage, the thin water layer covered the entire airside surface of the outdoor coil. However, the water layer was not in contact with the ambient air within the entire first stage. In the second stage, as the heat was transferred from the warmer water layer to the frost, the thickness of the frost layer was decreased and that of the water layer increased so that the water layer started to be in contact with the low-temperature ambient air. However, there was no melted frost flowing away from a circuit, as the mass of the melted frost held did not reach its maximum that could be held by surface tension [12]. The third stage began with the start of downward flowing of the melted frost, as the force of gravity was larger than the surface tension. In Model 1, the melted frost flowed downward from an up-control volume into a down-control volume due to gravity during this stage. However, in Model 2, the melted frost did flow out of a control volume but did not flow into a down-control volume as it was taken away by the water-collecting trays installed between the circuits, and was then collected by the respective cylinders. Finally, at the beginning of the fourth stage, or the water layer vaporizing stage, the entire outdoor coil surface was free of frost but covered by the retained water. When the tube surface temperature at the exit of the lowest refrigerant circuit (Circuit 3 in this section) in the outdoor coil reached 24°C [7,9,22], defrosting was terminated.

Modeling study on uneven defrosting

75

4.2.1.1 Assumptions and calculation conditions The two semiempirical models were developed based on the fundamentals of energy and mass conservation and heat and mass transfer within each of the control volumes at each stage of the defrosting process, and also using some of the experimental data obtained in experimental studies previously reported. When establishing the two models, the following was assumed: i. The convective heat transfer between the frost and ambient air in the first two stages was neglected. Therefore, the mass loss of frost due to sublimation during these two stages was neglected. ii. The thermal conductivities of the tubes and fins were much higher than those of the frost and retained water, and hence their heat transfer resistances were neglected. iii. The mass flow rate of refrigerant was evenly distributed into the three refrigerant circuits during defrosting, and the frost was assumed to be uniformly accumulated over the coil surface before starting defrosting. It is not the frost evenly accumulated on the surface of circuit, but the total frost accumulations on different circuits’ surface are assumed to be the same. iv. The movement of the melted frost layer was considered to be a flowing boundary. Because the velocity of the water flow was very small as observed during experiments, the melted frost layer flowed in a laminar regime. v. During the third stage, the retained water in each control volume was in a dynamic equilibrium, i.e., the difference between the mass of the water entering into a control volume and that flowing away from the control volume was equal to the rate of frost melting within the control volume. vi. During defrosting, the melted frost infiltrated into the porous structure of the frost. The contact area between the frost and the melted frost would increase as water flowed downward, suggesting that the flow resistance was increased downward along the surface of the outdoor coil. Therefore, the velocity of the water layer in each control volume was decreased from top to bottom. vii. There would be a thin water film between the frost and the surface of the fin, which is shaped like a wedge. The upper end is thin while the lower is thick. But in this section, it was assumed as a rectangle, so the water film surface contacting with frost is also parallel to the fin surface. And thus, the thermal conductivities between the fin and the water layer and between the water layer and the frost are the same for the whole water film. viii. During defrosting, there was no frost chip or debris flowing into a down circuit or a watercollecting tray. ix. During defrosting, the mass of melted frost left on the water-collecting trays or vaporized from the water-collecting trays and cylinders was neglected. x. In the process of the melted frost falling into a down-circuit or a water-collecting tray, the heat dissipated from the melted frost to the ambient air was negligible because the falling distance was small.

Furthermore, the following experimental data previously obtained were also used in assisting the development of the two semiempirical models: (a) The total mass of the frost was experimentally obtained at 1050 g, thus following Assumption (iii), the mass of frost formed on the surface of each circuit, Mf, j(j ¼ 1  3) , was 1050 / 3 g, or 350 g.

76

Defrosting for Air Source Heat Pump

(b) At 40 s into defrosting, the preheating stage (first stage) was over. (c) At 90 s into defrosting, the frost melting without water flowing away from a circuit stage (second stage) was over. (d) The following experimentally measured refrigerant flow rate, tube surface temperatures at the inlets and outlets of each circuit, temperature of the ambient air surrounding the outdoor coil, and compressor discharge pressure during defrosting were used as the inputs to the models developed.

4.2.1.2 Model development As mentioned, two semiempirical models were developed for the two settings of with and without the use of water-collecting trays between circuits. The use of trays would stop the melted frost from flowing into the circuits (or control volumes) underneath in Stage 3. Therefore, the two models were identical for Stages 1, 2, and 4, with only the modeling work in Stage 3 being different. In this section, for simplicity, the complete development of Model 1 is first presented. For Model 2, only the modeling work in Stage 3 is reported.

Model 1 development (all four stages) The model was developed by applying the energy and mass conservation in each of the three control volumes at each of the fourth defrosting stages. First stage: Preheating As shown in Fig. 4.1A, the energy and mass conservation in Control Volume j yielded:
 qj Δt ¼ Lsf mf , j Δt + cp Δ Mw, j Tw, j + qMe Δt ðj ¼ 1  3Þ

(4.1)

As Δt ! 0, Eq. (4.1) can be written as qj ¼ Lsf mf , j + cp


 d Mw, j Tw, j + qMe ðj ¼ 1  3Þ dt

(4.2)

where Mw, j is the accumulated mass of the retained melted frost in the Control Volume j: ðt Mw, j ¼ mf , j dt ðj ¼ 1  3Þ

(4.3)

0

where t is the ending time of Stage 1 defrosting and Tw, j the temperature of the melted frost on the surface of Circuit j (as shown in Fig. 1). On the other hand, the heat transferred from the water layer to the frost layer for frost melting was:
 hw Tw, j  Ttp A0 ¼ Ls, f mf , j ðj ¼ 1  3Þ

(4.4)

where Ttp is the triple point of water and A0 the equivalent airside surface area of a refrigerant circuit.

Modeling study on uneven defrosting

77

Also in Eq. (4.4), hw, the average coefficient of convective heat transfer caused by water flow inside a control volume, was evaluated by: ðj +ð1ÞH

hw ¼

hw, x dx=H ðj ¼ 1  3Þ

(4.5)

jH

where the convective heat transfer coefficient due to water flow downward was evaluated by [22]: λ 1 1 hw, x ¼ 0:332 Re2x Pr3 x

(4.6)

To evaluate hw, x, it was necessary to evaluate the velocity of the water layer in each control volume, which, based on Assumption (vi), would decrease from top to bottom within a circuit, and could be estimated by [24]: vj ¼

H 0:85j1 ðj ¼ 1  3Þ td,1

(4.7)

where td, 1 is the defrosting duration in Circuit 1, and an experimentally obtained value 168 s was used for td, 1. H is the height of a refrigerant circuit. Furthermore, qMe is the energy used to heat the metal of the outdoor coil and can be evaluated by [25]: qMe ¼ cPMe  ðmCu + mAl Þ

ΔTMe Δt

(4.8)

where ΔTMe is the average temperature difference of the outdoor coil metal and was evaluated by [25]: ΔTMe ¼ Tt  T0

(4.9)

1 T0 ¼ ðTin, 0 + Te,0 Þ 2

(4.10)

1 Tt ¼ ðTin,t + Te, t Þ 2

(4.11)

where T0 and Tt are the average temperatures of the outdoor coil metal at the start and end of a defrosting process, and Tin and Te are the inlet and outlet tube surface temperatures of the outdoor coil, respectively. In Eq. (4.8), cPMe is the average specific heat of metal (copper and aluminum) and can be evaluated by: cPMe ¼

mCu cCu + mAl cAl mCu + mAl

(4.12)

78

Defrosting for Air Source Heat Pump

Meanwhile, the heat transfer in Control Volume j can be expressed as [24], qj ¼

Tr, j TICW, j At Rr

(4.13)

where TICW is the temperature at the interface between the coil surface and the water layer and Tr, j is the average temperature of the refrigerant in Circuit j. The relationship between TICW and Tw, j in this stage was [22]: Tw, j ¼

TICW, j + Ttp ð j ¼ 1  3Þ 2

(4.14)

As the heat transfer resistances of the tubes and fins were neglected (Assumption (ii)), the thermal resistance of the refrigerant, Rr, was evaluated by an empirical experimental correlation for the refrigerant-side mean heat transfer coefficient, hTPM [24]: Rr ¼

1

(4.15)

hTPM

During defrosting, two different heat transfer regions existed in the refrigerant side of the outdoor coil, namely (1) a superheated region and (2) a two-phase region. In the superheated region, the convective heat transfer coefficient of the refrigerant, hr, sh, was evaluated by the standard Dittus-Boelter correlation [25]: hr,sh ¼ 0:023Resh 0:8 Prsh 0:3

λsh L , for  10, Resh  104 , Prsh ¼ 0:7  160 di di

(4.16)

where λsh is the thermal conductivity of the refrigerant in the superheated region, di the inner diameter of the refrigerant tube, and L the tube length of a refrigerant circuit. The convective heat transfer coefficient of the refrigerant in the two-phase region, hr, tp, was evaluated using the liquid refrigerant heat transfer coefficient, hr, L, which was evaluated by [27]: " 0:8

hr,tp ¼ hr, L ð1  xÞ

3:8x0:76 ð1  xÞ0:04 + pre 0:38

hr,L ¼ 0:023ReL 0:8 PrL 0:3

λL di

# (4.17)

(4.18)

where x is the thermodynamic vapor quality, λL thermal conductivity of the liquid refrigerant, and Pre the reduced pressure, determined by [27] Pre ¼

P Pc

(4.19)

Modeling study on uneven defrosting

79

where P is the actual compressor discharge pressure. Pc is the critical pressure for R22, and a value of 4.99 MPa was used. Because the boundary of the two-phase region was moving during defrosting, hTPM, a mean heat transfer coefficient of refrigerant in Eq. (4.15), was evaluated by Shah [24] as:   2:09 hTPM ¼ hr, L 0:55 + 0:38 pre

(4.20)

In Eq. (4.13), At is the total refrigerant tube surface area of each circuit, At ¼

Atube 3

(4.21)

where Atube is the total refrigerant tube surface area of the entire three-circuit outdoor coil. Moreover, in Eqs. (4.2) and (4.13), qj can also be evaluated by [22,26]: qj ¼ mr, j ðhr, in  hr, e Þ ðj ¼ 1  3Þ

(4.22)

where mr, j is the mass flow rate of refrigerant in Control Volume j. Enthalpies of the refrigerant at both the inlet and outlet of each circuit, hr, in and hr, e, were evaluated from the measured inlet and outlet refrigerant temperatures and the measured compressor discharge pressure. Second stage: Frost melting without water flowing away from a circuit As shown in Fig. 4.2B, energy conservation in Control Volume j, taking into account the convective heat transfer between the water layer and ambient air, required:

 d Mw, j Tw, j + hc, w Tw, j  Ta Af a + qMe ðj ¼ 1  3Þ qj ¼ Lsf mf , j + cp dt

(4.23)

where hc, w(Tw, j  Ta)Afa is the heat transferred to the ambient air from the effective airside surface area, Afa, which was covered by only the melted frost in Control Volume j. In the second and third stages, Afa, was evaluated [8,14]: 0ð t

11:5 B 0 mf , j dtC C Af a ¼ A0 B @ M f , j A ð j ¼ 1  3Þ

(4.24)

Because the outdoor fan was turned off during the entire defrosting process, natural convection was the most important form to transfer the heat from the outdoor coil or

Defrosting for Air Source Heat Pump

CV 1

Refrigerant

qMe

Lsf,mf,j Tw,j

Coil metal

Water layer

hc,w(Tw,j –Ta)Aw-a

qT,j

Coil metal

Tw,j Water layer

qT,1

Lsf,mf,j

Frost

hc,w(Tw,1 –Ta)Aw-a

Tw,1 Coil metal

Frost

Water layer

Lsf,mf,1 Frost cpmw,1Tw,1 (to Circuit 2)

Ambient air

(b) Stage 2 (All control volumes)

Refrigerant

qMe

qT,j

Ambient air

(a) Stage 1 (All control volumes)

Refrigerant

qMe

Ambient air

80

CV 2

Refrigerant

(c) Stage 3

qMe

qT,2

hc,w(Tw,2 –Ta)Aw-a

Tw,2 Coil metal

Water layer

Lsf,mf,2 Frost cpmw,2Tw,2 (to Circuit 3)

Ambient air

cpmw,1Tw,1 (from Circuit 1)

CV 3

Refrigerant

qMe

qT,3

hc,w(Tw,3 –Ta)Aw-a

Tw,3 Coil metal

Water layer

Lsf,mf,3 Frost

Ambient air

cpmw,2Tw,2 (from Circuit 2)

cpm w,3Tw,3 (to Tray C)

Refrigerant

(d) Stage 4 (All control volumes)

hc,w (Tw,j –Ta)Aw-a

qT,j Coil metal

Tw,j Water layer

mv,jLv

Ambient air

hc,d (Tr,j –Ta)Ad-a qMe

Fig. 4.2 Schematics of mass and energy flows in the four defrosting stages for Model 1.

Modeling study on uneven defrosting

81

the retained water to the ambient air. The Nusselt number and the corresponding coefficients of natural convective heat transfer were used [27]: NuL ¼

1 hc H ¼ 0:13ðGr PrÞ3 , for 109 < GrPr < 1013 λa

(4.25)

where Gr ¼

gβðTw  Ta Þ 3 H ν2

(4.26)

Furthermore, the heat transferred from the water layer to the frost layer for melting the frost and to the ambient air was:

 hw Tw, j  Ttp A0 ¼ Lsf mf , j + hc, w Tw, j  Ta Af a ðj ¼ 1  3Þ

(4.27)

Third stage: Frost melting with water flowing away from a circuit Unlike in Stages 1, 2, or 4, as shown in Fig. 4.2C, in this stage, there were differences for the energy and mass flows in the three control volumes, as the melted frost started to flow away from a circuit. For Circuit 1, no melted frost flowed into it as it was at the highest level, but the melted frost in this circuit flowed down into Circuit 2. For Circuit 2, the melted frost in this circuit flowed into Circuit 3. Finally, the melted frost in Circuit 3 flowed into water-collecting Tray C and was collected by Cylinder C, as shown in Fig. 4.1. As shown in Fig. 4.2C1, energy conservation in Control Volume 1 (Circuit 1) yielded: q1 ¼ Lsf mf ,1 + cp Mw, max

dT w,1 + cp mw,1 Tw,1 + hc, w ðTw, 1  Ta ÞAf a + qMe dt (4.28)

where Mw, max is the maximum of the melted frost held on the surface of a refrigerant circuit. Mass conservation in this control volume was: mw,1 ¼ mf ,1

(4.29)

For the other two control volumes, as shown in Fig. 4.2C2 and C3, energy conservation yielded: qj + cp mw, j1 Tw, j1 ¼ Lsf mf , j + cp Mw, max

dT w, j + cp mw, j Tw, j dt


 + hc, w Tw, j  Ta Af a + qMe ðj ¼ 2, 3Þ

(4.30)

82

Defrosting for Air Source Heat Pump

Mass conservation was: mw, j ¼ mf , j + mw, j1 ðj ¼ 2, 3Þ

(4.31)

Fourth stage: Water layer vaporizing In this stage, the surface of the outdoor coil was free of frost, and the vaporization of the retained water took place. As illustrated in Fig. 4.2D, energy conservation in Control Volume j yielded:


 d Mw, j Tw, j + hc, w Tw, j  Ta Awa + hc, d Tr, j  Ta Ada qj ¼ c p dt + mv, j Lv + qMe ðj ¼ 1  3Þ

(4.32)

where hc, w(Tw, j  Ta)Awa is the heat transferred to low-temperature ambient air from the water layer in Control Volume j, and hc, d(Tr, j  Ta)Ada is the heat transferred to the ambient air from the high-temperature dry coil surface in Control Volume j. Mass conservation in Control Volume jwas: ðt Mw, j ¼ Mw, max  mv, j dt ðj ¼ 1  3Þ

(4.33)

0

where mv, j is the mass flow rate of the retained water vaporized from Circuit j and was proportional to the difference in vapor pressure between the exposed water surface and ambient air, as described by Mills [28], expressed as:
 mv, j ¼ hD Awa ρvsj  ρva ðj ¼ 1  3Þ

(4.34)

In Eqs. (4.32) and (4.34), Awa is the effective wetted airside surface area of a refrigerant circuit. As the vaporizing process went on, the area was diminishing. The relationship between the effective wetted airside surface area and the equivalent airside surface area of a refrigerant circuit was [8,14]:  Awa ¼ A0

Mw, j Mw, max

1:5 (4.35)

Then, the effective dry airside surface area of a refrigerant circuit in this stage,Ada, was: Ada ¼ A0  Awa

(4.36)

Moreover, the coefficient of convective mass transfer, hD, was related to the coefficient of natural convective heat transfer, hc, according to the Lewis Analogy [29,30]:

Modeling study on uneven defrosting

hD ¼

hc

83

(4.37)

2

cp ρa ðLeÞ3

where Le is the Lewis number for air and water vapor mixtures and a fixed value of 0.845 was used in this section [24].

Model 2 development (only Stage 3) Schematics of mass and energy flows in defrosting Stages 1, 2, and 4 for Model 2 are the same as those for Model 1 shown in Fig. 4.2, except that in Stage 3 in Model 2, which is shown in Fig. 4.3. Hence, for Model 2, all the equations for Stages 1, 2, and 4 were identical to those in Model 1, except those for Stage 3. In Stage 3, energy conservation in Control Volume j yielded: qj ¼ Lsf mf , j + cp Mw, max


 dT w, j + cp mw, j Tw, j + hc, w Tw, j  Ta Af a + qMe ðj ¼ 1  3Þ dt (4.38)

where Mw, max is the maximum of the melted frost held on the surface of Circuit j. Mass conservation in Control Volume j was: mw, j ¼ mf , j ðj ¼ 1  3Þ

(4.39)

Modeling a water-collecting tray and a water-collecting cylinder As part of the entire setup of the three-circuit outdoor coil, a mathematical submodel for the heat and mass flows on a water-collecting tray and a water-collecting cylinder was also developed, and used together with the two models. As seen from Fig. 4.1, for Model 1, the melted frost flowing away from the three circuits was collected by Cylinder C via Collecting Tray C. However, for Model 2, the melted frost flowing away from Circuits 1 to 3 was collected by Cylinders A, B, and C, via Collecting Trays A, B, and C, respectively. As shown in Fig. 4.4, there were three steps for the process of the mass and energy flows in a water-collecting tray during defrosting. Step 1 started during Stage 3 defrosting and the melted frost started to flow away from a refrigerant circuit and into a water-collecting tray. The energy conservation of the melted frost collected on the tray in Step 1 required: d ðMtw Ttw Þ dt  hc Atw ðTa  Ttw Þ ðj ¼ 3 forModel 1 and j ¼ 1  3 for Model2Þ

cp mw, j Tw, j ¼ cp

(4.40) where Mtw is the accumulated mass of the retained melted frost in the water-collecting tray, which can be evaluated using Eq. (4.3). In addition, Tw, j is the temperature of the retained water on the surface of the tubes and fins of each circuit (as shown in Fig. 4.1),

Defrosting for Air Source Heat Pump

qMe

Refrigerant

Fig. 4.3 Schematics of mass and energy flows in Stage 3 of defrosting for Model 2.

qT,j Tw,j Coil metal

Water layer

hc,w(Tw,j-Ta)Aw-a

Ambient air

84

Lsf,mf,j Frost cpmw,jTw,j (to water collecting tray)

and could be evaluated using Eqs. (4.28) and (4.30). Ttw is the temperature of the melted frost collected and Ta is the temperature of the ambient air during defrosting. As time went by, more and more water would be accumulated on the tray. When the melted frost started to flow away from the tray and into its connecting cylinder, Step 1 was ended and Step 2 started. As shown in Fig. 4.4B, energy conservation in Step 2 yielded: dT tw + cp mw, j Ttw dt  hc Atw ðTa  Ttw Þ ðj ¼ 3 forModel 1 and j ¼ 1  3 for Model2Þ

cp mw, j Tw, j ¼ cp Mtw, max

(4.41) where Mtw, max is the maximum mass of the retained water that can be held on the collecting tray. When there was no more melted frost flowing away from the water-collecting tray, Step 2 was ended and Step 3 commenced. As shown in Fig. 4.4C, energy conservation for the collected melted frost in the cylinder in Step 3 required: cp

d ðMtw Ttw Þ ¼ mv, tw Lv + hc Atw ðTa  Ttw Þ dt

(4.42)

where mv, tw is the rate of vaporization for the melted frost vaporized from the cylinder. As shown in Fig. 4.4C, Step 3 started during Stage 3 defrosting, and ended after Stage 4 defrosting was over. Eqs. (4.40)–(4.42) are the governing equations for evaluating the temperatures of the melted frost collected in the collecting cylinder.

The method of solving the two models When solving the two semiempirical models, Euler’s method [31] was applied to solving all the differential Eqs. (4.2)–(4.4), (4.13), (4.22), (4.23), (4.27)–(4.33), and (4.38)–(4.42). The mass flow rate and temperature of the melted frost flowing away from the upper control volume were regarded as the same as those of the melted frost entering an adjacent lower control volume. Moreover, based on Assumption (viii), the thermal properties of the retained water leaving the control volume were regarded as those of the melted frost collected in the respective water-collecting cylinders. Fig. 4.5 shows the computational algorithm for the four defrosting stages for the two models.

Modeling study on uneven defrosting

hcAtw(Ta-Ttw) cpMtwTtw

Ttw (From Circuit j) cpmw,jTw,j Ttw

(b) Step 2

Fig. 4.4 Schematics of mass and energy flows in a water-collecting tray and a water-collecting cylinder during defrosting.

Tray j

hcAtw(Ta-Ttw) cpMtwTtw

Tray j

Stage 3

(a) Step 1

(From Circuit j) cpmw,jTw,j

85

cpmcTtw

(c) Step 3

mv,twLv

hcAtw(Ta-Ttw) cpMtwTtw

Ttw

Cylinder j

Stage 4

(To Cylinder j) (To ambient air)

Start Input initial parameters

Initial Stage = 1

Stage 1

Stage 2

Stage 3

Frost melting without water flowing away from a circuit

Preheating

N t

Frost melting with water flowing away from a circuit N

N t

40 s

Mf,j

90 s Y

Y Stage = 2

Y

Stage = 3

Stage 4

1050 g

Stage = 4

Water layer vaporizing

Te,3

24 oC

N

Y Outputs End

Fig. 4.5 The computational algorithm for the four defrosting stages in two models.

86

Defrosting for Air Source Heat Pump

4.2.2 Experimental validation of two models The validation of the two empirical models developed was carried out by comparing the experimental data and the predicted data using the models for the key operating parameters of the experimental ASHP unit, including the tube surface temperature at the outlet of each circuit, the total mass of melted frost collected, and the defrosting duration. The experimental data separately reported in the previous section were used for validation purposes.

4.2.2.1 Validation of Model 1 Model 1 was validated by comparing the predicted defrosting duration, the tube surface temperatures at the exit of each circuit, the temperature variations of the melted frost collected in Cylinder C, and the total mass of the melted frost collected with the corresponding experimental data. Both measured and predicted tube surface temperatures at the exit of each circuit are shown in Fig. 4.6. Model 1 predicted that, at 186 s into the defrosting process, the surface temperature at the exit of Circuit 3 reached 24°C, the same as the experimental defrosting duration reported in the previous section. As seen from the diagram, the curves representing the predicted and measured tube surface temperatures at the exits of the three circuits agreed very well during the entire defrosting process. Compared with the measured data, the maximum deviations of predicted results for Circuits 1–3 were 1.8°C, 1.2°C, and 2.4°C respectively. The average deviations between measured and predicted results for Circuits 1–3 were 0.2°C, 0°C, and 0.4°C, respectively. 28 C1 (Measured) C2 (Measured) C3 (Measured)

Tube surface temperatures (ºC)

24

C1 (Predicted) C2 (Predicted) C3 (Predicted)

20 16 12 8 4 0 80

90

100 110 120 130 140 150 160 170 180 190 200 Time (s)

Fig. 4.6 Comparison between the measured and predicted tube surface temperatures at the exit of each circuit (Model 1).

Modeling study on uneven defrosting

87

2.1

o

Temperature of melted frost collected ( C)

Measured

Predicted

1.8 1.5 1.2 0.9 0.6 0.3 100

120

140

160

180

200

220

240

260

280

Time (s)

Fig. 4.7 Comparison between the measured and predicted temperatures of melted frost in Cylinder C (Model 1).

Fig. 4.7 shows the measured and predicted temperatures of the melted frost collected in Cylinder C. As seen in Fig. 4.7, the predicted temperature of the melted frost collected using the model developed agreed reasonably well with the experimental data. The maximum and average deviations between the measured and the predicted results were 0.68°C and 0.05°C, respectively. Furthermore, as reported in the previous section, the experimental total mass of melted frost collected in Cylinder C was 931 g. Using Model 1, the predicted total melted frost mass was 933 g, with only 0.2% difference.

4.2.2.2 Validation of Model 2 Model 2 was validated by comparing the following four operating parameters: defrosting duration, tube surface temperatures at the exit of each circuit, temperature variations of the melted frost collected in three water collecting cylinders, and the mass of the melted frost collected in each cylinder. The measured and predicted refrigerant tube surface temperatures at the exit of the three circuits are shown in Fig. 4.8. Model 2 predicted that, at 168 s into the defrosting process, the surface temperature at the exit of each circuit reached 24°C, which was the same as the experimental results reported in the previous section. Overall, the curves representing the measured and predicted data agreed well during the entire defrosting process. Compared with the measured data, the maximum deviations of predicted results for Circuits 1–3 were 2.5°C, 1.8°C, and 1.2°C respectively. The average deviations between measured and predicted results for Circuits 1–3 were 0.7°C, 0.6°C, and 0.1°C, respectively.

88

Defrosting for Air Source Heat Pump

28 C1 (Measured) C2 (Measured) C3 (Measured)

o

Tube surface temperatures ( C)

24

C1 (Predicted) C2 (Predicted) C3 (Predicted)

20 16 12 8 4 0 80

90

100

110

120

130

140

150

160

170

180

Time (s)

Fig. 4.8 Comparison between the measured and predicted tube surface temperatures at the exits of the three circuits (Model 2).

0.6

o

Temperature of melted frost collected ( C)

Measured

Predicted

0.5 0.4 0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3 – 0.4 100

120

140

160

180

200

220

240

260

280

Time (s)

Fig. 4.9 Comparison between the measured and predicted temperatures of melted frost collected in Cylinder A (Model 2).

Moreover, Figs. 4.9–4.11 show the measured and predicted temperatures of the melted frost collected in each of the three water-collecting cylinders. Fig. 4.9 shows the measured and predicted temperatures of the melted frost collected in Cylinder A during defrosting. As seen, the predicted temperature of the melted frost collected

Modeling study on uneven defrosting

89

0.6

o

Temperature of melted frost collected ( C)

Measured

Predicted

0.5 0.4 0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3 –0.4 100

120

140

160

180

200

220

240

260

280

Time (s)

Fig. 4.10 Comparison between the measured and predicted temperatures of melted frost collected in Cylinder B (Model 2).

0.6 Measured

Predicted

0.4

o

Temperature of melted frost ( C)

0.5

0.3 0.2 0.1 0.0 –0.1 –0.2 –0.3 –0.4 100

120

140

160

180

200

220

240

260

280

Time (s)

Fig. 4.11 Comparison between the measured and predicted temperatures of melted frost collected in Cylinder C (Model 2).

in the cylinder agreed well with the experimental results. For melted frost collected in Cylinders B and C, their predicted and measured temperatures also agreed reasonably well, as shown in Figs. 4.10 and 4.11, respectively. Compared with the measured data, the maximum deviations of predicted results for melted frost collected in Cylinders A,

90

Defrosting for Air Source Heat Pump

B, and C were 0.084°C, 0.085°C, and 0.068°C, respectively. The average deviations between measured and predicted results were 0.002°C, 0.002°C, and 0.010°C, respectively. Furthermore, using Model 2, the predicted masses of the melted frost collected in Cylinders A, B, and C were all at 327 g. As reported separately in the previous section, their corresponding experimental values were 317 g, 328 g, and 324 g, respectively. Therefore, the largest difference was only at 3% for Cylinder A. From the comparisons presented above, it was considered that the two empirical models were experimentally validated, and the validated models can be further used to quantitatively analyze the defrosting performances of an ASHP unit as discussed in the following section.

4.2.3 Predicting results by using the validated Model 1 4.2.3.1 Conditions of model extrapolation As shown in Figs. 4.6 and 4.8, the deviation of tube surface temperature is very small, and could be accepted. Here, in this work, we used the validated defrosting Model 1, but changed the input parameters. The refrigerant temperature at the inlet and outlet of each circuit as well as the refrigerant mass flow rate during defrosting were input into this validated dynamic defrosting model. These parameters are separately shown in Figs. 4.12 and 4.13. As seen, the defrosting durations for each circuit reached 24°C at 168, 175, and 181 s for Circuits 1–3, respectively. In Fig. 4.13, the curve of the mass flow rate of the refrigerant clearly shows us three stages. From 0 to 70 s into defrosting, it fluctuated a lot as increasing. It should contain the preheating stage and part of the frost melting without water flowing away from a circuit stage. Then, 40 Inlet temp in Circuit 1 Inlet temp in Circuit 2 Inlet temp in Circuit 3 Outlet temp in Circuit 1 Outlet temp in Circuit 2 Outlet temp in Circuit 3

36

o

Temperature of refrigerant ( C)

32 28 24 20 16 12 8

175 s

4

168 s

0 0

20

40

60

80

181 s

100 120 140 160 180 200 220 240 Defrosting time (s)

Fig. 4.12 Refrigerant temperature in each circuit.

Modeling study on uneven defrosting

91

11 Circuit 1 Circuit 2 Circuit 3

Mass flow rate of refrigerant (g/s)

10 9 8 7 6 5 4 3

75 s

2

165 s

1 0

20

40

60

80

100 120 140 160 180 200 220 240 Defrosting time (s)

Fig. 4.13 Refrigerant mass flow rate during defrosting.

it steeply increased from 70 to 160 s. The melted frost took a lot of energy from the refrigerant, and thus the flow rate also increased quickly. When it reached 160 s into defrosting, the rate of energy consumption would decrease due to the melted frost flowing away. Therefore, the refrigerant mass flow rate decreased suddenly and was kept at a steady value. All the statuses of the refrigerant mass flow rate result from the heat and mass transfer status during physical defrosting, but also are reflected in the defrosting parameters in this model as the input values.

4.2.3.2 Predicting results Some of the modeling results were reported here, as shown in Figs. 4.14–4.17. The input values were from 0 to 225 s, and thus the horizontal axis, were also at this period. In Fig. 4.14, it is the temperature of the melted water on the surface of each circuit during defrosting, just after the melted frost flowing away from the circuit. This parameter is very hard or even impossible to measure in an experiment due to the dynamic heat and mass transfer, and is important to develop the model as an intermediate variable. This is the reason why we choose to investigate it with this model. As seen, it was kept at nearly 0°C from 0 to 85 s, which is at the first stage, the preheating stage. There is a water layer between the frost and tube surface, and thus the water temperature is always the temperature of the mixture of frost and water. When it comes to 85–145 s, it kept at increase slowly, this is because the water layer is increasing as the frost melting. Meanwhile, as the frost layer decreased, the water layer was heated. From 145 s to the end of defrosting, the melted water as kept heated, and thus its temperature always increased. At the end of the simulation, it reached nearly 7°C for Circuit 1. At the same time, it is easy to find that the temperature at the three circuits is different. As the melted frost flowed downward, the water temperature of

92

Defrosting for Air Source Heat Pump

8 Circuit 1 Circuit 2 Circuit 3

o

Temperature of melted water ( C)

7 6 5 4 3 2 1

170 s

145 s

0 0

20

40

60

80

100 120 140 160 180 200 220 240 Defrosting time (s)

2

40 35

-4

Thermal resistance of refrigerant , 10 (Km )/W

Fig. 4.14 Temperature of melted water on each circuit’s surface.

30

Circuit 1 Circuit 2 Circuit 3

25 Period 5

20 Period 4

15

Period 3

10 Period 2

5

Period 1 85 s

0 0

20

40

60

80

145 s 170 s 190 s

225 s

100 120 140 160 180 200 220 240 Defrosting time (s)

Fig. 4.15 Thermal resistance of refrigerant during defrosting.

Circuit 1 is the highest and that of Circuit 3 the lowest. At 170 s, there is a sudden decrease. This is also the start of a quick increase for the three curves. Thus, at this time point, the main heat transfer role changed from heating the melted frost to vaporizing it. Fig. 4.15 presents the thermal resistance of the refrigerant during defrosting, which is also impossible to measure in an experiment. It clearly shows the thermal resistance changing at five periods.

Modeling study on uneven defrosting

93

5.0 4.5

Circuit 1 Circuit 2 Circuit 3

The mass of melted frost (g)

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

85 s

0.0

145 s 155 s

–0.5 0

20

40

60

80

100 120 140 160 180 200 220 240 Defrosting time (s)

Fig. 4.16 The mass of melted frost during defrosting.

5500 Circuit 1 Circuit 2 Circuit 3

5000 Energy used from refrigerant (J)

4500 4000 3500 3000 2500 2000 1500 1000

Water vaporizing

500 Water flowing Forst melting 85 s 145 s

0 –500 0

20

40

60

80 100 120 140 160 180 200 220 240 Defrosting time (s)

Fig. 4.17 Energy used from the refrigerant at each 5 s.

Figs. 4.16 and 4.17 show the mass of the melted frost during defrosting and the energy used from the refrigerant at each 5 s, respectively. The two figures also show the changing at different stages, with the time points at 85 and 145 s into defrosting. This is just at the second and third stages in the model. As shown in Fig. 4.16, the frost is mainly melted from 85 to 145 s, and the difference between the three circuits is clear. The peak values for Circuits 1–3 are 4.3, 4.0, and 3.7 g, respectively.

94

Defrosting for Air Source Heat Pump

Meanwhile, as shown in Fig. 4.17, the sudden increase point occurs at 150, 155, and 160 s for Circuits 1–3, with their peak value at about 5000, 4700, and 4400 J, respectively.

4.2.4 Potential uses and limitations 4.2.4.1 Potential uses of the two models developed The two experimentally validated semiempirical models can have a number of potential applications. First, they can be used to quantitatively analyze not only the negative effects of the downward-flowing melted frost due to gravity, but also the impact of locally draining away the melted frost using water-collecting trays on RCD performance in an ASHP unit. Second, the two models can be used to optimize the design of an outdoor coil and positioning water-collecting trays. Third, a number of operating parameters for an ASHP unit, which are difficult to measure experimentally, could be predicted, for example, the frost melting rate and the temperature of retained water over the surface of the outdoor coil. Also, with the validated models, the energy consumption on melting frost, vaporizing the melted frost, heating low-temperature ambient air, and heating the outdoor coil metal during an RCD process could be evaluated. Finally, the models will help increase awareness and spur efforts in exploring and maximizing the potential of ASHP units to realize greater defrosting efficiency.

4.2.4.2 Limitations of the modeling work Although the two models were developed based on energy and mass flow conservations, there exist a few limitations. These included the 10 listed assumptions, which were necessary in model developments but introduced errors into the models. However, the errors were still within acceptable levels. Moreover, certain empirical formulas were used, such as Eq. (4.20), which had their limitations. Furthermore, experimental data were also used in assisting the model development, thereby making the two models empirical. Therefore, appropriate modifications might have to be introduced when the models are to be used for studying ASHP units with different configurations or operating conditions. Nonetheless, the models developed could adequately describe the defrosting performance for the experimental ASHP unit with local drainage of the melted frost from its outdoor coil. In this section, following on an experimental study on draining away locally the melted frost for an experimental ASHP unit having a three-refrigerant-circuit outdoor coil using water-collecting trays between circuits, a modeling study on the defrosting process, at the two experimental settings of with and without the use of watercollecting trays between circuits, was carried out. Two empirical models, corresponding to the two settings, were therefore developed. Based on the validated defrosting Model 1, without trays installed, the model extrapolation was also carried out, with some modeling results given. The validated models have their limitations, but also could adequately describe the defrosting performance for the experimental ASHP unit with local drainage of the melted frost from its outdoor coil.

Modeling study on uneven defrosting

4.3

95

Alleviating uneven defrosting for an ASHP unit

An effective alternative to experimentally investigate the defrosting performance in an ASHP unit is via a numerical approach and therefore, the last two decades saw a growing number of modeling studies on defrosting operations [32,33]. Noticeable, Krakow et al. first developed a hot-gas defrosting model for evaporators [11,12], and later presented an idealized RCD model for an ASHP unit with a receiver [13,14]. Similar to Krakow, a detailed transient simulation model for hot-gas bypass defrosting in an air-cooled evaporator was also developed by Dopazo et al. [21]. However, in the above-mentioned defrosting models, an uneven defrosting phenomenon was not considered. Thus, the effects of the downward flowing of the melted frost due to gravity along the surface of a multicircuit outdoor coil on the defrosting performance all were neglected by assuming either no water retention on the coil surface or a stable water layer. Only in 2012 when a semiempirical model was developed by Qu et al. were the negative effects of melted frost considered for the first time [22]. Thereafter, as a follow up to Qu’s study on uneven defrosting, a series of experimental and modeling studies on the defrosting performance of an ASHP unit when the melted frost was drained away locally from its three-circuit outdoor coil was carried out by the authors and separately reported in previous sections. While the outcomes from these studies demonstrated the effectiveness of locally draining away the melted frost from a vertical multicircuit outdoor coil, for existing ASHP units, however, it is hardly possible to install water-collecting trays between circuits. Nonetheless, for existing ASHP units, it is still possible to vary the heat input to each refrigerant circuit through varying the refrigerant supply to each circuit. This is because uneven defrosting was fundamentally caused by different thermal loads imposed to each circuit due to the downward flowing of the melted frost, when the supply of refrigerant or heat to each circuit was the same. Consequently, if the heat to be supplied to each circuit may be varied according to the actual defrosting thermal load each circuit is to deal with, then the problem of uneven defrosting may be alleviated. Modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to vary the refrigerant flow, thus the heat input to each circuit. Therefore, in this section, a modeling study on varying the heat (via refrigerant) supply to each refrigerant circuit in the three-circuit outdoor coil to alleviate uneven defrosting was carried out and reported. First, the methodology and three study cases are explained. Second, the results of the modeling study on defrosting durations and energy use in the three study cases are presented. Finally, a conclusion is given.

4.3.1 Methodology and study cases The reported study was carried out using a previously developed and validated semiempirical mathematical model at the experimental setting of not using watercollecting trays between circuits. In this section, a brief description of the previous experimental study is first introduced, and the validated models are then shown. This is followed by presenting three study cases in this section. Finally, the assumptions used for the three study cases will be given.

96

Defrosting for Air Source Heat Pump

4.3.1.1 Experimental study To quantitatively study the negative effects of the downward flowing of the melted frost during RCD, an experimental ASHP unit with a vertical three-circuit outdoor coil was specifically established. As introduced in previous sections, it was modified from a commercially available 6.5 kW heating-capacity variable speed ASHP unit. The experimental ASHP unit was installed in an existing environmental chamber having a simulated heated indoor space and a simulated outdoor frosting space. The sizes of both spaces were each measured at 3.8 m (L)  3.8 m (W)  2.8 m (H). Fig. 3.1 shows the schematics of the ASHP unit installed in the environmental chamber. The experimental ASHP unit was a split-type one, consisting of a swing-type compressor, an accumulator, a four-way valve, an electronic expansion valve, an indoor coil, and an outdoor coil. The outdoor coil was specially designed and made, as shown in Fig. 4.18. There were three individual and parallel refrigerant circuits and the airside surface areas corresponding to each of the three circuits were the same. The outdoor coil was vertically installed, and in each circuit a solenoid modulating valve and a manual stop valve were used. The specifications of the three-circuit outdoor coil are detailed in the previous section. The experimental conditions were jointly maintained by the use of a separate airconditioning system in the environmental chamber, and sensible and latent load generating units (LGUs), which were used to simulate the thermal load. During normal heating (or frosting) operation, a frosting environment in the outdoor space was maintained by running the experimental ASHP unit and LGUs together while an indoor heated environment by the experimental ASHP unit and the existing air conditioning system. All the parameters, such as temperature, pressure, relative humidity, refrigerant mass flow rate, voltage, current, etc., were measured. All sensors and measuring devices were able to output direct current signals of 4–20 mA or 1–5 V to a data-acquisition system (DAS) for logging and recording. All the measured data throughout both the frosting and defrosting periods were collected and recorded by the DAS at an interval of 5 s. In addition, during defrosting, photos for surface conditions of the outdoor coil were taken at an interval of 10 s. Three cases were designed and carried out in this experimental study, and two of them were prototypes of the following models developed. For Case 1, there were no water-collecting trays installed between circuits, as shown in Fig. 4.18A. During defrosting, the melted frost could downward flow from the up-circuits to the downcircuits freely due to gravity. As shown in Fig. 4.18B, for Case 3, there were two water-collecting trays (Tray A and Tray B) installed between circuits. The melted frost would be taken away when it was downward flowing away from the circuit during defrosting, and thus the negative effects of the melted frost were eliminated. The experimental results were compared, and the negative effects of downward-flowing melted frost were qualitatively studied. At the same time, a water-collecting tray was suggested for installation between circuits for a multicircuit outdoor coil when optimizing its structure, and thus improving system defrosting performance. On the other hand, more information was obtained from this experimental study. First, all the data used in the following modeling development as known parameters were

590 mm 44 mm

152 mm

Circuit 2

22 mm

i

Te

MV1

From compressor SV2

To compressor Te

MV2

500 mm

Circuit 1

Circuit 1

Heating mode Defrosting mode Ti Ti SV1 T

97

Circuit 2

To EEV

Circuit 3

MV3

Measuring cylinder

Distributor Water collecting tray (slope 37 )

Heating mode Defrosting mode Ti Ti SV1 Ti Te

590 mm 44 mm

MV1

From compressor

152 mm

22 mm

A SV2

To compressor Te

A

MV2

To EEV

B Te

Circuit 3

SV3 From EEV

B

MV3

C Distributor

(B)

Circuit 1

Header

(A)

Circuit 2

Te

Circuit 3

SV3 From EEV

Water collecting tray (slope 37 )

500 mm

Header

Modeling study on uneven defrosting

Measuring cylinder

C

Fig. 4.18 Details of the vertical three-parallel refrigerant circuit outdoor coil without and with a water-collecting tray installed between circuits. (A) Without water collecting tray installed between circuits (Case 1 in Section 3.3). (B) With water collecting tray installed between circuits (Case 3 in Section 3.3).

collected, such as the tube surface temperature at the inlet of each refrigerant circuit, the refrigerant mass flow rate, the total mass of melted frost collected, etc. Second, part of the experimental results would be used in the validation stage of the following two developed models. However, there are still some limitations in this experimental study. First, some parameters could not be measured, such as the rate of melted frost downward flowing along the surface of the outdoor coil. Second, some parameters such as the

98

Defrosting for Air Source Heat Pump

temperature of the melted frost when it was downward flowing, were hardly impossible to be measured. Third, some parameters were easy to measure, but not accurate due to their fluctuations, such as the temperature of the air around the outdoor coil. In the experimental study, this type of parameter was always measured by many sensors and/or tested many times. Therefore, the following modeling study takes the responsibility to measure the previous three types of parameters.

4.3.1.2 Modeling study The development of the validated semiempirical model at the setting of without using water-collecting trays between circuits for the experimental ASHP unit was separately introduced previously. However, for the completeness of the current section, it is briefly described here. The prototype of this model was Case 1 in the previous experimental study. Details of the vertical three-circuit outdoor coil without a watercollecting tray installed between circuits were shown in Fig. 4.18A, and a conceptual model for the airside of the three-circuit outdoor coil is shown in Fig. 4.19. Before the model was built, a series of assumptions were given, such as the frost at the surface of each circuit was even, the refrigerant mass flow was evenly distributed, and so on. At the same time, as mentioned previously, some of experimental results obtained were used as known parameters when the model was developed. In this model, a defrosting process on the airside of the outdoor coil was divided into four stages: (1) preheating, (2) frost melting without water flowing away from a circuit, (3) frost melting with water flowing away from a circuit, and (4) water layer vaporizing. The schematics of mass and energy flows in the four defrosting stages were shown in Fig. 4.2. Moreover, as part of the entire setup of the three-circuit outdoor coil, a mathematical submodel for the heat and mass flows on a water-collecting tray and a water-collecting cylinder was also developed and used together with the model. As shown in Fig. 4.4, there were three steps for the process of the mass and energy flows in a water-collecting tray during defrosting. Outdoor coil (air side)

Refrigerant in

Thermal-couple mw,j–1Tw,j–1

mw,jTw,j

mw,j+1Tw,j+1

Circuit 1 mf,j

Circuit 2

Circuit 3

Water collecting tray C Refrigerant out

C

Water collecting cylinder C

Fig. 4.19 Conceptual model for the airside of the three-circuit outdoor coil.

Modeling study on uneven defrosting

99

During the modeling study, 42 equations were used. Although the models were developed based on energy and mass flow conservations, there existed a few limitations. Those included the assumptions introducing errors, certain empirical formulas having their limitations, and experimental data making the models empirical. Therefore, appropriate modifications might have to be introduced when the models are to be used for studying ASHP units with different configurations or operating conditions. After this model was built, it was validated by comparing the predicted defrosting duration, the tube surface temperatures at the exit of each circuit, the temperature variations of the melted frost, and the total mass of the melted frost collected in a measuring cylinder shown in both Fig. 3.1 and Fig. 4.18A, with the corresponding experimental data [11, 17]. The average deviations between the measured and predicted results of the tube surface temperatures at the exits of Circuits 1–3 were 0.2°C, 0°C, and 0.4°C, respectively. The maximum and average deviations between the measured and the predicted results of the temperature variations of the melted frost were 0.68°C and 0.05°C, respectively. Finally, the difference between the measured and predicted total melted frost mass was only 0.2%. Therefore, this validated model at the setting of without using water-collecting trays between circuits could adequately describe the defrosting performance for the experimental ASHP unit and was used in the current modeling study.

4.3.1.3 The three study cases When an ASHP unit operates at defrosting mode, usually the refrigerant discharged from the compressor is assumed to be equally distributed into each circuit of a multicircuit outdoor coil. As shown in Fig. 4.13, the refrigerant mass flow rates in the three circuits during defrosting in the previous experimental study were calculated. From 0 to 70 s, the refrigerant mass flow rates were fluctuating, named Stage 1 in this section. At Stage 2, from 70 to 160 s, their values increased steadily, with their peak values at 10.52 g/s at 160 s into defrosting. The following period was named Stage 3, and the refrigerant mass flow rate decreased first and then kept fluctuating to the termination of defrosting. It could be found that the refrigerant mass flow rates of the three circuits calculated were always kept the same during defrosting. However, as the melted frost flows downward along the surface of the outdoor coil due to gravity, the heating load for each refrigerant circuit becomes different. In this section, therefore, to alleviate the uneven defrosting due to the downward flowing of melted frost for the three-circuit vertical outdoor coil during RCD, three study cases were included where different openings of modulating valves were applied so as to vary the heat supply to each of the three circuits. Table 4.2 details the opening of the modulating valve and other operational changes in the three study cases. At the same time, to clearly describe their differences, changes in the proportion of the refrigerant distribution into each circuit in the three study cases were illustrated in Fig. 4.20. A. The opening values of the three valves on each circuit, from top to bottom, were fixed at 92.5%, 97.8%, and 100%, respectively. B. Fully open all valves at the start of defrosting, and fully close the modulating valve on Circuit 1 when its defrosting was terminated.

100

Defrosting for Air Source Heat Pump

Table 4.2 Experiment conditions in the three study cases Item

Parameter

Case 1

Case 2

Case 3

1

Opening of modulating valve Other operational changes Refrigerant changes shown in

A

B

C

None

None

D

Figs. 4.20A and 4.22

Figs. 4.20B and 4.23

Figs. 4.20C and 4.24

2 3

Circuit 1 (Re = 100%) Circuit 2 (Re = 100%)

Circuit 1 (Re = 95.6%) Fix the opening values since the start of defrosting

Circuit 3 (Re = 100%)

(A)

Circuit 3 (Re = 103.3%) After

Before

Circuit 1 (Re = 100%) Circuit 2 (Re = 100%)

Close valve on Circuit 1 when its defrosting was terminated

Circuit 3 (Re = 100%)

(B)

Circuit 2 (Re = 100%) Circuit 3 (Re = 100%)

(C)

Before

Circuit 1 (Re = 0%) Circuit 2 (Re = 150%) Circuit 3 (Re = 150%)

Before

Circuit 1 (Re = 100%)

Circuit 2 (Re = 101.1%)

After Close valve on Circuit 1 when its defrosting was terminated Reduce the compressor speed to 66.7%

Circuit 1 (Re = 0%) Circuit 2 (Re = 100%) Circuit 3 (Re = 100%) After

Fig. 4.20 Changes in the proportion of the refrigerant distribution into each circuit in the three study cases.

C. Fully open all valves at the start of defrosting, and fully close the modulating valve on Circuit 1 when its defrosting was terminated. D. Keep the refrigerant flows to Circuit 2 and Circuit 3 unchanged by reducing the compressor speed to 66.7% of the original speed when the modulating valve on Circuit 1 is closed.

Modeling study on uneven defrosting

101

4.3.1.4 Case 1 Fig. 8 shows the measured tube surface temperature at the exit of each circuit during defrosting in the previous experimental study in Case 1 in Section 3.3. It could be seen that the temperature order of the three circuits was kept at T1 > T2 > T3 during defrosting. This is because of the negative effects of the downward flowing melted frost. Experimental results show that the defrosting durations for the three circuits from top to bottom were 172, 182, and 186 s, respectively. In other words, the defrosting durations for the top and the middle circuits were 92.5% and 97.8%, respectively, of that for the bottom circuit. To alleviate the uneven defrosting, Study Case 1 was then designed, where the modulating valve for the bottom circuit was fully opened and the openings of the modulating valve for the top and middle circuits were set at 92.5% and 97.8% of full opening, respectively. In this way, the heat supplies to the three circuits via the supply of refrigerant during defrosting were no longer the same, and the assumed refrigerant mass flow rates to each circuit during defrosting are 95.6%, 101.1%, and 103.3% of their previous values, respectively, as shown in Fig. 4.20. Fig. 4.22 shows the assumed refrigerant mass flow rate in the three circuits during defrosting in Study Case 1. The trends of refrigerant mass flow rates in the three circuits during defrosting at three stages are the same as those shown in Fig. 4.13. However, their peak values at 160 s into defrosting were 10.10 g/s for Circuit 1, 10.62 g/s for Circuit 2, and 10.83 g/s for Circuit 3, respectively. During defrosting, the mass flow rate order was always at R3 > R2 > R1, which met the designed experiment conditions in Table 4.2 and the changes in the proportion of the refrigerant distribution shown in Fig. 4.20A.

4.3.1.5 Case 2 As shown in Fig. 4.21, the results from the previous experimental study also revealed that the defrosting duration for Circuit 1 was the shortest. Hence, it was also possible to vary the heat input to the three refrigerant circuits by fully closing the modulating valve on Circuit 1 when its tube surface temperature at exit reached 24°C, which means its defrosting process was terminated. Therefore, in Study Case 2, it was designed that the three valves on the three circuits were fully open at the start of defrosting. When the defrosting of Circuit 1 was terminated, the modulating valve on it would be fully closed. Consequently, more refrigerant would flow into the other two refrigerant circuits to speed up their defrosting. Fig. 4.20B illustrates the changes in the proportion of the refrigerant distribution into each circuit, and Fig. 4.23 shows the assumed refrigerant mass flow rates to each circuit during defrosting in Study Case 2. It could be found that the trends of refrigerant mass flow rates in the three circuits during defrosting at Stages 1 and 2 are the same as those shown in Fig. 4.13. But at Stage 3, the refrigerant mass flow rate of Circuit 1 decreased to 0 g/s at 172 s, as designed in Table 4.2. At the same time, the values of the other two circuits increased first and then kept fluctuating, with their same peak values at 12.59 g/s at 195 s into defrosting. All their trends met the changes in the proportion of the refrigerant distribution shown in Fig. 4.20B.

102

Defrosting for Air Source Heat Pump

28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature ( C)

24 20 16

T1 > T2 > T3

12 8 186 s 4

182 s

0

172 s 80

100

120

140 Time (s)

160

180

200

Fig. 4.21 Measured tube surface temperature at the exit of each circuit during defrosting in the previous experimental study (Case 1 in Section 3.3).

4.3.1.6 Case 3 As shown in Figs. 4.20B and 4.23, in Study Case 2, when the modulating valve on Circuit 1 was closed, the refrigerant mass flow rates to the other two circuits were increased. To possibly reduce defrosting energy consumption, however, it was possible to decrease the compressor speed so that the refrigerant flow rates to Circuits 2 and 3 remained unchanged. Therefore, in Study Case 3, it was designed that the three modulating valves on the three circuits were fully open at the start of defrosting. When the tube surface temperature at the exit of Circuit 1 reached 24°C, its modulating valve would also be fully closed, the same as that in Study Case 2. However, the compressor speed was also reduced by 1/3 at the same time. The changes in the proportion of refrigerant distribution into each circuit during defrosting in Study Case 3 were shown in Fig. 4.20C, and the assumed refrigerant mass flow rates to each circuit are shown in Fig. 4.24. The same as that shown in Fig. 4.23, the trends of refrigerant mass flow rates in the three circuits during defrosting at Stages 1 and 2 are the same as those shown in Fig. 4.13. In addition, at Stage 3, the refrigerant mass flow rate of Circuit 1 decreased to 0 g/s at 172 s, as designed in Table 4.2. At the same time, the values of the other two circuits increased first and then kept fluctuating, with their same peak values at 12.59 g/s at 195 s into defrosting. All their trends met the changes in the proportion of the refrigerant distribution shown in Fig. 4.20C.

4.3.1.7 The seven assumptions In the current modeling study, the following were also assumed:

Modeling study on uneven defrosting

103

(i) In the previous experimental study, the refrigerant mass flow rate in the three refrigerant circuits was assumed to be evenly distributed. The calculated refrigerant mass flow rates in the three circuits in the previous experimental study were derived following this assumption. (ii) In the three study cases, the refrigerant mass flow rate passing through a modulating valve to each circuit during defrosting was assumed to be proportional to the respective percentage openings of the three modulating valves, under a constant total refrigerant flow rate. For example, when the percentage openings of the valves are 50% for the valve on Circuit 1 and 100% for the valves on Circuits 2 and 3, respectively, the ratio of the three valves’ openings is 1:2:2, and thus the percentage shares of the total refrigerant mass flow rate passing through the three modulating valves are 20%, 40%, and 40%, respectively. The assumed refrigerant mass flow rates to each circuit during defrosting in Study Case 1 shown in Figs. 4.20A and 4.22 were derived following this assumption. (iii) In Study Case 2, the total refrigerant mass flow was evenly distributed to the other two refrigerant circuits during defrosting after the modulating valve on Circuit 1 was closed. As a result, the refrigerant mass flow rates to Circuit 2 and Circuit 3 were each increased by 50%, as illustrated in Fig. 4.20B. (iv) In Study Case 3, as the modulating valve on Circuit 1 was closed, the refrigerant mass flow in Circuit 2 and Circuit 3 remained unchanged as a result of compressor speed reduction by 33%. (v) The energy consumption on the compressor would decrease 33% as a result of a compressor speed reduction by 33%. (vi) When the tube surface temperature at the exit of a refrigerant circuit reached 24°C, the defrosting operation on that circuit was considered ended. The experiment conditions illustrated in Figs. 4.20B and C were derived following this assumption. (vii) The defrosting duration for the ASHP unit was the same as that of Circuit 3.

14 Circuit 1

Refrigerant mass flow rate (g/s)

12

Circuit 2

Circuit 3

10.83 g/s

10 10.62 g/s 8

10.10 g/s

R3 > R2 > R1

6 Stage 3: Decrease and then fluctuate

4 Stage 1: Fluctuate

2

Stage 2: Increase steadily

70 s

0 0

20

40

60

160 s 80

100

120

140

160

180

200

Time (s)

Fig. 4.22 Assumed refrigerant mass flow rate in the three circuits during defrosting in Study Case 1.

104

Defrosting for Air Source Heat Pump

4.3.2 Results and analysis Using the validated empirical model introduced in the previous section, a modeling study for the three study cases was undertaken and the study results are shown in Figs. 4.25–4.27 for the three study cases. In addition, to illustrate the effectiveness of varying heat supply to the respective refrigerant circuit, the results of the previous experimental study for the setting of not using water-collecting trays between circuits shown in Fig. 4.21 were also used for comparison purposes. In Figs. 4.21 and 4.25–4.27, for their time (horizontal) axis, although defrosting starts at 0, 80 s is the chosen time point for these figures in order to more clearly show the temperature rise during defrosting. Further, Table 4.3 summarizes the defrosting durations in the previous experimental study and the three study cases. For the results presented here in the three study cases, the time difference in defrosting duration between Circuit 1 and Circuit 3, Δt, was used as a parameter to indicate the degree of uneven defrosting. The smaller the value of Δt, the uneven defrosting was eliminated much more effectively.

4.3.2.1 Case 1 Fig. 4.25 shows the variations of the predicted tube surface temperatures at the exit of each circuit during defrosting in Study Case 1. It can be seen that the defrosting durations were 178 s for Circuit 1 and Circuit 2, and 183 s for Circuit 3, respectively. From 95 to 180 s into defrosting, the temperature order of the three circuits was kept at T1 > T2 > T3. However, the temperature order changed to T2 > T1 > T3 after 180 s into defrosting, because the refrigerant mass flow rate distributed into Circuit 1 was less than that in the others. This phenomenon met the changes in the proportion of the refrigerant distribution into each circuit in Study Case 1, as shown in Figs. 4.20A and 4.22. It also can be found that the defrosting durations for Circuit 2 and Circuit 3 were shortened, but that for Circuit 1 was slightly extended as compared to the experimental results shown in Fig. 4.21. This was because the refrigerant supply to each circuit was no longer the same. Following Assumption (ii) specified in Section 2.4, as shown in Figs. 4.20A and 4.22, the refrigerant mass flow rate during defrosting in Circuit 1 decreased to 95.6% of the previous value, and that in Circuits 2 and 3 increased to 101.1% and 103.3% of the previous values, respectively. Compared

Table 4.3 Defrosting durations in the previous experimental study and the three study cases Defrosting durations for each circuit Case No.

Circuit 1

Circuit 2

Circuit 3

Shown in

Experimental study Study Case 1 Study Case 2 Study Case 3

172 s 178 s 172 s 172 s

182 s 178 s 177 s 179 s

186 s 183 s 179 s 187 s

Figs. Figs. Figs. Figs.

4.21 and 4.25 and 4.26 and 4.27 and

4.28 4.28 4.28 4.28

Modeling study on uneven defrosting

105

to the results from the previous experimental study, the defrosting duration for Circuit 3 or the ASHP unit was decreased by 3 s, or 1.6%. Also, as seen, Δt was 5 s, which is much shorter than the experimental value of 14 s, suggesting that the uneven defrosting was significantly alleviated.

4.3.2.2 Case 2 Fig. 4.26 shows the variations of the predicted tube surface temperatures at the exit of each circuit during defrosting in Study Case 2. The simulation results demonstrated that the defrosting for the durations was 173 s for Circuit 1, 176 s for Circuit 2, and 179 s for Circuit 3, respectively. From 95 to 175 s into defrosting, the temperature order of the three circuits was kept at T1 > T2 > T3. When the tube surface temperature at the exit of Circuit 1 reached 24°C at 173 s into defrosting, its modulating valve was closed so that the refrigerant supply to Circuit 1 was reduced to 0 g/s, as shown in Figs. 4.20B and 4.23. At the same time, because the compressor speed remained unchanged, the refrigerant supplies to Circuit 2 and Circuit 3 were consequently increased to 150% of previous values. As a result of the increase in refrigerant mass flow rate, it can be seen from Fig. 4.26 that the tube surface temperatures were increased for Circuit 2 and Circuit 3, but decreased for Circuit 1 after the closing valve, from 175 s into defrosting to the termination. Compared to the results of the previous experimental study, the defrosting duration for Circuit 3 or the ASHP unit was decreased by 7 s, or about 3.8%. The value of Δt was 7 s, also suggesting the alleviated uneven defrosting.

14 Circuit 1

Circuit 2

Circuit 3

12 Refrigerant mass flow rate (g/s)

12.59 g/s 10 10.52 g/s 8 6

Stage 3: Decrease and then fluctuate

4 Stage 1: Fluctuate

Stage 2: Increase steadily

2

172 s 70 s

0 0

20

40

60

160 s 80

100

120

140

160

180

200

Time (s)

Fig. 4.23 Assumed refrigerant mass flow rate in the three circuits during defrosting in Study Case 2.

106

Defrosting for Air Source Heat Pump

4.3.2.3 Case 3 Fig. 4.27 shows the variations of the predicted tube surface temperatures at the exit of each circuit during defrosting in Study Case 3. The simulation results showed that the defrosting durations were 172 s for Circuit 1, 182 s for Circuit 2, and 186 s for Circuit 3, respectively. As shown in Figs. 4.20C and 4.24, when the modulating valve on Circuit 1 was closed at 172 s into defrosting, its refrigerant mass flow rate was reduced to 0 g/s. For the other two circuits, the refrigerant mass flow rates remained unchanged after the compressor speed was reduced by one-third. As seen, the surface temperature for Circuit 1 was reduced after closing the valve at 172 s, but those for Circuit 2 and Circuit 3 continued their increasing trend. Unlike the other two study cases, the defrosting duration for Circuit 3 or the ASHP unit was slightly increased by 1–187 s, when compared with the results of the previous experimental study. At the same time, Δt was also slightly increased by 1–15 s, suggesting that the problem of uneven defrosting remained. The simulation results for the three study cases are summarized in Table 4.3, where the results of the previous experimental study are also included. To clearly illustrate the difference of these durations, the analysis on durations for the three study cases and the previous experimental and modeling studies are shown in Fig. 4.28. In this figure, the time difference in defrosting duration between the study case and the previous experimental study, △ dt, was used as a parameter to indicate the defrosting duration. The bigger the value of △ dt, the earlier the defrosting process of this study case was terminated. It can be seen from Table 4.3 and Fig. 4.28 that for the three study cases, Study Case 2 appeared to be the better one in terms of shortening the defrosting duration with the shortest duration of 179, 7 s earlier than the defrosting termination of the 12 Circuit 1

Circuit 2

Circuit 3

Refrigerant mass flow rate (g/s)

10 10.52 g/s 8 6

Stage 3: Decrease and then fluctuate

4 Stage 1: Fluctuate

2

Stage 2: Increase steadily

70 s

0 0

20

40

160 s 60

80

100

120

140

172 s 160

180

200

Time (s)

Fig. 4.24 Assumed refrigerant mass flow rate in the three circuits during defrosting in Study Case 3.

Modeling study on uneven defrosting

107

28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature at exit ( C)

24 20 T1 > T2 > T3

16

T2 > T1 > T3 12 8

183 s

4

95 s

178 s

0 80

100

120

140

160

180

200

Time (s)

Fig. 4.25 Predicted tube surface temperatures at circuit exits during defrosting in Study Case 1. 28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature at exit ( C)

24 20

T1 > T2 > T3

16 T 1 is decreasing.

12 8

179 s

4

177 s

95 s

172 s

0 80

100

120

140

160

180

200

Time (s)

Fig. 4.26 Predicted tube surface temperatures at circuit exits during defrosting in Study Case 2.

previous experimental study. Also, it can be seen from Fig. 4.28 that the values of Δt for Study Case 1 and Study Case 2 were significant smaller than the experimental value and that for Study Case 3, suggesting that using the methods in Study Cases 1 and 2 can help alleviate uneven defrosting for a better defrosting performance. However, because different defrosting durations for the three study cases were resulted in, the energy use for defrosting was therefore different. This is further discussed in the next section.

108

Defrosting for Air Source Heat Pump

28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature at exit ( C)

24 20 16 T1 was decreasing.

T1 > T2 > T3 12 8

187 s

4

179 s

95 s

172 s

0 80

100

120

140

160

180

200

Time (s)

Fig. 4.27 Predicted tube surface temperatures at circuit exits during defrosting in Study Case 3.

C1

C2

C3

Experimental study 175 s

170 s

180 s

185 s

190 s

185 s

190 s

Δt = 14 s C1,C2

Study Case 1 175 s

170 s

C3 180 s Δt = 5 s Δdt = 3 s

C2

C1

C3

Study Case 2 175 s

170 s

190 s

185 s

180 s

Δt = 7 s Δdt = 7 s C1

Study Case 3 170 s

C2 175 s

180 s

C3 185 s

190 s

Δt = 14 s Δdt = –1 s

Fig. 4.28 Analysis on durations for the previous experimental study and the three study cases.

4.3.3 Analysis on heat supply and energy consumption For an ASHP unit, during RCD, the energy is used to heat the outdoor coil metal, melt frost, heat the melted frost, heat the cold ambient air, and evaporate the retained water on the surface of the outdoor coil. In this section, the total energy use for defrosting was also evaluated for the three study cases, 715.9 kJ for Case 1, 693.2 kJ for Case 2,

Modeling study on uneven defrosting

109

90 Compressor

Supply fan

Indoor air

80 73.2 kJ 68.6 kJ

Heat supply for defrosting (kJ)

70 60

56.5 kJ

50 92.2%

86.3%

40

33.8 kJ

30

86.3% 86.6%

20 1.0%

10

1.0% 12.5%

12.7%

0 Experimental study

Study Case 1

1.1% 6.7%

1.0% 12.4%

Study Case 2

Study Case 3

Fig. 4.29 Heat supply for defrosting after 172 s into defrosting in the previous experimental study and the three study cases. 450 Frost melting

Vaporizing

400 363.2 kJ

Energy use for defrosting (kJ)

350

357.7 kJ

361.1 kJ

334.5 kJ

34.2 kJ

28.8 kJ

5.5 kJ

32.2 kJ

329.0 kJ

329.0 kJ

329.0 kJ

329.0 kJ

300 250 200 150 100 50 0 Experimental study

Study Case 1

Study Case 2

Study Case 3

Fig. 4.30 Energy consumption during defrosting in the previous experimental study and the three study cases.

and 728.0 kJ for Case 3, respectively. Compared with the experimental value of 732.6 kJ, the defrosting energy uses in the three study cases were all less, with that defrosting energy use in Study Case 2 being the lowest at about 94.6% of the experimental value. It is noted that in Study Case 3, the compressor speed was reduced by

110

Defrosting for Air Source Heat Pump

one-third for possible energy savings after the tube surface temperature at the exit on Circuit 1 reached 24°C. However, because the total durations of the defrosting operation were longer than those in Study Case 2, the total energy use in Study Case 3 was more than that in Study Case 2, and slightly lower than the experimental value. To clearly show the differences of heat supply, Fig. 4.29 shows the heat supply for defrosting after 172 s into defrosting in the previous experimental study and the three study cases. It is obvious that the heat supply for defrosting in Study Case 2 was the shortest, at 33.8 kJ. At the same time, the three study cases all could decrease the heat supply for defrosting because the value in the previous experimental study was the biggest, at 73.2 kJ. The heat supply mostly comes from the thermal energy of indoor air, accounting for more than 86% of the total heat supply. The least part is the heat supply from the compressor, at only around 1%. In addition, the energy consumption during defrosting in the previous experimental study and the three study cases is shown in Fig. 4.30. It could be found that the energy consumption on frost melting for different study cases was the same, at 329.0 kJ. This is because the frost accumulation on the surface of the outdoor coil was assumed to be the same as that in the previous experimental study. The differences totally come from the energy consumption on vaporizing. The biggest value of energy consumption on vaporizing is 34.2 kJ in the previous experimental study. And the smallest one is only 5.5 kJ, or 16.1% of the value in the previous experimental study. Figs. 4.29 and 4.30 show that the heat supply for defrosting and energy consumption in Study Case 2 were both the shortest. This further confirmed that system performance could be improved most by fully closing the modulating valve on the top circuit when its defrosting terminated in the three study cases for alleviating uneven defrosting. In this section, a modeling study on alleviating uneven defrosting for a vertical three-circuit outdoor coil in an ASHP unit during RCD was undertaken. The following conclusions could be received: (1) Three study cases were included and the study results suggested that the best operating defrosting performances in terms of defrosting durations and energy use were achieved in Study Case 2. In this section, defrosting energy use could be decreased to 94.6% as well as a reduction of 7 s in defrosting duration by fully closing the modulating valve on the top circuit when its defrosting terminated. (2) It is expected that with more refrigerant circuits in an outdoor coil in an ASHP, the method of fully closing the modulating valves on the top circuit when its defrosting is terminated will yield a better defrosting performance for the ASHP unit, as predicted by the modeling study reported in this section. (3) In this modeling study, frost accumulation on the surface of each circuit and the refrigerant distributed into each circuit in the multicircuit outdoor coil both were assumed to be even. However, when the frost accumulation and refrigerant distribution were uneven, the performances of the three study cases would be different. Therefore, a model should be further developed with the considerations of frost accumulation and refrigerant distribution for an ASHP unit with a multicircuit outdoor coil. (4) Compared with the previous experimental study, although the heat supply and energy consumption in Study Case 3 were both decreased by 4.6 kJ, its defrosting duration was extended by 1 s. Therefore, in consideration of the indoor thermal comfort requirements, this type of control strategy was not suggested.

Modeling study on uneven defrosting

4.4

111

Concluding remarks

Following on the experimental work on defrosting performance of an experimental ASHP unit having a three-circuit outdoor coil draining away locally the melted frost by using water-collecting trays, a modeling study on the dynamic defrosting process, at the two experimental settings of with and without the use of watercollecting trays between circuits, was carried out. Two empirical models corresponding to the two settings were therefore developed. It was the first time for a set of defrosting models focusing on the uneven defrosting performance for a multicircuit outdoor coil, but not for a whole heat pump system. It is also the first time for a defrosting model to fully consider the melted frost locally drained away and the energy stored in the metal of the coil. Additionally, the two models were validated with experimental data, with the totally same experimental conditions considered. Therefore, the two models were expected to be used to predict some physical parameters difficult or hardly possible to be measured in a dynamic defrosting process. Finally, the two models could help increase awareness and spur efforts in exploring and maximizing the potential uses of ASHP units to realize more energy savings. Moreover, by using the validated defrosting Model 1 corresponding to a typical traditional ASHP unit, without trays installed between circuits, the model extrapolation was carried out. A series of modeling results was quantitatively predicted, including the temperature of melted water on each circuit’s surface during defrosting, the thermal resistance of the refrigerant during defrosting, the mass of melted frost during defrosting, and the energy used from the refrigerant at each 5 s. Additionally, based on the validated Model 1, three control strategies were proposed and tested in the model. Results showed that the best operating defrosting performances in terms of defrosting durations and energy use were achieved by fully closing the modulating valve on the top circuit when its defrosting was terminated. However, as discussed, the two validated models have their limitations, including the listed assumptions, the validation errors, empirical formulas, and the experimental data from the fixed configuration and operating conditions for an ASHP unit, etc. For example, the frost accumulation on each circuit was assumed evenly distributed at the start of a defrosting process. But in practice, it is impossible that the frost was evenly formed and accumulated on the surface of each circuit in a vertically installed multicircuit outdoor coil. This was because of the uneven distribution for the refrigerant inside the tube and as the entrance air outside the tube. Moreover, the refrigerant was also assumed evenly distributed during defrosting for each circuit in the preliminary model extrapolation work. Clearly, this is impossible because the heating loads imposed on the outdoor coil kept changing. The uneven distributed frost on the surfaces of each circuit would further promote uneven refrigerant distribution, and thus result in uneven defrosting. Therefore, the frost accumulation condition at the start of a defrosting process and the refrigerant distribution during defrosting should be further studied, and are presented in the next chapter.

112

Defrosting for Air Source Heat Pump

References [1] Young DJ. Development of a northern climate residential air-source heat pump. ASHRAE Trans 1980;86(1):671–86. [2] Watters RJ, O’Neal DL, Yang JX. Frost/defrost performance of a three-row fin staged heat pump evaporator. ASHRAE Trans 2002;108(2):318–29. [3] Yang DK, Lee KS, Song S. Fin spacing optimization of a fin-tube heat exchanger under frosting conditions. Int J Heat Mass Transf 2006;49:2619–25. [4] Cai L, Wang RH, Hou PX, Zhang XS. Study on restraining frost growth at initial stage by hydrophobic coating and hygroscopic coating. Energy Build 2011;43:1159–63. [5] Mei VC, Gao Z, Tomlinson JJ. Frost-less heat pump. ASHRAE Trans 2002;108(1):452–9. [6] Hu WJ, Jiang YQ, Qu ML, Yao Y, Deng SM. An experimental study on the operating performance of a novel reverse-cycle hot gas defrosting method for air source heat pumps. Appl Therm Eng 2011;31(2):363–9. [7] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse-cycle defrosting method for air source heat pumps. Build Environ 2010;45 (11):2354–61. [8] O’Neal DL, Peterson KT, Anand NK, Schliesing JS. Refrigeration system dynamics during the reversing cycle defrost. ASHRAE Trans 1998;95(2):689–98. [9] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part I: experiments. Appl Energy 2012;91:122–9. [10] Wang ZY, Wang XX, Dong ZM. Defrost improvement by heat pump refrigerant charge compensating. Appl Energy 2008;85:1050–9. [11] Krakow KI, Yan L, Lin S. A model of hot gas defrosting of evaporators, Part 1: heat and mass transfer theory. ASHRAE Trans 1992;98(1):451–61. [12] Krakow KI, Yan L, Lin S. A model of hot gas defrosting of evaporators, Part 2: experimental analysis. ASHRAE Trans 1992;98(1):462–74. [13] Krakow KI, Lin S, Yan L. An idealized model of reversed-cycle hot gas defrosting of evaporators, Part 1: theory. ASHRAE Trans 1993;99(2):317–28. [14] Krakow KI, Lin S, Yan L. An idealized model of reversed-cycle hot gas defrosting of evaporators, Part 2: experimental analysis and validation. ASHRAE Trans 1993;99 (2):329–38. [15] Alebrahim AM, Sherif SA. Electrical defrosting analysis of a finned tube evaporator coil using the enthalpy method. Proc Inst Mech Eng Part C J Mech Eng Sci 2002;216 (6):655–73. [16] Sherif SA, Hertz MG. A semi-empirical model for electric defrosting of a cylindrical coil cooler. Int J Energy Res 1998;22(1):85–92. [17] Al-Mutawa NK, Sherif SA. An analytical model for hot-gas defrosting of a cylindrical coil cooler, Part I: model development. ASHRAE Trans 1998;104(1):1722–30. [18] Al-Mutawa NK, Sherif SA. An analytical model for hot-gas defrosting of a cylindrical coil cooler, Part II: model results and conclusions. ASHRAE Trans 1998;104(1):1731–7. [19] Liu ZQ, Tang GF, Zhao FY. Dynamic simulation of air source heat pump during hot-gas defrost. Appl Therm Eng 2003;23(6):675–85. [20] Hoffenbecker N, Klein SA, Reindl DT. Hot gas defrost model development and validation. Int J Refrig 2005;28(4):605–15. [21] Alberto Dopazo J, Fernandez-Seara J, Uhı´a FJ, Diz R. Modeling and experimental validation of the hot-gas defrost process of an air-cooled evaporator. Int J Refrig 2010;33 (4):829–39.

Modeling study on uneven defrosting

113

[22] Qu ML, Xia L, Deng SM, Jiang YQ. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part II: modeling analysis. Appl Energy 2012;91:274–80. [23] Dong JK, Deng SM, Jiang YQ, Xia L, Yao Y. An experimental study on defrosting heat supplies and energy consumptions during a reverse cycle defrost operation for an air source heat pump. Appl Therm Eng 2012;37:380–7. [24] Shah MM. A general correlation for heat transfer during film condensation inside pipes. Int J Heat Mass Transf 1979;22(4):547–56. [25] Dittus FW, Boelter LMK. Heat transfer in automobile radiators of the tubular type. Univ Calif Publ Eng 1930;2(13):443–61. [26] Ye HY, Lee KS. Performance prediction of a fin-and-tube heat exchanger considering airflow reduction due to the frost accumulation. Int J Heat Mass Transf 2013;67:225–33. [27] Mills AF. Basic heat and mass transfer. 2nd ed. Upper Saddler River; NJ: Prentice Hall; 2000. [28] Jaluria Y. Natural convection. HMT, the science and applications of heat and mass transfer. vol. 5. New York: Pergamon Press; 1980. [29] Padki MM, Sherif SA, Nelson RM. A simple method for modeling the frost formation phenomenon in different geometries. ASHRAE Trans 1989;95:1127–37. [30] Threlkeld JL. Thermal environmental engineering. 2nd ed. Englewood cliffs, NJ: Prentice-Hall, Inc.; 1970. [31] Gerald CF, Wheatley PO. Applied numerical analysis. 7th ed. Addison-Wesley Publishing Co.; 2003. [32] Qu ML, Xia L, Deng SM, Jiang YQ. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part II: modeling analysis. Appl Energy 2012;91:274–80. [33] Song MJ, Deng SM, Xia L. A semi-empirical modeling study on the defrosting performance for an air source heat pump unit with local drainage of melted frost from its three-circuit outdoor coil. Appl Energy 2014;136:537–47.

Investigation of effect on uneven defrosting performance 5.1

5

Introduction

With the growing demand for electricity worldwide, environmental aspects in connection with energy consumption, such as global warming, ozone layer depletion, and high-levels of pollution, especially the PM 2.5 air pollution in Beijing in China, have become a main concern while heavily influencing global energy policy. It is necessary to emphasize the use of emerging and well-known renewable energy resources and different energy conservation approaches. An ASHP unit, utilizing low-grade energy in the air as a source, has the advantages of simple operation, high efficiency, no pollution, ability to provide both cooling and heating, etc. [1]. Accordingly, as a key technology under the clean development mechanism (CDM) to mitigate climate change and avoid global warming [2], it has become widely used as cooling and heating sources for heating, ventilation, and air-conditioning over the recent decades [3]. However, when operated at heating mode under an extremely cold and high humidity environment, frost would appear and accumulate over the outdoor coil’s surface in an ASHP unit, which severely deteriorates the system operating performance. Therefore, it is necessary to implement periodical defrosting to maintain its normal operation. Currently, there are many defrosting methods investigated for ASHP units, and the most widely used standard defrosting method is RCD. When an ASHP unit is operated at RCD mode, its outdoor coil, which is usually installed vertically for space saving, acts as a condenser and the indoor coil acts as an evaporator. On the other hand, in order to minimize the refrigerant pressure loss along the tube inside and enhance the heat transfer between the inside refrigerant and the outside ambient air via tubes and fins, a multicircuit outdoor coil is usually used in ASHP units. The downwardflowing melted frost helps form or reinforce a water layer between the frost and the coil surface, which introduces a thermal resistance [4] and thus reduces the heat transfer between the two [5]. It is reported that when defrosting at the top circuits is terminated, the bottom ones are still covered with frost [6]. Also, when the tube surface temperature at the exit of the top circuit reaches the preset defrosting termination temperature, the temperature of the bottom circuit is much lower [7–10]. Thereafter, the negative effects of downward-flowing melted frost due to gravity are demonstrated [7, 11, 12], with water-collecting trays installed between circuits to improve defrosting efficiency. However, for a vertically installed multicircuit outdoor coil, it is hard to avoid the uneven defrosting phenomenon. This means it is nearly impossible for each circuit to reach the preset defrosting termination temperature at the same time. Thus, the energy consumption due to uneven defrosting can hardly be saved. Consequently, avoiding uneven defrosting for a multicircuit outdoor coil in an ASHP unit becomes a technical Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00005-9 © 2019 Elsevier Ltd. All rights reserved.

116

Defrosting for Air Source Heat Pump

problem. After uneven defrosting was eliminated, the defrosting performance should be quantitatively evaluated. Moreover, as mentioned in previous model developments, the melted frost flows downward because the gravity force is larger than the surface tension. This implies that the melted frost could remain on the surface of the outdoor coil due to surface tension. In experiments, it was also observed that some residual water was retained on the downside surface of the circuit in the outdoor coil due to surface tension. As demonstrated, the melted frost downward flowing due to gravity has negative effects on defrosting performance. Therefore, how the residual water influences defrosting should also be explained. Also, it is valuable to quantitatively evaluate the defrosting performance when considering the residual water retained on the downside of the circuit considered. In this chapter, studies around the aforementioned problems will be introduced.

5.2

Effects of melted frost elimination on uneven defrosting

On identification of the negative effects of downward-flowing melted frost, a traditional vertical multicircuit outdoor coil is suggested to be installed horizontally to reduce the flow path of melted frost during defrosting [13]. As shown in Fig. 5.1A–B, when the vertically installed three-circuit outdoor coil [11] is horizontally installed, the maximum flow path of melted frost over the coil surface can be shortened from 500 mm to 44 mm, being reduced 11.36 times. As illustrated in Fig. 5.1C–D, the flow directions of hot refrigerant and cold melted frost during defrosting are also changed from opposite to orthogonal, which effectively shortened their heat transfer length. Consequently, a better defrosting performance is expected. However, it was found in the previous experimental studies that there was some melted frost remaining on the downside of each circuit due to surface tension during defrosting [11, 12, 14]. From the definition of surface tension [15], it is concluded that the total mass of retained water is directly proportional to the total area of circuit downsides. During defrosting, the retained water would consume energy [16], and thus delay the defrosting process. As shown in Fig. 5.1E–F, when the installation type of the three-circuit outdoor coil is changed, the total area of retained water is increased, from 590 mm
44 mm to 590 mm
500 mm, being increased 11.36 times. Therefore, it is contradictory for the maximum flow path of melted frost and the total area of retained water to improve system defrosting performance. On the other hand, although horizontal heat exchangers are reported by many studies [9, 11, 12, 14, 17–19], few of them are related to a coiled heat exchanger. Most of them are horizontal ground heat exchangers [20, 21], tube heat exchangers [27,28], or flat-panel heat exchangers [9,29]. Notably, Abdel-Wahed RM et al. experimentally studied a horizontal flat-plated cooling surface. Their results indicate that the decrease in the thickness of the frost layer is approximately linear with the defrosting time [9]. However, it is not RCD, but hot water defrosting. Later, Hambraeus et al. carried out an experimental setup with a horizontal evaporator to study the heat transfer of a special refrigerant, with the effects of melted frost neglected [22]. In 2012, an

Investigation of effect on uneven defrosting performance

m m

Side A (Topside) Side B (Back side)

44

590 mm

117

Circuit 1 Circuit 3

m m

500 mm

Side C (Front side)

Circuit 2

50 0

Circuit 2

Circuit 1 Circuit 3 Side A 590 mm (Front side)

Side B (Downside)

(B)

Flow direction of cold melted frost

Flow direction of hot refrigerant

500 mm

Circuit 1

Refrigerant in

Refrigerant out

(A)

Side C (Downside) 44 mm

Refrigerant out Flow direction of cold melted frost

Circuit 2 500 mm Flow direction of hot refrigerant

Circuit 3

(C)

Refrigerant in

(D)

m m

Circuit 3 m m

Circuit 2

50 0

44

Side B 590 mm

Circuit 1

(E)

(F)

Side C

590 mm

Fig. 5.1 The vertically installed outdoor coil changed to horizontally installed. (A) Vertically installed three-circuit outdoor coil. (B) Horizontally installed three-circuit outdoor coil. (C) Opposite flow directions. (D) Orthogonal flow directions. (E) Area of retained water (Side B). (F) Area of retained water (Side C).

experimental study on the comparison of heat transfer and pressure drop in a horizontal and vertical helically coiled heat exchanger with CuO/water-based nanofluids was reported, in which the convective heat transfer coefficient and friction factors were comparatively studied. However, their heat transfer performance on the airside was also neglected [23].

118

Defrosting for Air Source Heat Pump

In this section, to solve the previous contradictory problem and study the heat transfer performance of a horizontally installed heat exchanger, an experimental study on RCD performance for an ASHP unit with a horizontal multicircuit outdoor coil has been carried out. First, the ASHP unit under experiment is presented, followed by the experimental procedures and conditions. Thereafter, the experimental cases and their results are given. The defrosting durations and energy consumptions for each case study are measured and discussed, with a conclusion given at the end.

5.2.1 Experimental work 5.2.1.1 Experimental ASHP unit An experimental ASHP unit was specifically established for carrying out the experimental work reported in this section. Similar to the experimental setup used in previous sections, it was also modified from a commercially available 6.5 kW heating-capacity variable speed ASHP unit and was installed in the same existing environmental chamber, with a simulated indoor heated space and a simulated outdoor frosting space. As previously introduced, the sizes of both the indoor and outdoor spaces were each measured at 3.8 m (L)
3.8 m (W)
2.8 m (H). Fig. 3.1 shows the schematics of the ASHP unit installed in the environmental chamber. The experimental ASHP unit was a split-type one consisting of a swing-type compressor, an accumulator, a four-way valve, an electronic expansion valve, an indoor coil, and an outdoor coil. Additionally, to maintain the suitable experimental conditions in both the indoor and outdoor spaces, there was a separate A/C system in the environmental chamber, and the same LGUs were used to simulate thermal loads in the experiment. Finally, during normal heating (frosting) operation, a frosting environment in the outdoor space was maintained by running the experimental ASHP unit and LGUs together while an indoor heated environment was maintained by the experimental ASHP unit and the existing A/C system. Different from the previous experimental setup in Chapter 3, the outdoor coil was specially designed and made for this study, as shown in Fig. 5.2. There were three individual and parallel refrigerant circuits and the airside surface areas corresponding to each of the three circuits were equal. There were four wind boards installed on the two air sides of the outdoor coil, which were used to prevent the air passing the outdoor coil through separations between circuits. The outdoor coil was horizontally installed, and in each circuit an SV and an MV were fixed, with their locations shown in Fig. 5.2. In order to easily describe the process of frost melting, the topside and downside of the primary vertically installed three-circuit outdoor coil were named Side A and Side B, as shown in Figs. 5.1 and 5.2. Side C was the inlet air side of the outdoor coil, where the frost will be formed and accumulated. A 700 mm
750 mm water-collecting tray made of PVC placed under the outdoor coil was added to the experiment rig, and a 2000 mL water-collecting cylinder made of PVC was connected to the tray. Both of them would be used for collecting and measuring the melted frost. At the same time, in the experiments, the retained water on the surface of the fins, especially on the downside of each circuit, was absorbed by

Investigation of effect on uneven defrosting performance

To compressor From compressor Header

119

Heating mode Defrosting mode

To EEV From EEV

650 mm

Te

Side B

44 mm

Circuit 3

MV Ti SV 3 3 Circuit 2

Wind board

m 22 m m 2 15

ib ut or

Circuit 1

Side C

590 mm

D

ist r

m

Te MV1 Ti SV 1

M 326 70 elte mm 0 m d fr m os t

Te MV2 Ti SV2

Side A Water collecting tray (slope 37 )

Melted frost

750 mm Water collecting cylinder (2000 mL)

Fig. 5.2 Details of the horizontally installed three-circuit outdoor coil and locations of SVs and MVs.

preweighed cotton tissues. In this way, the melted frost from the outdoor coil during defrosting was collected and weighed. The specifications of the three-parallel refrigerant circuit outdoor coil are shown in Table 5.1. Fig. 5.3 shows the airside details of the outdoor coil in the experimental ASHP unit installed in the outdoor frosting space. On the windward side (Side C in Figs. 5.1–5.3), the air dry-bulb temperatures were measured at 6 points using thermocouples (Type K, of 0.75% accuracy) and the air wet-bulb temperatures at 3 points using temperature sensors (PT100, class A). In this way, there were two dry-bulb temperature sensors and two wet-bulb temperature sensors for each circuit. As seen in Fig. 5.3, they were placed at the same positions for each circuit. The average values from these measurements were used as the inlet air dry-bulb temperature and the wet-bulb temperature in the following calculations. On the other hand, the air temperature and humidity downstream of the outdoor coil were measured by a hygrosensor ( 0.2°C and 1.0% RH accuracy, respectively; the Testo Hygrotest 650 was located inside an air duct 700 mm downstream of the outdoor coil outlet. To ensure the best possible measuring accuracy, the air wet-bulb temperature sensors positioned on Side C of the outdoor coil were also calibrated using the hygrosensor. Furthermore, the air flow rate passing through the outdoor coil was measured by using a flow hood (of 3% accuracy) with a 16-point velocity grid located at the center of a 400 mm
400 mm air duct 600 mm long, as shown in Fig. 5.3. Precalibrated K-type thermocouples were also used for measuring the temperatures of the tube/coil and fin surfaces of the outdoor coil. Six were for measuring the refrigerant tube surface temperatures at both the inlets (Ti in Fig. 5.2) and exits (Te in Fig. 5.2) of the three refrigerant circuits. Three were affixed on the fin surface at the center of each circuit. These fin surface temperature sensors are located at the

120

Defrosting for Air Source Heat Pump

Table 5.1 Specifications of the outdoor coil in the experimental ASHP unit Item

Parameter

Value

Unit

1 2 3 4 5 6 7 8 9 10 15 11 12 13 14 15 16 17 19 20 21 22 23

Height of the outdoor coil Width of the outdoor coil Length of the outdoor coil Fin height Fin width Fin thickness Fin pitch Fin type Tube external diameter Tube thickness Tube spacing Circuit pitch Number of tube rows Number of circuits Number of water-collecting trays Number of water-collecting cylinders Number of wind boards Material of tube Material of fin Material of water-collecting tray Material of water-collecting cylinder Material of wind board Volume of water-collecting cylinder

44 590 500 44 152 0.115 2.1 Plate 10 0.5 20 22 2 3 1 1 2 Copper Aluminum PVC PVC Wood 2000

mm mm mm mm mm mm mm – mm mm mm mm – – – – – – – – –

600 mm

44 mm

Grid for measuring air velocity Air duct (1000×600×550 mm)

Side A

600 mm

Temperature and humidity sensor

Outdoor air fan

440 mm

1000 mm

Air duct (400×400 mm)

600 mm 250 mm Outdoor coil Dry bulb temperature sensor Wet bulb temperature sensor

Side C

Fig. 5.3 The airside details of the outdoor coil in the experimental ASHP unit.

mL

Investigation of effect on uneven defrosting performance

121

upward surface of the outdoor coil, the opposite side to Side C, so that the measured fin surface temperatures can avoid the effects of downward-flowing melted frost during defrosting. Furthermore, one more thermocouple was placed inside the watercollecting cylinder to measure the temperature of the melted frost collected, as shown in Fig. 5.2. On the other hand, for the experimental ASHP unit, refrigerant pressures were measured using pressure transmitters (Danfoss Pressure Transmitter, Type AKS 32 and AKS 33) with an accuracy of 0.3% of full scale reading and refrigerant volumetric flow rate by a variable area flow meter with a reported accuracy of 1.6% of full scale reading (KROHNE VA Flowmeter, H250). All sensors and measuring devices were able to output a direct current signal of 4–20 mA or 1–5 V, which can be transferred to a data acquisition system (DAS) for logging and recording. The DAS collected and recorded all the measured data throughout both frosting and defrosting at an interval of 5 s. In addition, during defrosting, photos for surface conditions of the outdoor coil were taken at an interval of 10 s.

5.2.1.2 Experimental procedures and conditions Prior to the defrosting operation, the experimental ASHP unit was operated in the heating (frosting) mode for 60 min at an outdoor frosting ambient temperature of 0.5  0.2°C (dry-bulb temperature) and 90  3% relative humidity, which was jointly maintained by the use of both the experimental ASHP unit and the LGUs placed in the outdoor frosting space. During heating (frosting), the air temperature inside the heated indoor space was maintained at 20  0.5°C, which was jointly maintained by the use of both the experimental ASHP unit and the existing A/C system. The experimental conditions are summarized in Table 5.2. Before defrosting was started, the compressor was first switched off. One minute after the compressor shutdown, the four-way valve was switched to defrosting mode. Four seconds later, the compressor was powered on again manually, and a defrosting operation was started. The defrosting operation was also manually terminated. However, different from the previous experimental studies introduced in Chapter 3, the defrosting operation was terminated when the tube surface temperature at the exits of the three refrigerant circuits in the outdoor coil all reached the preset Table 5.2 Experimental conditions Item

Parameter

Value

Unit

1 2 3 4 5 6 7

Air temperature in indoor heated space Air temperature in outdoor frosting space Air relative humidity in outdoor frosting space Face velocity of outdoor coil Face velocity of indoor coil at defrosting mode Face velocity of indoor coil at heating (frosting) mode Heating (frosting) operation duration

20 0.5  0.2 90  3% 1.3a 2.31 3.68 60

°C °C – m s1 m s1 m s1 min

a

The average value during a heating (frosting) operation. During heating (frosting), the face velocity decreased due to frost growth.

122

Defrosting for Air Source Heat Pump

Table 5.3 Measurement/calculation errors of system parameters Item

Parameter

1

6

Air dry-bulb temperature at upstream of the outdoor coil Air wet-bulb temperature at upstream of the outdoor coil Air temperature at downstream of the outdoor coil Air relative humidity at downstream of the outdoor coil Air flow rate passing through the outdoor coil Temperatures of tube/coil and fin surfaces

7

Refrigerant pressure

8

Refrigerant volumetric flow rate

9 10

Mass of the melted frost collected Temperature of the melted frost collected

11 12

Total energy supply for defrosting Total energy consumption during defrosting Defrosting efficiency Defrosting evenness value

2 3 4 5

13 14

Measurement/ calculation error

Unit

1 (K-type thermocouple) 0.1 (PT100, class A)

°C °C

0.2 (hygrosensor)

°C

1.0% (hygrosensor)

–

3% (flow hood)

–

1 (K-type thermocouple) 0.3% (pressure transmitters) 1.6% (variable area flow meter) 0.1 (weighing scale) 1 (K-type thermocouple) 1% (calculated) 1% (calculated)

°C

kJ kJ

0.15% (calculated) 5.9% (calculated)

– –

– – g °C

temperature, 24°C in this section [11, 12, 14, 18]. To supply enough energy for defrosting and keep the ASHP unit working safely, the indoor air fan during defrosting remained operational at a lower speed. However, the outdoor air fan was turned off or turned to blowing at the opposite direction as the same air quantity in different experimental cases, which was controlled by the inverter (Model: KASUGA KIDEN INVERTER, KKVF-F204ESB). Table 5.3 summarizes the measuring accuracy for various sensors/instruments used in the experimental ASHP unit, and the calculated relative standard errors.

5.2.1.3 Experimental cases A series of experimental works using the experimental ASHP unit have been carried out to study the defrosting performance for an ASHP unit with a horizontally installed multicircuit outdoor coil. In order to obtain meaningful experimental results, it was necessary to ensure that the frost that accumulated on the surface of the three circuits was even at first. For an ASHP unit with a multicircuit outdoor coil, it is hard to make

Investigation of effect on uneven defrosting performance

123

the frost evenly accumulate on the surface of the outdoor coil, as many parameters affect frosting performance [24]. However, in this section, modulating valves installed at the inlet refrigerant pipe to each circuit (as shown in Fig. 5.2) were deployed to vary the refrigerant flow to each circuit adjusted. Therefore, the frost accumulation on the surface of the three circuits was close to each other, with their biggest difference being less than 10% [11, 25, 26]. Second, to make the comparative study results meaningful, the frost accumulations in different cases should be close to each other. Therefore, in this section, the frosting duration was fixed at 60 min, as listed in Table 5.2. The frost accumulations could be calculated with the total mass of the melted frost collected, with the water vaporized into the ambient air during defrosting neglected. Fig. 5.4 shows the force analysis of retained water droplets on the surface of the outdoor coil, on the conditions of (a) on the side of a single fin, (b) between double fins, and (c) at the bottom of the fins. Four types of force—gravity force (G), fraction force (Ff), normal force (N), and surface tension (Fs)—would work on the water droplets. As the frost melted during defrosting, the melted frost would be held on the fin surface due to surface tension. As the water accumulated, the gravity effects of water increased. The melted frost would flow downward when its gravity exceeded the maximum surface tension [23]. When a vertical outdoor coil was horizontally installed, blowing the melted frost may be one of methods to increase the melted frost flowing away. As shown in Fig. 5.4, after the force of wind blowing (Fw) is added, the maximum surface tension was easy to exceed. Therefore, to comparatively and quantitatively study the defrosting performance of an ASHP unit having a horizontal multicircuit outdoor coil, Case 1 and Case 2 were designed and carried out in this section. In the two cases, the outdoor coils were vertically and horizontally installed, respectively. Therefore, their experimental results could be meaningfully compared. Furthermore, to study the effects of blowing the melted frost by reversing the outdoor air fan, Case 3 was also designed and conducted. In Case 3, the outdoor air fan was turned on and reversed blowing by a control strategy adjustment during defrosting when the tube surface temperature of one circuit reached 3°C, and the blowing was kept up for approximately 40 s. In this method, the wind was expected to blow the melted frost away by destroying its surface tension. Therefore, the effects of wind blowing could be conducted by comparative analysis of the Fin

Fin

Ff Droplet

Ff Droplet

Fb

Fs

Fin

Fs

Fs Fs Fs Fs

G Fw

G Fw

G

(A)

(B)

(C)

Droplets

Fw

Fig. 5.4 Force analysis of retained water droplets on the surface of the outdoor coil. (A) On side of single fin. (B) Between double fins. (C) At the bottom of fins.

124

Defrosting for Air Source Heat Pump

Fin

Fin Vaporizing due to heating from fins

Vaporizing due to heating from fins

Droplets

Droplets

Freely flowing due to gravity

Freely flowing due to gravity and blowing the water away

(A)

(B)

Fig. 5.5 Mass transfer of the retained water during defrosting in the three cases. (A) Case 1 and case 2. (B) Case 3. Table 5.4 Experimental conditions and relative results in the three cases Item

Parameter

Case 1

Case 2

Case 3

1

Installation type of outdoor coil Operation of outdoor air fan during defrosting Defrosting duration Total mass of the melted frost collected Total mass of the retained water collected Results shown in

Vertically installed Turn it off

Horizontal installed Turn it off

186 s

186 s

Horizontal installed Turn it on and reverse its direction to blow the coil 204 s

921 g

948 g

957 g

91 g

566 g

344 g

Figs. 5.6–5.8, 5.11; Tables 5.5–5.7

Figs. 5.6, 5.7, 5.9, 5.12; Tables 5.5–5.7

Figs. 5.6, 5.7, 5.10, 5.13; Tables 5.5–5.7

2

3 4

5

6

experimental results of Case 2 and Case 3. In addition, the melted frost kept vaporizing due to heating from the fins during defrosting. Fig. 5.5 illustrates the mass transfer of the retained water during defrosting in the three cases, and all the experimental conditions and relative results are listed in Table 5.4.

5.2.2 Results and analysis Six photographs illustrating the airside surface conditions of the outdoor coil at the start and end of defrosting in the three cases are shown in Fig. 5.6. As shown in Fig. 5.6A1, B1, and C1, it is visually the same and even for the frost accumulated on the surface of the outdoor coil in the three cases, which met the requirements

Investigation of effect on uneven defrosting performance

125

Fig. 5.6 Airside surface conditions of the outdoor coil at the start and end of defrosting in the three cases (six photographs).

previously described in Section 5.2.1. As listed in Table 5.4, the frost accumulations, or the total mass of the melted frost collected, were 921 g in Case 1, 948 g in Case 2, and 957 g in Case 3, respectively. Their biggest difference, between Case 1 and Case 3, was just 36 g, or about 3.76%, which was small and acceptable in this section. Meanwhile, as predicted, a lot of residual water was retained over the outdoor coil downside surface when the defrosting operation terminated, especially in Case 2. As shown in Fig. 5.6B2 and C2, the melted frost retained on the downside surface over

126

Defrosting for Air Source Heat Pump

the horizontal multicircuit outdoor coil in an ASHP unit due to surface tension during RCD can be visually observed. Their differences on the mass of retained water could be found in the white dotted rectangles, as shown in Fig. 5.6A2, B2, and C2. It is obvious that the retained water in Case 2 was much more than that in Case 3. In Case 1, there was nearly no melted frost found. The phenomenon met the mass of the retained water collected listed in Table 5.4, at 94 g in Case 1, 566 g in Case 2, and 344 g in Case 3, respectively. Therefore, a similar mass of frost accumulations, frost evenly accumulated, and total mass of the retained water collected with obvious differences make this comparative study meaningful. To analyze the effects of wind blowing the melted frost, eight photographs illustrating the frost melting and downward-flowing process on the airside of the horizontal three-circuit outdoor coil in Case 2 and Case 3 are shown in Fig. 5.7. These photographs show the airside conditions of the outdoor coil from 60 s to 120 s into the defrosting operation. At this period, there was a lot of melted frost downward flowing away from circuits in the two cases. Meanwhile, in Case 3, the fan was turned on at reversed direction while a lot of melted frost was blowing away. As shown in Fig. 5.7A1 and B1, when the defrosting came to 60 s, the fin surface started directly contacting the ambient air. However, there was still no melted frost flowing away from the circuit due to surface tension, as illustrated in Fig. 5.4. Also due to surface tension, after 60 s into the defrosting operation, the melted frost started flowing from side A to side B along the downside surface of each circuit, which was east to observe in Fig. 5.7A2 and B2. As the defrosting process went by, the mass of melted frost increased as it accumulated. As illustrated in Figs. 5.4 and 5.5, when the gravity direction total force of the melted frost exceeded the maximum of the surface tension, the melted frost began downward flowing from the circuit to the water-collecting tray. Therefore, from 60 s to 80 s, there were few melted frost drops downward flowing away from the circuit in the two cases. In addition, as shown in Fig. 5.7A3 and B3, at the positions indicated by the white arrows, a lot of melted frost kept downward flowing to the water-collecting tray. Especially at 100 s into defrosting, there was a lot of melted frost flowing away from the circuit. Meanwhile, due to the effects of wind blowing, more melted frost was flowing away from the circuit in Case 3. At 120 s into defrosting, there was still melted frost flowing away from the circuits in Case 2, as shown in Fig. 5.7A4. However, in Case 3, there was no frost flowing after the air fan was turned off, as shown in Fig. 5.7B4. Therefore, the effects of wind blowing on draining the melted frost were very obvious, which also met the total mass of the retained water collected listed in Table 5.4. The measured operating performances of the experimental ASHP unit during defrosting, corresponding to the three experimental cases are presented in Figs. 5.8–5.13. In all these figures, for their time (horizontal) axis, 60 s and 80 s are the chosen starting time in order to clearly show the temperature rise during defrosting. Figs. 5.8–5.10 present the measured tube surface temperatures at the exit of the three refrigerant circuits, and Figs. 5.11–5.13 show the measured fin surface temperatures at the center point of the three refrigerant circuits during defrosting. It is noted that the variation trends of these temperatures are similar to those reported by Qu and O’Neal [12, 22].

Investigation of effect on uneven defrosting performance

127

Fig. 5.7 Airside surface conditions of the outdoor coil during defrosting in Case 2 and Case 3 (eight photographs).

It can be seen from Fig. 5.8 that, in Case 1, the tube surface temperatures of the three circuits remained around 0°C during the first 100 s, and started to rise steadily thereafter. Temperatures reached 24°C at 172 s, 182 s, and 186 s, respectively [11]. It is demonstrated that the negative effects of downward-flowing melted frost due to

128

Defrosting for Air Source Heat Pump

28 Circuit 1

Circuit 2

Circuit 3

24 o

o

Tube surface temperature ( C)

4 C 20 15 s

16

T1 > T2 > T3

12 8 4

173 s T1 – T2 > T2 – T3

182 s 186 s

0 80

100

120

140

160

180

200

220

Time (s)

Fig. 5.8 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 1.

28 Circuit 1

Circuit 2

Circuit 3

24 o

o

Tube surface temperature ( C)

4 C 20

14 s

16 12

T1 = T2 = T3

T2 > T3 > T1

8

o

9.6 C

4

o

1.87 C 106 s

145 s

172 s

186 s

40 s

0 60

80

100

120 140 Time (s)

160

180

200

Fig. 5.9 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 2.

gravity, from 80 s to 186 s, the relationship of tube surface temperatures kept at T1 > T2 > T3 clearly. In addition, from 110 s to 150 s, the order of the curves was T1  T2 > T2  T3 because the mass of the melted frost flowing into Circuit 3 was much more than that into Circuit 2. In this section, this case was mainly used to be compared with Case 2.

Investigation of effect on uneven defrosting performance

129

28 Circuit 1

Circuit 2

Circuit 3

24 o

o

Tube surface temperature ( C)

4 C 20 16

Stage 1: No effect

12

T3 > T2 > T1

Stage 3: Positive effect T3 > T1 Stage 2: Negative effect

13 s

T 3 > T 1 > T2

T2 > T3 > T1

8

165 s o

4.2 C

4

204 s

40 s

115 s

199 s 118 s

0 80

100

120

196 s 140 160 Time (s)

180

200

220

Fig. 5.10 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 3.

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

4 C

20

o

Fin surface temperature ( C)

24

13 s

16

T1 > T2 > T3

12 8

195 s

4

190 s 185 s

0 80

100

120

140 160 Time (s)

180

200

220

Fig. 5.11 Measured fin surface temperatures at the center of the three refrigerant circuits during defrosting in Case 1.

As shown in Fig. 5.9, in Case 2, the tube surface temperatures of the three circuits left 0°C at 90 s, and started to rise steadily thereafter. From 80 s to 145 s, the order of the tube surface temperatures was T2 > T3 > T1. This possibly results from their uneven frosting accumulations and refrigerant distributions [32,33]. The difference

130

Defrosting for Air Source Heat Pump

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

4 C

20

o

Fin surface temperature ( C)

24

15 s 16 T2 > T 3; T 2 > T1

12

T1 > T3

T1 = T3

8

T1 = T2 = T3

o

8.1 C

188 s

4

173 s

135 s 40 s

0 60

80

100

141 s

120 140 Time (s)

160

180

200

Fig. 5.12 Measured fin surface temperatures at the center of the three refrigerant circuits during defrosting in Case 2.

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

20

Stage 1: No effect

o

Fin surface temperature ( C)

24

16

o

4C

Stage 2: Stage 3: Negative effect Positive effect

13 s

12 T1 = T2 = T3

8

T3 = T 1> T 2

T1 > T3 > T2

o

1.2 C 4

207 s

114 s

0

40 s 80

100

120

123 s

145 s

140 160 Time (s)

205 s 180

200

220

Fig. 5.13 Measured fin surface temperatures at the center of the three refrigerant circuits during defrosting in Case 3.

between the three tube surface temperatures was small, with the biggest difference of 1.87°C at 110 s. It is very obvious that the temperature curves steeply increased from 100 s to 130 s, which met Fig. 5.7 well. From 145 s to the end of defrosting, the relationship of tube surface temperatures kept at T1 ¼ T2 ¼ T3. All the temperature curves

Investigation of effect on uneven defrosting performance

131

for each circuit in Case 2 reached 24°C at the same time, 186 s into defrosting. That means the defrosting durations in Case 1 and Case 2 were the same. However, if their frost accumulations were the same, the defrosting duration in Case 2 may be shorter. Therefore, the defrosting performance would be better when the vertically installed multicircuit outdoor coil changed vertically installed. In addition, compared with the trends of the tube surface temperature for the outdoor coil vertically installed in Case 1, as shown in Fig. 5.8 [11], the coincidence curves show that the negative effects of melted frost downward flowing due to gravity could be eliminated after the outdoor coil was horizontally installed. As shown in Fig. 5.10, the tube surface temperatures of the three circuits left 0 °C at approximately 100 s, and orderly reached 24°C at 199, 204, and 196 s, respectively. Obviously, compared with Case 2, the defrosting duration was prolonged. To clearly describe the effects of wind blowing, the three stages were divided. From 80 to 118 s, named Stage 1 in Case 3, the temperature curves’ order kept at T3 > T2 > T1. At Stage 2, from 118 to 150 s, the wind blowing showed negative effects on defrosting. Particularly, the temperature of Circuit 3 became lower than that of Circuit 2, and their order changed to T2 > T3 > T1. The negative effects resulting from the heat transfer were enhanced between the hot tube and fins in each circuit and the cold ambient air (around 0.5°C). After 150 s, at Stage 3, the temperature of Circuit 3 returned to the highest one. The curves’ order became to T3 > T1, especially after 165 s, was at T3 > T1 > T2. T3 was the highest one, which may be because the retained water left on Circuit 3 was the least. And T2 was the smallest, which indicates that the wind quality at the middle circuit of the outdoor coil was the highest. At this stage, the temperature increased very quickly. This is because most melted frost was drained. It could be found that the turning point of Stage 2 and Stage 3 was at 150 s, a little earlier than the time the air fan turned off at 155 s. That means the positive effects of wind blowing were shown before the air fan turned off. Finally, compared with Case 2, it could be concluded that the operation of turning on the outdoor air fan and reversing the direction during defrosting could not decrease the energy waste fundamentally. As shown in Fig. 5.11, unlike the tube surface temperatures, the fin surface temperatures remained at 0°C at the first 110 s into defrosting. The rise in fin surface temperature was later than that in the tube surface temperature because the tube was in direct contact with the hot refrigerant; however, the fin indirectly contacted the refrigerant via the tube. In Case 1, it took 185, 190 , and 195 s for the fin surface temperatures to reach 24°C in the three circuits, respectively. During defrosting, the fin surface temperature curves’ order, the same as that of the tube surface temperature, kept at T1 > T2 > T3 clearly. As demonstrated in the previous studies [11, 14], this also resulted from the negative effects of downward-flowing melted frost due to gravity. Fig. 5.12 shows the measured fin surface temperatures at the center points of the three refrigerant circuits during defrosting in Case 2. From 60 to 135 s, T2 was always the highest in the three circuits, which was also because of their uneven frosting accumulations and refrigerant distributions [32,33]. At 95 s, the relationship of the temperature curves of Circuit 1 and Circuit 3 was changed from T1 ¼ T3 to T1 > T3. This was because the melted frost was downward flowing, and the mass of flowing melted frost

132

Defrosting for Air Source Heat Pump

on the surface of Circuit 3 was less than that on Circuit 1. After 135 s, the three curves were coincidental, which also demonstrated that the negative effects of downwardflowing melted frost were eliminated when the vertical multicircuit outdoor coil was horizontally installed. The fin surface temperatures all reached 24°C at 188 s, which was shorter than the duration in Case 1 by about 7 s, or 3.7% less. Therefore, it is further proved that the defrosting performance was better after the outdoor coil was horizontally installed. As shown in Fig. 5.13, the fin surface temperatures of the three circuits left 0°C at approximately 100 s, and orderly reached 24°C at 205, 207, and 205 s, respectively. Obviously, compared with Case 2, this duration was not shortened, but prolonged. Therefore, it further proved the negative effects of wind blowing the melted frost during defrosting for an ASHP unit with a horizontally installed multicircuit outdoor coil. The same as that shown in Fig. 5.10, three stages were also clearly divided in the fin surface temperature curves. Obviously, at Stage 2, the temperature rose slowly, although there was a lot of melted frost drained away by the wind blowing, as shown in Fig. 5.7. At Stage 3, especially at the later part of this stage, the temperature rose much more steeply than that in Case 2, due to less melted frost remaining on the surface of the fins. Therefore, to improve the defrosting performance, the melted frost remaining on the surface of the fins should be drained clearly.

5.2.3 Discussions For an ASHP unit with a multicircuit outdoor coil, to evaluate the energy consumption on heating ambient air due to waiting other circuit’s tube surface temperature reaching the preset defrosting termination temperature, defrosting evenness coefficient (DEC) was defined as the ratio of the minimum defrosting duration of circuit to the maximum one. Clearly, the higher the DEC, the more energy used for heating the ambient air due to waiting for the other circuit to terminate its defrosting could be saved. A DEC ¼
100% B A: The minimum defrosting duration for the tube or fin of the circuit. B: The maximum defrosting duration for the tube or fin of the circuit. Table 5.5 listed all the durations for tube and fin surface temperature of each circuit reaching 24°C, and the calculated DECs for the tubes and fins in the three cases. The calculation errors of the DEC were listed in Table 5.3. It could be found that the DECs for the circuits are 93.01% for Case 1, 100% for Case 2, and 96.08% for Case 3, respectively. And the DECs for the fins in the three cases are 94.87%, 99.47%, and 99.03%, orderly. Obviously, although the fin surface temperature was later, the DEC orders for the tube and fin were the same, at DEC2 > DEC3 > DEC1. In Case 2, the DECs were the highest, which means that there would be the least energy consumed on heating the ambient air, with the shortest defrosting duration. Meanwhile, due to the lowest DECs in Case 1, this part of the energy wasted on heating ambient air was the most. Consequently, to improve the defrosting performance, the DEC should

Investigation of effect on uneven defrosting performance

133

Table 5.5 Durations for the tubes and fins and the DECs in the three cases Item

Parameter

Case 1

Case 2

Case 3

Figs.

Unit

1

Duration for tube of Circuit 1 Duration for tube of Circuit 2 Duration for tube of Circuit 3 DEC for tube surface temperature Duration for fin of Circuit 1 Duration for fin of Circuit 2 Duration for fin of Circuit 3 DEC for fin surface temperature

173

186

199

5.8–5.10

s

182

186

204

5.8–5.10

s

186

186

196

5.8–5.10

s

93.01%

100%

96.08%

–

–

185

187

205

5.11–5.13

s

190

188

207

5.11–5.13

s

195

187

205

5.11–5.13

s

94.87%

99.47%

99.03%

–

–

2 3 4

5 6 7 8

be higher. Meanwhile, the DEC could be used as one index to evaluate the defrosting performance for an ASHP unit with a multicircuit outdoor coil. To clearly and comparatively study the effects of wind blowing on melted frost, Table 5.6 listed all the different durations and their differences in Case 2 and Case 3. The starts of all the tube surface temperatures leaving 0°C in the two cases were Table 5.6 Different durations and their differences in Case 2 and Case 3 Item

Parameter

Case 2

Case 3

Differencea

Figs.

Unit

1

Start of all the tube surface temperatures leaving 0°C Start of all the fin surface temperatures leaving 0°C Start of one tube surface temperature reaching 3°C Start of one fin surface temperature reaching 3°C Temperature of the tube surface increased during 40 s

90

90

0

5.9, 5.10

s

90

100

10

5.12, 5.13

s

106

115

9

5.9, 5.10

s

101

114

13

5.12, 5.13

s

9.6

4.2

5.4

5.9, 5.10

o

2

3

4

5

C

Continued

134

Defrosting for Air Source Heat Pump

Table 5.6 Continued Item

Parameter

Case 2

Case 3

Difference

Figs.

Unit

6

Temperature of the fin surface increased during 40 s Duration of the tube surface temperatures all increasing from 20°C to 24°C Duration of the fin surface temperatures all increasing from 20°C to 24°C Defrosting duration (tube surface temperatures all reaching 24°C) Duration of the fin surface temperatures all reaching 24°C

8.1

1.2

6.9

5.12, 5.13

o

14

13

1

5.9, 5.10

s

15

13

2

5.12, 5.13

s

186

204

18

5.9, 5.10

s

188

207

19

5.12, 5.13

s

7

8

9

10

C

The difference of values in Case 2 and Case 3 is calculated by: Difference ¼ Value (Case 2)  Value (Case 3).

a

the same at 90 s. However, the starts of all the fin surface temperatures leaving 0°C and the starts of one tube and fin surface temperature reaching 3°C were a little earlier in Case 2. This may be because the frost accumulation in Case 3 was 9 g more than that in Case 2. Compared with Case 3, the temperature of the tube and fin surface increases during the 40 s, when the outdoor air fan reversed to blowing, were obviously much higher. This further confirms the negative effects of wind blowing during defrosting. Compared with Case 3, the duration of the tube and fin surface temperatures increasing from 20°C to 24°C was much longer in Case 2. This resulted from less melted frost remaining on the surface in Case 3. Compared with Case 2, the duration of the tube and fin surface temperatures reaching 24°C was longer in Case 3. This further confirmed the negative effects of wind blowing melted frost during defrosting. As listed in Table 5.7, the energy supply, energy consumption, and defrosting efficiency in the three cases were also calculated [18,32]. In this experimental study, the total energy used for defrosting was 727.1 kJ in Case 1, but 697.9 kJ in Case 2, or 4.0% less. In Case 3, the total energy consumed was 812.0 kJ, or 16.3% more than that in Case 2. The defrosting efficiencies calculated for the three cases were calculated at 43.5%, 53.3%, and 47.8%, respectively. Comparing the defrosting efficiency in Case 1 with that in Case 2, it could be concluded that the defrosting performance would be better when the outdoor coil was horizontally installed. Also, it could be demonstrated that blowing the melted frost could not improve the defrosting performance, although the mass of retained water was effectively decreased. Therefore, to destroy the surface tension and thus improve defrosting performance, fin structure adjustment and fin surface treatment may be a direction for system optimization.

Investigation of effect on uneven defrosting performance

135

Table 5.7 Energy performance analysis for the three experimental cases Item

Parameter

Case 1

Case 2

Case 3

Unit

1 2 3 4 5 6 7

The power input to compressor The power input to indoor air fan The power input to outdoor air fan The energy from the indoor air Total energy supply during defrosting Energy consumption on melting frost Energy consumption on vaporizing the retained water Total energy consumption for defrosting Defrosting efficiency

111.0 6.3 0 609.9 727.1 307.6 55.0

112.6 1.7 0 538.6 697.9 323.3 48.9

140.3 1.6 7.2 663.0 812.0 313.0 66.0

kJ kJ kJ kJ kJ kJ kJ

362.6

372.2

379.0

kJ

43.5%

53.3%

46.7%

–

8 9

In this section, to decrease the flow path of downward-flowing melted frost due to gravity, the multicircuit outdoor coil was changed to be horizontally installed. In addition, the operation of reversing the outdoor air fan and blowing the remaining water away was undertaken to decrease the total mass of the remaining water. Finally, the following conclusions were reached: (1) When a vertically installed multicircuit outdoor coil in an ASHP unit was changed to horizontally installed, the defrosting efficiency increased from 43.5% to 53.3%, or at an increasing value of 9.8%. Meanwhile, the negative effects of melted frost downward flowing due to gravity were eliminated. The positive effects of a horizontally installed multicircuit outdoor coil on defrosting performance for an ASHP unit are obvious. (2) For an ASHP unit with a horizontal multicircuit outdoor coil, when the outdoor air fan was used to blow the melted frost during defrosting, the total mass of the retained water collected obviously decreased, from 566 g to 344 g, or 222 g less. However, the defrosting efficiency was not increased, but decreased from 53.3% to 46.7% due to the heat transfer being enhanced between the hot coil and the cold wind. Therefore, destroying the surface tension to enhance the melted frost’s local drainage as well as an outdoor coil structure adjustment and fin surface treatment may be a better choice. (3) To evaluate the energy consumption on heating the ambient air due to waiting for the other circuit to terminate its defrosting, the DEC was innovatively defined and first used here. The experimental results demonstrated that, to improve the defrosting performance, the DEC should also be improved. Therefore, there would be less energy consumed on heating the ambient air to wait for the other circuit’s defrosting termination. Furthermore, the DEC was suggested to be used as an index to evaluate the defrosting performance of an ASHP unit with a multicircuit outdoor coil. (4) During an operational cycle of frosting-defrosting for an ASHP unit, the frosting duration is always much longer than the defrosting duration. Consequently, frosting studies account for more important parts in studies of system performance improvement. Therefore, system frosting performance should also be studied for an ASHP unit with a horizontally installed outdoor coil before a traditional vertical outdoor coil is changed to being horizontally installed.

136

5.3

Defrosting for Air Source Heat Pump

Effect of surface tension on uneven defrosting

Among these exploratory experimental studies on improving system defrosting performance for ASHP units with vertical multicircuit outdoor coils, the phenomenon that different circuits terminate their defrosting processes at different times was found and reported. For example, Qu et al. [12] and O’Neal et al. [22] both investigated the transient defrosting performances of ASHP units, each with a vertical four-circuit outdoor coil. It was reported that when a defrosting process was terminated, the surface temperature at the exit of the lowest circuit was much lower than that at the exit of the highest circuit. In the study by Wang et al. [19], it was also shown that for a vertical seven-circuit outdoor coil, at 6 min into defrosting, the surfaces of the lower refrigerant circuits, which accounted for almost one-fourth of the entire coil surface area, were still covered by frost while those of the up-circuits were already free of frost. Thereafter, to qualitatively and quantitatively study the effects of the melted frost downward flowing over the surface of the outdoor coil due to gravity, a series of experimental and modeling studies were carried out in an ASHP unit having a vertical multicircuit outdoor coil [11, 18, 19]. Finally, it was demonstrated that downward flowing of the melted frost due to gravity from the up-circuits would adversely affect the defrosting performance of the down-circuits during an RCD operation, thus prolonging the defrosting operation and negatively impacting the indoor thermal comfort and energy efficiency of the ASHP unit [20]. While the outcomes from these studies [11, 18] demonstrated the effectiveness of locally draining away the melted frost from a vertical multicircuit outdoor coil with water-collecting trays installed between circuits, for an ASHP unit, however, there is always some melted frost remaining on the downside surface of each circuit due to surface tension. As shown in Fig. 5.14, for a fixed surface area multicircuit outdoor coil, the maximum flow path of melted frost on each circuit would decrease as its circuit number increases, from 2H in Fig. 5.14A, A two-circuit outdoor coil, to 4H/3 in Fig. 5.14B, A three-circuit outdoor coil, and then to only H in Fig. 5.14C, A four-circuit outdoor coil. The energy consumption on heating the cold melted frost downward Area of remained water H

h

A

2H

4H+h

2H

4H/3

Circuit 2

4H/3

A

(A)

2H

2A

(B)

4H/3

Circuit 2 A

H

A H

(C)

Circuit 2

A H

Circuit 3 A 3A

Circuit 1

A

Circuit 1 A

4H+2h

4H/3 Circuit 1

H

A

Circuit 3

4H+3h

Maximum flow path

Circuit 4 4A

Fig. 5.14 Maximum flow path of melted frost and total area of remaining water in multicircuit outdoor coil. (A) Two-circuit outdoor coil. (B) Three-circuit outdoor coil. (C) Four-circuit outdoor coil.

Investigation of effect on uneven defrosting performance

137

flowing along the flow path would be reduced, and then the defrosting performance was expected to be improved. However, at the same time, the total area of the circuit downside surface, or the total area of the remaining water, would increase exponentially, from 2A in Fig. 5.14A, A two-circuit outdoor coil, to 3A in Fig. 5.14B, A three-circuit outdoor coil, and then even to 4A in Fig. 5.14C, A four-circuit outdoor coil. The remaining water would consume energy [26] and thus adversely affect the system defrosting performance. Therefore, it is contradictory for the maximum flow path of the melted frost and the total area of the remaining water on improving system defrosting performance, when increasing circuit number of a fixed surface area evaporator. To solve this contradictory problem, the most effective method is to eliminate the surface tension on the melted frost, and thus decrease the total area of remaining water. Therefore, as a fundamental problem, the surface tension effects on melted frost, and thus on the defrosting performance for an ASHP unit having a multicircuit outdoor coil, should be quantitatively studied. However, as shown in Fig. 5.15, when an outdoor coil is vertically installed and without any separations between circuits, the area of remaining water is very small, just 44 mm
590 mm (Side B). To enlarge the area of remaining water, and thus clearly show the negative effects of surface tension on

m m

Side A (Topside) Side B (Back side)

44

590 mm

Circuit 1 Circuit 3

m

m

500 mm

Side C (Front side)

0

Circuit 2

50

Circuit 2

Circuit 1 Circuit 3 Side A 590 mm (Front side)

Side B (Downside)

(A)

Side C (Downside) 44 mm

(B)

m

Circuit 3

m m

Circuit 2

50 0

44

m

Side B 590 mm

Circuit 1

(C)

(D)

Side C

590 mm

Fig. 5.15 Area of remaining water for vertically and horizontally installed three-circuit outdoor coil. (A) Vertically installed three-circuit outdoor coil. (B) Horizontally installed three-circuit outdoor coil. (C) Area of remained water (Side B). (D) Area of remained water (Side C).

138

Defrosting for Air Source Heat Pump

defrosting performance, the outdoor coil could be changed to horizontally installed, from Fig. 5.14A to B. With this method, the total area of remaining water could be enlarged to 500 mm
590 mm (Side C), or 11.36 times the previous area, as shown in Fig. 5.15C–D. Therefore, to quantitatively study the negative effects of surface tension, in this paper, an experimental study on defrosting performance for an ASHP unit with a horizontal multicircuit outdoor coil was carried out. First, a detailed description of the specially built ASHP unit is presented. This is followed by reporting the experimental procedures and conditions. After the experimental cases and their results are detailed, a discussion and energy analysis are reported. Finally, a conclusion is given.

5.3.1 Experimental cases 5.3.1.1 Experimental setup, procedures, and conditions In this section, the experimental setup was totally the same as that previously used. So, it is briefly introduced here. The experimental ASHP unit was modified from a commercially available 6.5 kW heating-capacity variable speed ASHP unit. It was installed in an existing environmental chamber. In the chamber, there are a simulated indoor heated space and a simulated outdoor frosting space, with the same size of 3.8 m (L)
3.8 m (W)
2.8 m (H). The experimental ASHP unit was a split-type one consisting of a swing-type compressor, an accumulator, a four-way valve, an EEV, an indoor coil, and an outdoor coil. To control the indoor and outdoor spaces to meet the experimental conditions, a separate DX A/C system and two suits of sensible and latent LGUs were used in the environmental chamber. Finally, frosting environment in the outdoor space could be reached by running the experimental ASHP unit and LGUs together while an indoor heated environment by the experimental ASHP unit and the existing A/C system. Detailed information about the experimental setup, such as the measuring parameters, sensor locations, etc., can be found in the previous section. Moreover, the experimental procedures and conditions in this section are also totally the same as the previous section. To keep the compressor safe, it as always switched off before defrosting was started. About 1 min after the shutdown of the compressor, the four-way valve was switched to defrosting mode. After about 4 s, the compressor was then powered on again manually. Until now, a defrosting operation was started in the experiment. The defrosting operation was also manually terminated when the tube surface temperature at the exits of the three refrigerant circuits in the outdoor coil reached 24°C. Prior to defrosting, the experimental ASHP unit was operated in the heating (frosting) mode for 60 min, at an outdoor frosting ambient temperature of 0.5  0.2°C (dry-bulb temperature) and 90  3% relative humidity. During heating (frosting), the air temperature inside the heated indoor space was maintained at 20  0.5°C. To supply enough energy for defrosting and keep the ASHP unit working safely, the indoor air fan during defrosting remained operational at a lower speed. Fig. 5.16 shows the fluctuation of the measured face velocity of the outdoor coil during a heating (frosting) operation.

Investigation of effect on uneven defrosting performance

139

2.0 1.9 1.8 1.7

–1

Vair (m s )

1.6

The average value during a heat (froting) operation

1.5 1.4 1.3 1.2

–1

1.3 m s

1.1 1.0

0

400

800

1200 1600 2000 2400 2800 3200 3600 Time (s)

Fig. 5.16 Measured face velocity of outdoor coil during a heating (frosting) operation.

5.3.1.2 Experimental cases A series of experimental works using the experimental ASHP unit have been carried out to study the effects of surface tension on defrosting performance. In order to obtain meaningful experimental results, it was necessary to ensure that the frost that accumulated on the surface of the three circuits was even at first. For an ASHP unit with a multicircuit outdoor coil, it is hard to make the frost evenly accumulate on the surface of the outdoor coil, as many parameters affect frosting performance [31]. However, in this section, modulating valves installed at the inlet refrigerant pipe to each circuit (as shown in Fig. 5.2) may be deployed to vary the refrigerant flow to each circuit adjusted [32,33]. Therefore, to adjust the refrigerant flow into each circuit, a series of trial-and-error manual adjustments of the opening degrees of the stop valves was carried out. Then, a set of fixed valve opening degrees were obtained, and frost accumulation on the surface of the three circuits was close to each other [11]. Second, to make the comparative study results meaningful, the frost accumulations of different cases should be close to each other. To comparatively and quantitatively study the effects of surface tension for an ASHP unit having a multicircuit outdoor coil, two settings of experiments were finally designed and conducted. The mass transfer of the retained melted frost during defrosting in the two cases is shown in Fig. 5.17. In Case 1, the melted frost was melted and downward flowed freely. In Case 2, the melted frost was manually cleaned with a brush when the tube surface temperature at the exit of Circuit 1 reached 2°C, and the operation was sustained for approximately 30 s. This case stands for the conditions of fin surface treatment or structure optimizing, and thus the surface tension of the remaining melted frost was destroyed during defrosting. Table 5.8 shows the experimental conditions and relative results in the two cases.

140

Defrosting for Air Source Heat Pump

Fin

Fin

Vaporizing due to heating from fins

Vaporizing due to heating from fins

Droplets

Droplets

Clear the water manually

Freely flowing due to gravity

(A)

(B)

Fig. 5.17 Mass transfer of the retained melted frost during defrosting in the two cases. (A) Case 1. (B) Case 2.

Table 5.8 Experimental conditions and relative results in the two cases Item

Parameter

Case 1

Case 2

1

The surface tension of remaining water Defrosting duration Total mass of frost accumulated Total mass of melted frost collected Mass of the retained water collected Shown in

Kept

Destroyed

186 s 958 g

167 s 916 g

948 g

909 g

566 g

42 g

Figs. 5.17–5.20, 5.22; Tables 5.9 and 5.10

Figs. 5.17, 5.18, 5.21, 5.23; Tables 5.9 and 5.10

2 3 4 5 6

5.3.2 Results and analysis Four photographs illustrating the airside surface conditions of the outdoor coil at the start and end of defrosting in the two cases are shown in Fig. 5.18. As shown in Fig. 5.18A1 and B1, it is visually the same and even for the frost accumulated on the surface of the outdoor coil in the two cases, which met the requirements previously described in Section 5.3.1. As listed in Table 5.8, the frost accumulations were calculated at 958 g in Case 1 and 916 g in Case 2, respectively. Their difference was 42 g, or about 4.38%, which was small and acceptable in this section. Meanwhile, as predicted, a lot of residual water was retained over the outdoor coil downside surface when the defrosting operation terminated. As shown in Fig. 5.18A2 and B2, the melted frost retained on the downside surface over the horizontal multicircuit outdoor coil in an ASHP unit due to surface tension during RCD can be visually observed. The differences on the mass of retained melted frost in the two cases were shown in the whitedotted rectangles. It is obvious that the retained water in Case 1 was much more than

Investigation of effect on uneven defrosting performance

141

Fig. 5.18 Airside surface conditions of the outdoor coil at the start and end of defrosting in the two cases (four photographs).

that in Case 2. There is water remaining at the position where the arrow points in Fig. 5.18A2, and no water remained at the position where the arrow points in Fig. 5.18B2. It is obvious that the retained water in Case 1 was much more than that in Case 2, especially on Circuit 1. At the same time, Table 5.8 lists the mass of the retained water collected, at 566 g in Case 1 and 42 g in Case 2, respectively, in which the difference was obvious. Therefore, a similar mass of frost accumulation, frost evenly accumulated, and obvious differences on the mass of the remaining melted frost make this comparative study meaningful. Eight photographs illustrating the frost melting and downward-flowing process on the airside of the horizontal three-circuit outdoor coil in Case 1 are shown in Fig. 5.19. These photographs show the airside conditions of the outdoor coil up to 180 s into the defrosting operation when there was no solid frost left on the coil surface. It further took 6 s for all the tube surface temperatures at the exits of the three refrigerant circuits in the outdoor coil to reach 24°C, when the defrosting process was ended. From 0 to 40 s, the defrosting process was at the frost melting stage [25]. When it came to 60 s, the fin surface was directly contacting the ambient air; however, there was still no melted frost flowing away from the circuit. After 60 s into the defrosting operation, the melted frost started flowing from side A to side B (as shown in Fig. 5.19) along the downside surface of each circuit due to surface tension. As the defrosting process went by, the mass of melted frost increased with its accumulation. As illustrated in Figs. 5.4 and 5.17, when the gravity of the melted frost exceeded the maximum of the surface tension, the melted frost began downward flowing from the circuit to

142

Defrosting for Air Source Heat Pump

Fig. 5.19 Airside surface conditions of the outdoor coil during defrosting in Case 1 (eight photographs).

the water-collecting tray. Therefore, from 60 to 80 s, there were few melted frost drops downward flowing away from the circuit. In addition, as shown in Fig. 5.19, at the places indicated by the white arrows, a lot of melted frost kept downward flowing to the water-collecting tray for about 60 s, from 80 to 140 s. It is very obvious that at 100 s into defrosting, there was a lot of melted frost flowing away from the circuit. The melted frost downward flowed away from these circuits, reduced the energy consumption on vaporizing, and improved defrosting performance.

Investigation of effect on uneven defrosting performance

143

The measured operating performances of the experimental ASHP unit during defrosting, corresponding to the two experimental cases, are presented in Figs. 5.20–5.23. In all these figures, for their time (horizontal) axis, 60 s is the chosen starting time in order to clearly show the temperature rise during defrosting. Figs. 5.20 and 5.21 present the measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting. Figs. 5.22 and 5.23 show the measured fin

28 Circuit 1

Circuit 2

Circuit 3

24 o

o

Tube surface temperature ( C)

4 C 20 14 s 16

T2 > T3 > T1

12

T2 = T3 = T1

8 o

6.5 C 4

103 s

172 s

145 s

0

186 s

30 s 60

80

100

120

140

160

180

200

Time (s)

Fig. 5.20 Measured tube surface temperatures at the exit of refrigerant circuits in Case 1.

28 Circuit 1

Circuit 2

Circuit 3

24 o

o

Tube surface temperature ( C)

4 C 20 11 s

T2 > T3 ; T1 > T3

16

T1 = T2 = T3

T1 > T2 > T3

12 8 o

8.7 C 4

145 s

105 s

0

167 s

156 s

30 s 60

80

100

120

140

160

180

200

Time (s)

Fig. 5.21 Measured tube surface temperatures at the exit of refrigerant circuits in Case 2.

144

Defrosting for Air Source Heat Pump

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

4 C

20

o

Fin surface temperature ( C)

24

15 s

T2 > T 3; T 2 > T1

16

T1 = T3

12

T1 > T3

T1 = T2 = T3

8 o

5.5 C 95 s

4

135 s

0

173 s

188 s

30 s 60

80

100

120

140

160

180

200

Time (s)

Fig. 5.22 Measured fin surface temperatures at center of refrigerant circuits in Case 1. 28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

4 C

20

o

Fin surface temperature ( C)

24

11 s 16 T1 = T2 = T3

12

T1 > T2 > T3

T1 = T2 = T3

8 o

8.2 C 4

105 s

145 s

0

161 s

172 s

30 s 60

80

100

120

140

160

180

200

Time (s)

Fig. 5.23 Measured fin surface temperatures at center of refrigerant circuits in Case 2.

temperatures at the center points of the three circuits. It is noted that the variation trends of these temperatures are very similar to those reported by O’Neal [22] and Qu [12]. As shown in Fig. 5.20, all the curves for each circuit in Case 1 reached 24°C at the same time, at 186 s into defrosting. Compared with the trends of the tube surface temperature at the exit of each circuit for the vertically installed outdoor coil, as shown in

Investigation of effect on uneven defrosting performance

145

Fig. 3.13 in Section 3.3, the coincidence curves show that the negative effects of melted frost downward flowing due to gravity could be eliminated after the outdoor coil was horizontally installed. From 80 to 145 s, the tube surface temperature order for three circuits was kept at T2 > T3 > T1, which is mostly possible because of their uneven frost accumulations. The frost a that accumulated on the surface of Circuit 2 was the least, and then its tube surface temperature was obviously higher than the others in the three circuits. After the frost melted, from 145 s to the termination of defrosting, their curves kept at T1 ¼ T2 ¼ T3. In addition, it is very obvious that the temperature curves steeply increased from 100 s to 130 s, which met Fig. 5.19 well. Fig. 5.21 shows the measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 2. It is obvious that all the curves for each circuit reached 24°C at the same time, at 167 s into defrosting. Compared with the defrosting duration in Case 1, the defrosting duration in Case 2 was decreased a lot, at a reduction of 19 s, or about 10.2% less. Therefore, the decrease in the negative effects of surface tension by the melted frost drained away was proved. From 80 to 145 s, the same as that in Case 1, the tube surface temperature in Circuit 3 was kept the lowest. In addition, from 145 s to the termination of defrosting, the tube surface temperature curves of three circuits kept at T1 ¼ T2 ¼ T3. However, from 120 to 145 s, their temperature order kept at T1 > T2 > T3. This is because the melted frost was drained away at the order of Circuit 1–3. The negative effects of surface tension on system defrosting performance were further proved. The measured fin surface temperatures at the center points of the three refrigerant circuits during defrosting in Case 1 are shown in Fig. 5.22. The length of time it took for the fin surface temperatures in the three circuits to all reach 24°C was 188 s, which was about 2 s later than the tube surface temperature curves. This may be because of the delay of the heat transfer from the tube to the fin. The fin surface temperature curves were also kept at coincidence from 135 s into defrosting to the termination. From 60 to 135 s, the fin surface temperature at Circuit 2 was kept the highest. The same as its tube surface temperature, this is because the frost accumulated on this circuit was less than that on the surface of the other two circuits. However, the fin surface temperature was affected a lot by the melted frost, such as the fin surface temperature of Circuit 1 increasing from 95 to 100 s suddenly after the melted frost drained away, and then the temperature order kept at T1 > T3. Fig. 5.23 shows the measured fin surface temperatures at the center points of the three refrigerant circuits during defrosting in Case 2. The fin surface temperatures all reached 24°C at 172 s, which was shorter than the duration in Case 1 by about 16 s, or 8.5% less. Therefore, it is further proved that the negative effects of surface tension could be decreased by the melted frost being manually drained away. In this figure, only from 100 to 140 s was the curves’ order not kept at T1 ¼ T2 ¼ T3, but at T1 > T2 > T3 clearly. The reason is that the melted frost was drained away at the order of Circuit 1–3, too. Different durations and their differences in the two cases are summarized in Table 5.9. As shown in Figs. 5.20–5.23, in the two cases, the starts of the tube surface temperature leaving 0°C were both at 80 s. The starts of the tube surface temperatures reaching 2 °C were at 103 and 105 s, respectively. It could be found that the starts for

146

Defrosting for Air Source Heat Pump

Table 5.9 Different durations and their differences in the two cases Item

Parameter

Case 1

Case 2

Difference

Unit

1

Start of tube surface temperature leaving 0°C Start of fin surface temperature leaving 0°C Start of tube surface temperature reaching 2°C Start of fin surface temperature reaching 2°C Temperature of tube surface increased during 30 s Temperature of fin surface increased during 30 s Duration of tube surface temperatures all increased from 20°C to 24°C Duration of fin surface temperatures all increased from 20°C to 24°C Defrosting duration (tube surface temperatures all reached 24°C) Duration of fin surface temperatures all reaching 24°C

80

80

0

s

60

90

30

s

103

105

2

s

95

105

10

s

6.5

8.7

2.2

°C

5.5

8.2

2.7

°C

14

11

3

s

15

11

4

s

186

167

19

s

188

172

16

s

2 3 4 5 6 7

8

9

10

the two cases are nearly the same, which met the similar total mass of melted frost collected listed in Table 5.8 and even frost accumulations shown in Fig. 5.18A1 and B1. Only there is difference on fin surface temperature, with the starts of leaving 0°C at 60 and 90 s, and the starts of reached 2°C at 95 and 105 s, respectively. The difference on fin surface temperature results from the effect of the melted frost during defrosting. Therefore, this comparative study is meaningful. In addition, compared with the increased temperatures of the tube and fin surfaces during 30 s in Case 1, at the operation period of the melted frost being manually drained away in Case 1, those values in Case 2 always were much higher. This shows the negative effects of retained melted frost due to surface tension. After cleaning the melted frost, the durations of the tube and fin surface temperatures that increased from 20°C to 24°C in Case 1 were all less than those in Case 2. This phenomenon results from the different masses of retained melted frost on the surface of the outdoor coil, which also met the mass of collected retained melted frost listed in Table 5.9. Finally, the length of time it took the tube and fin surface temperatures to reach 24°C in Case 1 was less than that in Case 2. Therefore, the negative effects of surface tension were further demonstrated.

Investigation of effect on uneven defrosting performance

147

5.3.3 Discussions and energy analysis Due to the surface tension, there would be melted frost retained on the downside surface of the outdoor coil. In addition, the mass of the remaining melted frost is positively proportional with the area of remaining water, as shown in Fig. 5.14. However, for a fixed surface area multicircuit outdoor coil, adjusting its placement or structure may be an optimizing direction to decrease its area of remaining water, as shown in Fig. 5.24. Moreover, to destroy the surface tension like that in Case 2, a fin structure adjustment or surface treatment could also optimize the system defrosting performance, as shown in Fig. 5.25. When the downside of the outdoor coil was one angle, as shown in Figs. 5.25B–C, two angles as shown in Fig. 5.25D, and several angles as shown in Fig. 5.25E, the surface tension could be destroyed easily. Finally, the defrosting efficiency could be improved moreso than that in Case 2 if the melted frost was totally drained away by any method. As shown in Table 5.10, the energy supply and consumption during RCD in two cases were also calculated, with their calculation errors listed in Table 5.10. In this experimental study, the total energy used for defrosting was 697.9 kJ in Case 1, but 526.0 kJ in Case 2, or 24.6% less. Clearly, most energy came from the indoor air, with 583.6 kJ in Case 1 and 425.9 kJ in Case 2, respectively. Total energy consumption for defrosting for the two cases was 344.4 and 323.0 kJ, with their defrosting efficiencies calculated at 49.4% and 61.4%, respectively. Therefore, the negative

Fin

Fin

q

q

(A)

(B)

(C)

(D)

Fig. 5.24 Placement adjustment for an outdoor coil.

(A)

(B)

q

(C)

q

Fig. 5.25 Fin structure adjustment for an outdoor coil.

q

(D)

q

q

(E)

q

q

q

148

Defrosting for Air Source Heat Pump

Table 5.10 Energy and defrosting efficiency calculations in the two cases Item

Parameter

Case 1

Case 2

Unit

1 2 3 4 5 6 7

The power input to compressor fan The power input to indoor air fan The power input to outdoor air fan The energy from the indoor air Total energy supply during defrosting Energy consumption on melting frost Energy consumption on vaporizing the retained water Total energy consumption for defrosting Defrosting efficiency

112.6 1.7 0 583.6 697.9 320.0 24.4

98.8 1.3 0 425.9 526.0 305.9 17.1

kJ kJ kJ kJ kJ kJ kJ

344.4 49.4%

323.0 61.4%

kJ –

8 9

effects of surface tension on defrosting performance were further demonstrated, and the optimization by structure adjustment or fin surface treatment to eliminate the surface tension of the melted frost could decrease the energy waste fundamentally. For an ASHP unit with a vertical multicircuit outdoor coil, the RCD efficiency could be improved by installing water-collecting trays between circuits. However, there was still some melted frost on the downside surface of each circuit due to surface tension. That melted frost consumed energy during defrosting, and thus had negative effects on the system defrosting performance. Therefore, to quantitatively study the negative effects due to surface tension, the effects are enlarged as enlarging the area of the downside surface of the outdoor coil, by changing the vertical three-circuit outdoor coil into horizontally installed. Finally, a comparative experimental study with the surface tension kept and destroyed was undertaken and the relative results analyzed. The following conclusions could be reached from this section. (1) For a horizontal multicircuit outdoor coil, it is proved that the defrosting process of each circuit reached the preset termination temperature (24°C in this section) at the same time. The condition that a circuit terminated its defrosting and was waiting for the others could be avoided. The energy consumption on heating the cold ambient air due to uneven defrosting could be saved [21]. Therefore, the same as installing watercollecting trays between circuits, the negative effects of downward-flowing melted frost due to gravity on defrosting performance for an ASHP unit with a vertical multicircuit outdoor coil could also be eliminated by the coil being horizontally installed, as suggested in Reference [13]. (2) Compared with the melted frost remaining on the downside surface of each circuit due to surface tension in Case 1, the defrosting duration was shortened from 186 to 167 s, or 19 s less, and the length of time it took the fin surface temperatures to reach 24°C was shortened from 188 to 172s, or 16 s less, when the melted frost was cleaned off due to the surface tension being manually destroyed in Case 2. The defrosting performances are obviously improved by decreasing

Investigation of effect on uneven defrosting performance

149

their durations at the ratio of 10.2% and 8.6%, respectively. (3) Experimental results show that the total energy used for defrosting was 697.9 kJ in Case 1, but 526.0 kJ in Case 2, or 24.6% less. Total energy consumption for defrosting for the two cases was 344.4 kJ and 323.0 kJ, respectively, or a 6.2% difference. Most energy came from the indoor air, and was used on melting frost. Their defrosting efficiencies were calculated at 49.4% and 61.4%, with about 12.0% higher in Case 2. (4) As shown in Fig. 5.15A, for a vertical three-circuit outdoor coil with separations to install water-collecting trays, the total area of the downside surface of the outdoor coil is three times the area of Side B, or about 0.264 times the area of Side C shown in Fig. 5.15D. It means that the defrosting efficiency could be improved by about 0.264
12.0%, or 3.2%, when the remaining melted frost at the downside of each circuit was drained away during RCD. (5) For an ASHP unit with a vertical three-circuit outdoor coil, installing watercollecting trays between circuits could improve the defrosting efficiency from 43.5% to 56.7%, or about 13.2% [11]. Consequently, for an ASHP unit with a vertically installed three-circuit outdoor coil, the defrosting efficiency is supposed to be improved about 16.4% after downward-flowing melted frost is locally drained by water-collecting trays installed between circuits and the remaining water is cleaned off by destroying the surface tension.

5.4

Concluding remarks

In this chapter, the following conclusions could be reached. (1) The uneven defrosting phenomenon could be eliminated by horizontally installing a multicircuit outdoor coil. After the uneven defrosting was avoided, the defrosting performance of the ASHP unit would be improved with less energy consumption and a shortened defrosting duration. (2) It was experimentally demonstrated that the residual water retained on the downside of the circuit due to surface tension would have negative effects on defrosting performance for an ASHP unit. After the residual water was wiped off manually, the optimization of the defrosting performance was quantitatively investigated. (3) After installing horizontally a vertically installed multicircuit outdoor coil, the uneven defrosting problem was alleviated. However, before we consider this modification, the frosting or heating performance of the ASHP unit should be further tested. This is because the defrosting period accounts for a smaller proportion of time in a frosting-defrosting cycle. (4) As mentioned in Chapter 2, after the circuit number was increased from a two circuit to a three circuit for a multicircuit outdoor coil in an ASHP unit, the negative effects of the melted frost were increased due to a larger increase of defrosting efficiency after the water-collecting trays were installed. However, the number of circuits is always limited by the dimension of the structure of an outdoor coil. Here, the negative effects of the residual water retained on the downside of the circuit due to surface tension also limits the number of circuits for an outdoor coil having a fixed heat exchanger area.

150

Defrosting for Air Source Heat Pump

References [1] Wang F, Liang CH, Yang MT, Zhang XS. Preliminary study of a novel defrosting method for air source heat pumps based on superhydrophobic fin. Appl Therm Eng 2015;90: 136–44. [2] Ye XM, Xia XH, Zhang JF. Optimal sampling plan for clean development mechanism energy efficiency lighting projects. Appl Energy 2013;112:1006–15. [3] Ni L, Dong JK, Yao Y, Shen C, Qv DH, Zhang XD. A review of heat pump systems for heating and cooling of buildings in China in the last decade. Renew Energy 2015;84: 30–45. [4] Payne V, O’Neal DL. Defrost cycle performance for an air-source heat pump with a scroll and a reciprocating compressor. Int J Refrig 1995;18(2):107–12. [5] Song MJ, Deng SM, Xia L. A semi-empirical modeling study on the defrosting performance for an air source heat pump unit with local drainage of melted frost from its three-circuit outdoor coil. Appl Energy 2014;136:537–47. [6] Wang ZY, Wang XX, Dong ZM. Defrost improvement by heat pump refrigerant charge compensating. Appl Energy 2008;85:1050–9. [7] Song MJ, Pan DM, Li N, Deng SM. An experimental study on the negative effects of downwards flow of the melted frost over a multi-circuit outdoor coil in an air source heat pump during reverse cycle defrosting. Appl Energy 2015;138:598–604. [8] Song MJ, Deng SM, Pan DM, Mao N. An experimental study on the effects of downwards flowing of melted frost over a vertical multi-circuit outdoor coil in an air source heat pump on defrosting performance during reverse cycle defrosting. Appl Therm Eng 2014;67: 258–65. [9] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil in an air source heat pump-Part I: experiments. Appl Energy 2012;91:122–9. [10] O’Neal DL, Peterson KT, Anand NK, Schliesing JS. Refrigeration system dynamics during the reversing cycle defrost. ASHRAE Trans 1989;95(2):689–98. [11] Ciriello V, Bottarelli M, Di Federico V. Uncertainty-based analysis of variations in subsurface thermal field due to horizontal flat-panel heat exchangers. Proc Environ Sci 2015;25:50–7. [12] Qi GP, Jiang F. Numerical investigation on prevention of fouling in the horizontal tube heat exchanger: particle distribution and pressure drop. Desalination 2015;367:112–25. [13] Dong JK, Deng SM, Jiang YQ, Xia L, Yao Y. An experimental study on defrosting heat supplies and energy consumptions during a reverse cycle defrost operation for an air source heat pump. Appl Therm Eng 2012;37:380–7. [14] Shen C, Jiang YQ, Yao Y, Deng SM. Experimental performance evaluation of a novel dryexpansion evaporator with defouling function in a wastewater source heat pump. Appl Energy 2012;95:202–9. [15] Adamson AW. Physical chemistry of surfaces. New York, NY: John Wiley; 1982. [16] Li LT, Wang W, Sun YY, Feng YC, Lu WP, Zhu JH, Ge YJ. Investigation of defrosting water retention on the surface of evaporator impacting the performance of air source heat pump during periodic frosting-defrosting cycles. Appl Energy 2014;135:98–107. [17] Kannadasan N, Ramanathan K, Suresh S. Comparison of heat transfer and pressure drop in horizontal and vertical helically coiled heat exchanger with CuO/water based nano fluids. Exp Thermal Fluid Sci 2012;42:64–70. [18] Yoon S, Lee SR, Go GH. Evaluation of thermal efficiency in different types of horizontal ground heat exchangers. Energy Buildings 2015;105:100–5.

Investigation of effect on uneven defrosting performance

151

[19] Simms RB, Haslam SR, Craig JR. Impact of soil heterogeneity on the functioning of horizontal ground heat exchangers. Geothermics 2014;50:35–43. [20] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse-cycle defrosting method for air source heat pumps. Build Environ 2010;45(11):2354–61. [21] Song MJ, Deng SM, Mao N, Ye XM. An experimental study on defrosting performance for an air source heat pump unit with a horizontally installed multi-circuit outdoor coil. Appl Energy 2016;165:371–82. [22] Hambraeus K. Heat transfer of oil-contaminated HFC134a in a horizontal evaporator. Int J Refrig 1995;18(2):87–99. [23] Liang CH, Wang F, Lv Y, Wu CX, Zhang XS, Zhang YF. Experimental study of the effects of fin surface characteristics on defrosting behavior. Appl Therm Eng 2015;75:86–92. [24] Yao Y, Jiang YQ, Deng SM, Ma ZL. A study on the performance of the airside heat exchanger under frosting in an air source heat pump water heater/chiller unit. Int J Heat Mass Transf 2004;47:17–8. [25] Song MJ, Wang ZH, Mao N, Li Z, Chen Y. An experimental study on the uneven refrigerant distribution over a vertically installed multi-circuit outdoor coil in an air source heat pump unit during reverse cycle defrosting. Appl Therm Eng 2015;91:975–85. [26] Kim JH, Braun JE, Groll EA. A hybrid method for refrigerant flow balancing in multicircuit evaporators: upstream versus downstream flow control. Int J Refrig 2009;32(6): 1271–82.

Frosting evenness coefficient 6.1

6

Introduction

For an outdoor coil used in an ASHP unit, on its refrigerant side, multiple parallel circuits are commonly used for minimized refrigerant pressure loss and enhanced heat transfer efficiency. For an ASHP unit having a multicircuit outdoor coil, uneven defrosting was reported in limited previous chapters. This phenomenon reflects that different circuits’ defrosting processes were terminated (refrigerant temperature at exit of circuit reaching 24°C) at different times. Uneven defrosting would result in the delay of a defrosting process, and thus prolong its duration, with more energy consumption for heating the top circuits and surrounding air. Thus, the defrosting performance would be adversely affected due to uneven defrosting. It has been demonstrated that the melted frost flowing downward is one of reasons for the uneven defrosting. When the negative effects of downward-flowing melted frost were quantitatively investigated, the frost on the surface of each circuit was always adjusted to be nearly evenly distributed for a multicircuit outdoor coil, with the frost mass difference between any two circuits smaller than 10%. Also, in the modeling studies, the frost accumulation was also assumed to be totally evenly distributed before defrosting. However, in practice, it is impossible for the frost to be evenly distributed on the surface of a multicircuit outdoor coil due to the uneven distributions of refrigerant inside the tube and the inlet air outside the tube. That means the uneven frosting state should also be considered when the optimization of ASHP units was carried out. When the frost is unevenly distributed on the surface of the outdoor coil, there are three typical conditions. The first one is that the frost density is not even on the whole surface of a coil. The second one is that the frost accumulation on the windward and leeward sides of the outdoor coil is the same, and the last one is that the frost accumulation on the surface of each circuit is not equal. For the first one, it is easy to understand that the inlet and outlet refrigerant temperature of a coil is different, and thus the frost density at the inlet section is higher. For the second, it is hard to measure or calculate the frost mass on two sides of the outdoor coil. Also, it is clear that more frost would be formed on the windward side due to the higher relative humidity than on the leeward side. Therefore, in this chapter, only the third type of uneven frosting is considered. To further clearly describe this type of uneven defrosting, frosting evenness coefficient (FEC) was defined as the ratio of the minimum frost accumulation among three circuits to the maximum one, and could be calculated by the melted frost collected from each water collecting cylinder, with the water vaporized into the ambient air neglected. For example, as shown in Fig. 6.1, the masses of frost accumulation on the three circuits’ surface are FA1, FA2, and FA3, respectively. When FA1 ¼ FA2 ¼ FA3, the FEC is 100%, and thus this is even frosting. If no Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00006-0 © 2019 Elsevier Ltd. All rights reserved.

154

Defrosting for Air Source Heat Pump

FA1

Circuit 1

FA2

Circuit 2

FA3

Circuit 3

Fig. 6.1 Frost accumulated on the surface of three circuits.

FA1 ¼ FA2 ¼ FA3, the FEC is not 100%, and that is uneven frosting. When is FA1 > FA2 > FA3, FEC is the ratio of FA3 to FA1, which is less than 100%. It is hard to avoid uneven frosting in practice, which always influences the operating performance of an ASHP unit at the heating or frosting mode. To evaluate the influence, COP can be used as a performance indicator. In this chapter, uneven frosting performance will be comprehensively investigated using an experimental approach. Moreover, when defrosting is started following an uneven frosting operation, the frosting condition would influence the defrosting performance. Therefore, the defrosting performance with frost unevenly distributed on the outdoor coil surface will be examined in this chapter. As mentioned in the previous chapter, the downwardflowing melted frost due to gravity and the residual water due to surface tension both have negative effects on defrosting performance. When frost is unevenly formed on the surface of a multicircuit outdoor coil, it would be different for with and without melted frost locally drained away using water collecting trays. Therefore, the influences of uneven frosting on the defrosting performance of an ASHP unit with the melted frost locally drained will also be examined in this chapter.

6.2

Even frosting performance for a multicircuit outdoor coil

Uneven defrosting might result from an uneven frosting start, which means the frost accumulations on each circuit surface of a multicircuit outdoor coil in an ASHP were different. Wang et al. reported an experimental study on the performance of an ASHP unit for a kind of mal-defrost phenomenon appearing in moderate climate conditions [1]. As shown in Fig. 6.2A, the frost accumulated on the downside is much more than that on the upside of the outdoor coil. A similar phenomenon can also be found in Figs. 6.2B and C [2]. Unfortunately, in the existing literature, no study on even frosting performance and corresponding control method for an ASHP unit with a multicircuit outdoor coil was found. For an ASHP unit with a multicircuit outdoor coil, not only might uneven frosting be the reason for the uneven defrosting phenomenon, but it could also affect the heat transfer between the outdoor air and the refrigerant, and thus the frosting COP of the whole system. Therefore, in this section, a comparative and experimental study on

Frosting evenness coefficient

155

(A)

11.17 08:00

11.18 11:00

11.19 08:00

11.19 17:00

11.20 07:00

11.20 11:00

11.21 08:00

11.22 09:00

11.23 09:00

11.23 12:40

(B)

(C)

frosted Fig. 6.2 Airside surface conditions of the outdoor coil at the start of defrosting.

system performance when frost accumulates on a three-circuit outdoor coil’s surface in an ASHP unit at different FECs has been carried out. Also, a quantitative analysis has been conducted using the experimental data. In this section, a detailed description of an experimental ASHP unit and the experimental conditions are first reported. Thereafter, different experiment cases were designed and carried out, with their relative results followed. After the result analyses, the conclusions are finally given.

6.2.1 Experimental cases 6.2.1.1 Experimental setup In this section, the experimental setup was totally the same as that previously used in Chapter 3. So, it will be briefly introduced here. The experimental ASHP unit was modified from a commercially available 6.5 kW heating-capacity variable-speed

156

Defrosting for Air Source Heat Pump

ASHP unit. It was installed in an existing environmental chamber. In the chamber, there was a simulated indoor heated space and a simulated outdoor frosting space, with the same size of 3.8 m (L)  3.8 m (W)  2.8 m (H). The experimental ASHP unit was a split-type one consisting of a swing-type compressor, an accumulator, a four-way valve, an EEV, an indoor coil, and an outdoor coil. To control the indoor and outdoor spaces to meet the experimental conditions, a separate DX A/C system and two suits of sensible and latent LGUs were used in the environmental chamber. Finally, a frosting environment in the outdoor space could be reached by running the experimental ASHP unit and LGUs together while an indoor heated environment by the experimental ASHP unit and the existing A/C system. Detailed information about the experimental setup, such as the measuring parameters, sensor locations, etc., can be found in the previous section.

6.2.1.2 Control method of even frosting In order to obtain meaningful experimental results, it was necessary to ensure that the evenness of frost accumulations on the surface of three circuits in different experimental cases was different. In this section, all the frost accumulations on the surface of each circuit could be calculated, with the melted frost collected in the water-collecting cylinder and the residual water collected on the surface of the fins and tubes in consideration while the vaporized water during defrosting was neglected. The outcomes from previous studies [3,4] demonstrated that frost accumulation on the outdoor coil surface is decided by the distribution of inlet air passing by each circuit, the distribution of refrigerant flowing into each circuit, and the structure of an outdoor coil. Distributions of inlet air and refrigerant are affected by temperature, humidity, the tube’s inside frictional resistance, etc. Also, the structure of each circuit could not be made the same. Therefore, for an ASHP unit with a multicircuit outdoor coil, it is hardly possible to make the frost evenly accumulated on each circuit. However, it is still possible to vary the cold input to each circuit through varying the refrigerant supply to each circuit. This is because uneven frosting was fundamentally caused by a different heat transfer between the cold refrigerant and the ambient air, when the supplies of the refrigerant to each circuit were the same. Consequently, if the cold supplied to each circuit can be varied according to the actual frosting thermal load that each circuit should support, then the problem of uneven frosting may be alleviated. Modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to vary the refrigerant flow, thus adjusting the cold input to each circuit. In this section, only the refrigerant flow rate of each circuit was controlled, oriented by the tube surface temperature at the exit of each circuit. Experimental work was then carried out at two experimental cases, so that the system frosting performances at different FECs could be comparatively and quantitatively analyzed. In Case 1, all the stop valves on three circuits were fully open; therefore, the opening degrees of the stop valves were kept constant. However, in Case 2, the opening degrees of the valves were not constant. At the start of the frosting experiment, a suite of suitable opening degrees was obtained and fixed for three circuits after a series of trial-and-error

Frosting evenness coefficient

157

manual adjustments oriented by the tube surface temperature at the exit of each circuit. Thereafter, the valves were still randomly adjusted to make the tube surface temperature at the exit of each circuit the same during frosting. Finally, a higher FEC could be reached, and thus this comparative study on frosting performance at different FECs could be conducted. Figs. 6.3 and 6.4 show the measured tube surface temperatures at the exits of the three refrigerant circuits on heating mode before frosting growth. As shown in Fig. 6.3, the temperatures of Circuit 1 and Circuit 2 are the same at 9.2°C at 2800 s, which is the start point of 1 h frosting. The temperature of Circuit 3 is obviously much higher than the temperatures of the other two circuits, more than 6°C at about 1.8°C. This may be result from that refrigerant distribution into Circuit 3 is the fewest among the three circuits. However, in Case 2 as shown in Fig. 6.4, the temperatures of Circuit 1 and Circuit 2 are nearly the same at about 7.6°C, with the temperature of Circuit 2 about 0.2°C higher. The temperature of Circuit 3 is about 4°C at 2800 s. It could be found out that the temperature difference between Circuits 1 and 3 in Case 2 is just about 3.5°C, which is much smaller than that in Case 1, more than 6°C. Moreover, the sum value of the temperatures of the three circuits at 2800 s in Case 2 is 19.2°C and in Case 1 –20.2°C. The small temperature difference, about 1.0°C, demonstrates that the cold loads for the two cases are the same during experiments.

Fig. 6.3 Measured tube surface temperatures at the exits of the three refrigerant circuits on heating mode before frosting growth in Case 1.

158

Defrosting for Air Source Heat Pump

Fig. 6.4 Measured tube surface temperatures at the exits of the three refrigerant circuits on heating mode before frosting growth in Case 2.

6.2.2 Results and analysis Fig. 6.5 presents 12 photographs showing the airside surface conditions of the outdoor coil during frosting in the two cases. As observed from Fig. 6.5A1 and B1, the surface conditions at the start of defrosting for each circuit in the two cases were virtually the same, except that there was less frost on the upside of Circuit 3 in Case 1 than the downside, which agreed well with the trends of measured tube surface temperatures on heating mode before frosting growth, as shown in Figs. 6.3 and 6.4. From Fig. 6.5A1–A6, it can be seen that the frost accumulated on the airside of the outdoor coil increased with time. However, the frost accumulations were not even for the three circuits. It is obvious that the frost accumulated on Circuit 3 keeps fewer than that on the other two circuits. As seen from Fig. 6.5B1–B6,the frost accumulated evenly on the three circuits with time. As shown in Table 6.1, the results of two experimental cases are listed. The total mass of frost accumulated is 1000 g in Case 1 and 1001 g in Case 2, which are nearly the same. In Case 1, the masses of frost accumulated on each circuit’s surface are 355 g for Circuit 1, 367 g for Circuit 2, and 278 g for Circuit 3, respectively. The mass order of frost accumulation is Circuit 1  Circuit 2 > Circuit 3, which well agrees with the temperature order of TCircuit 1  TCircuit 2 < TCircuit 3, as shown in Fig. 6.3. This is because the opening degrees of the stop valves are kept constant during frosting in Case 1. The FEC is calculated at about 75.7%. However, in Case 2, the masses of frost accumulated on each circuit’s surface are orderly at 346 g, 313 g, and 342 g,

Frosting evenness coefficient

159

Fig. 6.5 Airside surface conditions of the outdoor coil during frosting in two cases. Table 6.1 Results of two experimental cases Item

Parameter

Case 1

Case 2

1 2 3 4 5 6

Frost accumulation on Circuit 1 Frost accumulation on Circuit 2 Frost accumulation on Circuit 3 Total mass of melted frost FEC Results shown in

355 g 367 g 278 g 1000 g 75.7% Figs. 6.5, 6.6, 6.8–6.13

342 g 313 g 346 g 1001 g 90.5% Figs. 6.5, 6.7–6.13

respectively, from Circuit 1 to 3. The mass order of frost accumulation is Circuit 1  Circuit 2  Circuit 3, which does not agree with the temperature order of TCircuit 1  TCircuit 2 < TCircuit 3, as shown in Fig. 6.4. This is because the opening degrees of the stop valves are not constant during frosting in Case 2, but randomly adjusted to make the tube surface temperature at the exit of each circuit the same. Therefore, although the temperature of Circuit 3 in Fig. 9 is the highest from 300 s to 2800 s on heating mode before frosting growth, it is no matter with the frost accumulation at the end of the frosting process. Finally, the FEC is calculated at about 90.5%, which is about 14.8% higher than that in Case 1. Moreover, the measured operating performances of the experimental ASHP unit during frosting, corresponding to the two experimental cases, are presented in Figs. 6.6–6.12 . Figs. 6.6 and 6.7 present the measured tube surface temperatures at the exits of the three refrigerant circuits. Figs. 6.8 and 6.9 present the Gauss fit of the measured refrigerant volume flow rate and the measured refrigerant pressure drop across the outdoor coil in the two cases. Figs. 6.10–6.12 show the measured air temperature difference between the indoor coil inlet and the outlet, the tube surface

160

Defrosting for Air Source Heat Pump

4 Circuit 2

Circuit 3

o

Tube surface temperature ( C)

Circuit 1 0 –4 –8 –12 –16 –20 0

500

1000

1500 2000 Time (s)

2500

3000

3500

4000

Fig. 6.6 Measured tube surface temperatures during frosting in Case 1.

4 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature ( C)

0 –4 –8 –12 –16 –20 0

500

1000

1500 2000 Time (s)

2500

3000

3500

4000

Fig. 6.7 Measured tube surface temperatures during frosting in Case 2.

temperatures at the exit of the outdoor coil, and the average fin surface temperatures at the center of the three circuits in the two cases, respectively. Finally, the variation of the calculated COP of the ASHP unit during frosting in the two cases is shown in Fig. 6.13. As shown in Fig. 6.6, the measured tube surface temperature of Circuit 1 and Circuit 2 was kept the same in Case 1. This is the reason why the frost that accumulated on

Frosting evenness coefficient

161

Measured refrigerant volume flow rate (L/M)

1.15 Case 1 Case 2

Gauss Fit of Case 1 Gauss Fit of Case 2

1.10

1.05

1.00

0.95

0.90 0

500

1000

1500 2000 Time (s)

2500

3000

3500

4000

Fig. 6.8 Measured refrigerant volume flow rate during frosting in two cases.

0.40 Case 1 Case 2

Gauss Fit of Case 1 Gauss Fit of Case 2

Pressure drop (Bar)

0.35 0.30 0.25 0.20 0.15 0.10 0

500

1000

1500

2000

2500

3000

3500

4000

Time (s)

Fig. 6.9 Measured refrigerant pressure drop across the outdoor coil during frosting.

the two circuits was nearly the same, as listed in Table 6.1. However, the temperature of Circuit 3 did not decrease, but increased as frosting process. And the temperature difference between Circuit 3 and the other two circuits increased. Therefore, the frost accumulated on Circuit 3 was very little, which agreed well with Fig. 6.3. In Fig. 6.7, the temperature of the three circuits was kept nearly the same during frosting. This may be the reason why the FEC is as high as 90.5%, with frost accumulation on Circuit

162

Defrosting for Air Source Heat Pump

9.5 Case 1

Case 2

8.5

o

Air temperature difference ( C)

9.0

8.0 7.5 7.0 6.5 6.0 5.5 5.0 0

500

1000

1500

2000

2500

3000

3500

4000

Time (s)

Fig. 6.10 Measured air temperature difference during frosting in two cases.

–6 Case 1

Case 2

o

Tube surface temperature ( C)

–8 –10 –12 –14 –16 –18 –20 –22 0

500

1000

1500

2000

2500

3000

3500

4000

Time (s)

Fig. 6.11 Measured tube surface temperatures at the exit of the outdoor coil during frosting in the two cases.

2 a little less. Therefore, it is demonstrated that the measured tube surface temperatures at the exits of the three refrigerant circuits during frosting could reflect the FEC. As shown in Figs. 6.8 and 6.9, the measured refrigerant volume flow rate and pressure drop across the outdoor coil kept steady from 0 s to 2500 s during frosting. However, 2500 s into frosting, the refrigerant volume flow rate decreased and the

Frosting evenness coefficient

163

0 Fin in Case 1

Fin in Case 2

–4

o

Fin surface temperature ( C)

–2

–6 –8 –10 –12 –14 –16 0

500

1000

1500

2000 Time (s)

2500

3000

3500

4000

Fig. 6.12 Measured average fin surface temperatures at the center of the three circuits during frosting in the two cases.

4.8 Case 1

Case 2

4.4

COP (/)

4.0

3.6

3.2

2.8 0

500

1000

1500

2000

2500

3000

3500

4000

Time (s)

Fig. 6.13 COP variation of the ASHP unit during frosting in the two cases.

pressure drop increased. This resulted from the heat transfer between the ambient air and the refrigerant becoming hard as frost accumulated on the surface of the outdoor coil. And this agrees well with the temperature decrease of the refrigerant, as shown in Figs. 6.6 and Figs. 6.7. Due to the higher FEC in Case 2 than that in Case 1, the trends of increasing and decreasing in Case 1 were bigger than those in Case 2.

164

Defrosting for Air Source Heat Pump

During frosting, the measured air temperature difference between the indoor coil inlet and outlet, the tube surface temperatures at the exit of the outdoor coil, and the average fin surface temperatures at the center of the three circuits during frosting in the two cases, were all measured and presented in this paper, and energy consumption in the two cases was calculated. The measured air temperature difference between the indoor coil inlet and the outlet stands for the energy taken from the indoor air during frosting. Measured tube surface temperatures at the exit of the outdoor coil and average fin surface temperatures at the center of the three circuits reflect the energy given from the refrigerant to the outdoor coil during frosting in the two cases, respectively. As shown in Fig. 6.13, from 0 s to 2500 s into frosting, the COP kept above 4.2. However, after the 2500 s point, the COP of the two cases decreased sharply because of frost accumulation on the airside surface of the outdoor coil. Although the COP difference from 0 s to 3600 s between the two cases is small, with 4.10 in Case 1 and 4.26 in Case 2, the COP difference in the last 10 min (3000–3600 s) is obvious, with 3.18 in Case 1 and 4.00 in Case 2. The difference accounted for as high as 20.5%. Moreover, all the curve trends in Figs. 6.10–6.13 are the same. Therefore, the FEC would positively affect the COP of the whole system for an ASHP unit with a multicircuit during frosting. The higher the FEC for a multicircuit outdoor coil, the higher the system’s COP. In this section, an experimental study on even frosting for a vertical three-circuit outdoor coil in an ASHP unit during heating mode was undertaken and the following conclusions could be reached: (1) There are many factors that could affect the frost distribution on each circuit’s surface for an ASHP unit with a multicircuit outdoor coil during frosting, such as the distribution of refrigerant, the distribution of inlet ambient air, the structure of the outdoor coil, etc. For an ASHP unit with a multicircuit outdoor coil, all the factors would finally decide its FEC. (2) The measured tube surface temperature, which means the temperature of the refrigerant, at the exit of each refrigerant circuit during frosting could directly reflect the FEC. Therefore, the tube surface temperature at the exit of each refrigerant circuit could be used as the comprehensive evaluation index of even frosting for an ASHP unit with a multicircuit outdoor coil. (3) Because the measured tube surface temperature at the exit of each refrigerant circuit during frosting was decided by the refrigerant mass flow rate on each circuit, the FEC, for a multicircuit outdoor coil, could be adjusted by the opening degree of the valves, which directly adjust the refrigerant mass flow rate on each circuit. (4) The higher the FEC for an ASHP unit with a multicircuit outdoor coil, the higher the COP could reach. The tube surface temperature at the exit of the outdoor coil and the average fin surface temperature at the center of each circuit during frosting can play a supporting role on analyzing the system’s energy consumption. (5) For an ASHP unit with a three-circuit outdoor coil, when the FEC increased from 75.7% to 90.5%, the COP could increase from 4.10 to 4.26 at a 3600 s frosting process, and increased from 3.18 to 4.00 in its last 600 s.

6.3

Defrosting performances at different FECs

For a multicircuit outdoor coil in an ASHP unit, uneven defrosting might result from an uneven frosting start, which means the frost accumulations on each circuit’s surface

Frosting evenness coefficient

165

of a multicircuit outdoor coil in an ASHP unit were different. It is easy to understand that the FEC on an outdoor coil surface is decided by the distribution of inlet air passing through each circuit, the distribution of refrigerant flowing into each circuit, the structure of the outdoor coil, the fin’s space, type, and surface characteristics, and so on. Therefore, it is hardly possible to make the frost accumulations on each circuit’s surface equal, leading to different FECs as an RCD operation start for an ASHP unit. For example, as shown in Fig. 6.2, the frost that accumulated on the surface of the outdoor coil is not even. Moreover, for an uneven defrosting operation, the most important possibility is the RCD starts at an uneven frosting. At the same time, at the start of defrosting, different FECs should affect the heat transfer between the frost, the melted frost, the ambient air, and the refrigerant, and thus the defrosting performance of the whole system. This is a fundamental and meaningful problem for ASHP units with multicircuit outdoor coils. Understanding defrosting performance of an ASHP unit with a multicircuit outdoor coil at different FECs is of importance for the ASHP units’ application, but studies are scarce in the open literature. Therefore, an experimental study on defrosting performance when frost accumulates on the surface of the outdoor coil at different FECs has been carried out and a comparative and quantitative analysis conducted using the experimental data. In this section, the detailed description of an experimental ASHP unit and the experimental conditions are first reported. This is followed by presenting the experimental results. Result analyses and conclusions are finally given.

6.3.1 Experimental cases In this section, the experimental setup, procedures, and conditions are totally the same as those introduced in Chapter 3. Therefore, the experimental case design process is directly given here. A series of experimental works using the experimental ASHP unit was carried out to study the defrosting performance when frost accumulated on the outdoor coil surface at different FECs. In order to obtain meaningful experimental results, first, it was necessary to ensure that the frost accumulated on the surface of the three circuits was at different FECs. For an ASHP unit with a multicircuit outdoor coil, it is hard to adjust the FEC, as many parameters affect frosting performance. However, in this section, the modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to vary the refrigerant flow to each circuit, and thus the frost accumulations on each circuit were adjusted. Therefore, to adjust the refrigerant flow into each circuit, a series of trial-and-error manual adjustments of the opening degrees of the stop valves was carried out. Then, a set of fixed valve opening degrees was obtained, and the frost that accumulated on the surface of an outdoor coil at different FECs could be reached. Second, during the experiments, water-collecting trays were used to collect the melted frost from its upside circuit, and thus to calculate the relative FEC. Once the opening degrees of the stop valves on the three circuits were fixed, the water-collecting trays were taken away during defrosting. Finally, experimental work was carried out on three experimental cases, as listed in Table 6.2, so the system defrosting performances at different FECs could be comparatively and quantitatively analyzed.

166

Defrosting for Air Source Heat Pump

Table 6.2 Results of the three experimental cases Item

Parameter

Case 1

Case 2

Case 3

1

Melted frost collected from Circuit 1 Melted frost collected from Circuit 2 Melted frost collected from Circuit 3 FEC Total mass of melted frost collected Results shown in

301 g

280 g

288 g

316 g

309 g

305 g

261 g

305 g

295 g

82.6% 878 g

90.6% 894 g

96.6% 881 g

Figs. 6.14, 6.15, 6.18, 6.21–6.24

Figs. 6.14, 6.16, 6.19, 6.21–6.24

Figs. 6.14, 6.17, 6.20–6.24

2

3

4 5

6

6.3.2 Experimental results Fig. 6.14 presents three photographs showing the airside surface conditions of the outdoor coil at the start of defrosting operations in the three cases. As observed from these pictures, the surface conditions at the start of defrosting for each circuit in the three cases were visually the same, which agreed well with the frost accumulated on the surface of each circuit, as listed in Table 6.4. Moreover, it could be found that it is hardly possible to visually distinguish the FEC difference among the three cases, although their values are calculated totally differently. Therefore, the study on exploring and concluding the defrosting performance of an ASHP unit with a vertically

Fig. 6.14 Airside surface conditions of the outdoor coil at the start of defrosting operation in the three cases. (A) Case 1, (B) Case 2, (C) Case 3.

Frosting evenness coefficient

167

installed multicircuit outdoor coil when its defrosting operation starts at different FECs is very meaningful and fundamental. As shown in Table 6.2, the results of the three experimental cases are listed in detail. The total mass of frost accumulated was weighed at 878 g in Case 1, 894 g in Case 2, and 881 g in Case 3. The three values are nearly the same, with the biggest difference of 16 g, or just 1.8% ( T1 > T3

24

o

Temperature of tube surface ( C)

28

20 16 T2 > T3 > T1

12 8 4

205 s

126 s 144 s

0 80

100

120

140

160

202 s 180

200

220

Time (s)

Fig. 6.15 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 1.

168

Defrosting for Air Source Heat Pump

32 Circuit 1

Circuit 2

Circuit 3 T1 > T 2 > T3

24

o

Temperature of tube surface ( C)

28

20 16

T2 > T1 > T3

12 8 4 148 s

0 80

100

120

140

160

198 s

196 s 180

200

220

Time (s)

Fig. 6.16 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 2.

32 Circuit 1

Circuit 2 Circit 3 T1 > T2 > T3

24

o

Temperature of tube surface ( C)

28

20 16 12

T2 > T1 > T3

8 4

185 s

138 s

0 80

100

120

140 160 Time (s)

180

200

220

Fig. 6.17 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 3.

Frosting evenness coefficient

169

32

Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

24

T 3 > T1 > T2

o

Temperature of fin surface ( C)

28

20 16 12 8 4 0 80

100

120

140

125 s

220 s

160 180 Time (s)

200

235 s 220

240

260

Fig. 6.18 Measured fin surface temperature at the center of three refrigerant circuits during defrosting in Case 1.

32 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

T1 > T2 > T3

24

o

Temperature of fin surface ( C)

28

20 16 12 T3 > T2 > T1

8 4

125 s

150 s

0 80

100

120

140

160

185 s 180

200

218 s 220

240

260

Time (s)

Fig. 6.19 Measured fin surface temperature at the center of three refrigerant circuits during defrosting in Case 2.

170

Defrosting for Air Source Heat Pump

32 T1 > T2 > T3

T1 > T3 > T2

24

o

Temperature of fin surface ( C)

28

Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

20 16 12 8 4

185 s

0 80

100

120

140

196 s

160 180 Time (s)

200

220

240

260

Fig. 6.20 Measured fin surface temperature at the center of three refrigerant circuits during defrosting in Case 3.

Measured refrigerant volumetric flow rate (L/M)

1.8 Case 1

Case 2

Case 3

1.6 1.4 Fluctuating

1.2 1.0 0.8 0.6

R Case 3 > R Case 2 > R Case 1 0.4 180 s

174 s

0.2

180 s 0.0 0

40

80

120 Time (s)

160

200

240

Fig. 6.21 Measured refrigerant volumetric flow rate during defrosting in the three cases.

Frosting evenness coefficient

171

2.8 Case 2 Case 3 TCase 1 > TCase 2 > T Case 3

o

Temperature of melted frost collected ( C)

Case 1 2.4 2.0

T Case 3 > TCase 2 > TCase 1

1.6

Temperature of surrounding air

1.2 Temperature of melted frost collected

0.8 0.4

145 s

135 s

150 s

187 s

0.0 0

30

60

90

120 150 Time (s)

180

210

240

Fig. 6.22 Measured temperature of the surrounding air and measured temperature of the melted frost collected in Cylinder C during defrosting in the three cases.

Finally, a variation of the measured refrigerant volumetric flow rate during defrosting in the three cases is shown in Fig. 6.22. Due to the negative effects of downward-flowing melted frost on defrosting performance, the temperature curves of the three circuits’ order should be at T1 > T2 > T3, the same as that shown in Fig. 3.13 in Chapter 3. However, as shown in Fig. 6.15, the temperatures were observed at T2 > T3 > T1 from 126 s to 144 s, and at T2 > T1 > T3 from 144 s to 205 s into defrosting. This contradictory phenomenon might owe to the fact that the FEC was calculated at just 82.5% in Case 1, which was less than that reported in Chapter 3, at a FEC higher than 90%. In addition, the same phenomenon came out in the other two cases, although the FECs were both calculated higher than 90%, at 90.6% in Case 2 and 96.6% in Case 3. In Case 2, as shown in Fig. 6.16, the temperature curves’ order kept at T2 > T1 > T3 from 80 s to 148 s. After 148 s, the order was changed to at T1 > T2 > T3. In Case 3, as shown in Fig. 6.17, the order was at T2 > T1 > T3 from 80 s to 138 s, and then changed to at T1 > T2 > T3 from 138 s to 185 s. Therefore, an uneven refrigerant distribution for three circuits during defrosting might be another reason resulting in the temperature curves of three circuits order not at T1 > T2 > T3 during defrosting. This also may be the reason why the temperature curve of Circuit 2 was always higher than the others in the three cases. Moreover, as shown in Figs. 6.15–6.17, the defrosting durations were 202 s in Case 1, 196 s in Case 2, and 185 s in Case 3, respectively. It means the defrosting duration would be decreased when an RCD operation starts at a higher FEC. Therefore, the conclusion that the system defrosting performance could be improved by a higher

172

Defrosting for Air Source Heat Pump

FEC for an ASHP unit with a vertically installed multicircuit outdoor coil was first demonstrated. The same as the tube surface temperature shown, it could be seen from Figs. 6.18–6.20, that the orders of the fin surface temperature curves were different from that shown in Fig. 3.16 in Chapter 3. In Case 1, the order was observed at T3 > T1 > T2 from 202 s to 220 s into defrosting. With respect to the order in Case 2, it was at T3 > T2 > T1 from 125 s to 150 s, and at T1 > T2 > T3 from 185 s to 240 s. When it came to Case 3, the order was at T1 > T2 > T3 from 80 s to 185 s, and at T1 > T3 > T2 from 185 s to 220 s. However, in Chapter 3, the fin temperature curves were always shown at T1 > T2 > T3 because of the negative effects of downwardflowing melted frost. The same as the previous reason that tube surface temperature curves show, this contradictory phenomenon should be resulted from a lower FEC and uneven distribution of refrigerant. In addition, as shown in Figs. 6.18–6.20, the durations of fin surface temperature that reached 24°C were 235 s in Case 1, 218 s in Case 2, and 196 s in Case 3, respectively. The durations decreased with their FECs increasing. This further demonstrates the negative effects of a lower FEC on the RCD performance of an ASHP unit with a multicircuit outdoor coil. Fig. 6.21 presents the measured refrigerant volumetric flow rate during defrosting in the three cases. It was observed that the measured refrigerant volumetric flow rate kept fluctuating severely from 0 s to 80 s, especially during the first 40 s into defrosting. This was because that compressor discharge pressure increased suddenly at the start of an RCD operation, and the internal diameter of the EEV was very small. In addition, a lot of energy was consumed during defrosting at the frost-melting stage described in Chapter 4, with a lot of refrigerant changing phases from the gas state to the two-phase state. Therefore, the measured refrigerant volumetric flow rate fluctuated with severe pressure changes. When the defrosting process came into the water layer vaporizing stage described in Chapter 4, the pressures of compressor suction and discharge both increased, leading to the refrigerant volumetric flow rate change from increasing to decreasing. As shown in Fig. 6.21, from Case 1 to Case 3, their peak values came out at 180 s, 180 s, and 174 s, respectively. Here, it is further confirmed that the defrosting performance would be improved with a higher FEC as a defrosting start for an ASHP unit with a vertically installed multicircuit outdoor coil. Fig. 6.22 shows the variation of the measured temperature of the surrounding air and the measured melted frost collected in Cylinder C during defrosting in the three cases. Before the water was collected, the temperature of the surrounding air was measured. When the frost melted and flowed into the water-collecting cylinder, the temperature of the melted frost collected was measured. As shown in Fig. 6.22, the temperatures of the melted frost collected reached their lowest value at 150 s in Case 1, 145 s in Case 2, and 135 s in Case 3, respectively. This phenomenon could also directly demonstrate the negative effects of a lower FEC on the RCD performance of an ASHP unit with a multicircuit outdoor coil. It was obvious that the temperature of the melted frost collected was very low, at about 0.2–0.4°C at the beginning of the melted frost collection. The temperatures of the melted frost collected would increase sharply with the heat from the surrounding air and the later high temperature melted

Frosting evenness coefficient

173

frost coming from water-collecting Tray C. As shown in Fig. 6.22, at 187 s into defrosting, the temperature curve order was observed at T1 > T2 > T3, although the surrounding air temperature order was totally different at T3 > T2 > T1, as shown in the first 140 s in Fig. 6.22. It could be concluded that the water temperature is mainly decided by the heat transferred from the outdoor coil through the later warmer melted frost flowing from the circuit to the tray and then to the cylinder. In addition, the temperature curve order confirmed that less heat was taken from the outdoor coil by melted frost for a higher FEC as a defrosting start for an ASHP unit. Therefore, the negative effects of a lower FEC on the defrosting performance of an ASHP unit with a multicircuit outdoor coil were further demonstrated.

6.3.3 Energy analysis and discussions The energy used for the RCD comes from three sources: the power input to the compressor, the power input to the indoor air fan, and the thermal energy from indoor air. As shown in Fig. 6.23, the energy supplies for defrosting in the three cases were calculated, with the calculated relative standard errors listed in Table 6.3. In this experimental study, the total energy used for defrosting was calculated at 768.6 kJ in Case 1, 743.8 kJ in Case 2, or 3.2% less, and 648.8 kJ in Case 3, or 15.6% less than that in Case 1, respectively. The main difference came from the thermal energy from indoor air, with a difference of 102.8 kJ between Case 1 and Case 3. However, the ratio of this part of the energy was kept at around 83%–85%, without an obvious change with the energy supply decreasing.

Energy supply (kJ)

1,000 900

Input to the compressor and indoor air fan Heat supply from indoor air

800

768.6 kJ

743.8 kJ

700

648.8 kJ

600 500 400

644.0 kJ (83.8%)

632.6 kJ (85.1%)

541.2 kJ (83.4%)

124.6 kJ

111.2 kJ

107.6 kJ

Case 1

Case 2

Case 3

300 200 100 0

Fig. 6.23 Energy supplies for defrosting in the three cases.

174

Defrosting for Air Source Heat Pump

Table 6.3 Differential analysis of durations and defrosting efficiencies in the three cases Item

Parameter

D12a

D23b

D12/D23c

1 2 3 4 5 6 7

FEC Defrosting duration Fin surface temperatured Refrigerant volumetric flow ratee Temperature of melted frost collectedf Total energy for defrosting Defrosting efficiency

8.0% 6s 17 s 0s 5s 24.8 kJ 1.6%

6.0% 11 s 22 s 7s 10 s 95 kJ 5.2%

133.0% 54.6% 77.3% 0.0% 50.0% 26.1% 31.7%

a

The differential value between those in Case 1 and 2. The differential value between those in Case 2 and 3. c The ratio of D12 and D23. d Duration of fin surface temperature reached 24°C. e Duration of refrigerant volumetric flow rate reached its peak value. f Duration of melted frost collected temperature reached its lowest value. b

The defrosting efficiency can be used to evaluate the performance of a defrosting operation. It is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation. In this section, the defrosting efficiencies were calculated at 42.0% in Case 1, at 43.4% in Case 2, and at 48.7% in Case 3, as shown in Fig. 6.24. The difference of defrosting efficiency between Case 1 and Case 3 was 6.8%. Moreover, Fig. 6.24 shows the defrosting durations, and the durations of the fin surface temperature

Defrosting duration Melted frost collected 39 s

240 220

6.77%

15 s

160 140

Case 2 (90.6%)

44

40 Case 3 (96.6%)

Case 1 (82.6%) 100 82

46

42

120

80

52

48

6s

180

54

50

17 s

200 Time (s)

Fin Refrigerant Defrosting efficiency

84

86

88

90

92

94

96

98

FECs (%)

Fig. 6.24 Different durations and defrosting efficiency in the three cases.

38 100

Defrosting efficiency (%)

260

Frosting evenness coefficient

175

reached 24°C, the refrigerant volumetric flow rate reached its peak value, and the temperature of the melted frost collected reached its lowest value, in the three cases, respectively. It could be found that the differences between Case 1 and Case 3 were 39 s for fin surface temperature that reached 24°C, 17 s for defrosting duration, 15 s for the temperature of the melted frost collected to reach its lowest value, and 6 s for the refrigerant volumetric flow rate to reach its peak value, from high to low. All the previous five parameters could demonstrate that the defrosting performance could be improved when an RCD operation starts at a higher FEC for an ASHP unit with a multicircuit outdoor coil. Furthermore, Table 6.3 shows the differential analysis of durations and defrosting efficiencies in the three cases, which listed differential values between those of Case 1 and 2, and those of Case 2 and 3. To study the relationship of different durations, defrosting efficiency, and FEC, their differential value ratios were calculated with the following formula in this table. It could also be found that only the D12/D23 value of FEC was calculated at higher than 100%, at 133.0%. All other values were less than 100%, especially the D12/D23 value of the refrigerant volumetric flow rate at 0.0%. The relationship between different parameters and the FEC should be further studied. In conclusion, a comparative experimental study on the defrosting performance of an ASHP unit with a vertically installed multicircuit outdoor coil at different FECs was undertaken and the following conclusions could be reached: (1) The negative effects of uneven frosting as a defrosting operation starts on defrosting performance were confirmed, and the performance would be better when it starts at a higher FEC. However, the defrosting duration, defrosting efficiency, and other parameters to evaluate the defrosting performance were not found to be positively correlated with the FEC. Therefore, to study the relationship of defrosting performance evaluating indexes and the FEC, more experimental and numerical studies should be conducted. (2) During experiments, the water-collecting trays were taken away after the FEC was calculated. As reported in Chapter 3, when the melted frost is locally drained with water-collecting trays installed between circuits, the negative effects of downward flowing melted frost on the defrosting performance of an ASHP unit would be eliminated. Therefore, it would be meaningful to carry out an experimental study on the defrosting performance of an ASHP unit at different FECs with melted frost locally drained. (3) There are many parameters that could be used to confirm the negative effects on defrosting performance for a lower FEC as the RCD starts. Therefore, these parameters could be used to regulate the termination of an RCD operation, especially the fin surface temperature reaching some preset value and the refrigerant volumetric flow rate reaching its peak value. This work is valuable, and should be further experimentally studied.

6.4

Defrosting performance at different FECs with local drainage of the melted frost

In order to minimize the refrigerant pressure loss along the tube’s inside and enhance the heat transfer between the inside refrigerant and the outside ambient air via the tube and fins, an outdoor coil used in an ASHP unit usually consists of a multicircuit

176

Defrosting for Air Source Heat Pump

structure. The uneven defrosting phenomenon was found and reported in several previous experimental studies, which means that different circuits’ defrosting processes terminated (refrigerant temperature at the exit of the circuit reached the preset termination temperature) at different times. For a multicircuit outdoor coil in an ASHP unit, uneven defrosting might result from the negative effects of downward-flowing melted frost, and the uneven refrigerant distribution into each circuit during defrosting. In addition, it is hardly possible to make the frost accumulations on each circuit’s surface equal because the frosting performance was affected by the distribution of inlet air passing through each circuit, the distribution of refrigerant flowing into each circuit, structure of the outdoor coil, the fin space, type, and its surface characteristics, and so on. Therefore, a defrosting process with an uneven frosting start, which means the frost accumulations on each circuit’s surface of a multicircuit outdoor coil are different, may be another important and easily found reason for uneven defrosting. To clearly describe the uneven frosting phenomenon, the FEC was defined as the ratio of the minimum mass of frost accumulated among three circuits to the maximum one in this section. Not only uneven frosting may lead to uneven defrosting, which also would affect the heat transfer between the frost, the melted frost, the ambient air, and the refrigerant, and thus the defrosting performance of the whole defrosting system. In the previous section, the experimental study on the defrosting performance of an ASHP unit with a multicircuit outdoor coil at different FECs was conducted, and an increase of 6.8% in defrosting efficiency was confirmed when the FEC changed from 82.6% to 96.6%. However, during defrosting in that study, the negative effects of downward-flowing melted frost were not eliminated. As shown in Fig. 6.25, when the total mass of frost accumulation is assumed at 270 g on the airside surface of a three-circuit outdoor coil, there are six situations with the same FEC at 80%. Table 6.4 lists all the masses of downward-flowing melted frost into the down circuits during defrosting in the six situations, with no water-collecting trays installed between the circuits and vaporized water neglected. It is obvious that the total mass of downward-flowing melted frost into Circuit 2 is 100 g and 80 g, and into Circuit 3190 g and 170 g, respectively, in Figs. 6.25A and 6.25F. Their total mass of downward flowing melted frost into down circuit(s) are 290 g and 250 g, with 40 g or 13.8% difference, as listed in Table 6.4. Consequently, different negative effects of downward-flowing melted frost would make the effects of FEC on defrosting performance hard to quality and quantitatively confirmed. This is a fundamental and meaningful problem for ASHP units with multicircuit outdoor coils. Understanding the defrosting performance of an ASHP unit with a multicircuit outdoor coil at different FECs is of importance for ASHP units’ application, but studies are scarce in the open literature. In this section, an experimental study on the defrosting performance for an ASHP unit at different FECs with melted frost local drainage has been carried out. First, the ASHP unit under experiment is presented, followed by the experimental procedures and conditions. Thereafter, the experimental cases and their results are given. The defrosting durations, the energy sources for defrosting, and the energy consumption during defrosting for each case study are measured and discussed, with a conclusion given at the end.

Frosting evenness coefficient

177

C f1 > C f 2 and C f 1 > C f3

C f 1 < C f 2 or C f 1 < C f 3

C f 1 < C f 2 and C f 1 < C f 3

Circuit 1

90 g

Circuit 1

80 g

Circuit 1

90 g

Circuit 2

100 g

Circuit 2

100 g

Circuit 2

80 g

Circuit 3

80 g

Circuit 3

90 g

Circuit 3

Cf 2 > Cf 3

100 g

(A)

(C)

(E)

Circuit 1

90 g

Circuit 1

80 g

Circuit 1

80 g

Circuit 2

80 g

Circuit 2

90 g

Circuit 2

90 g

Circuit 3

100 g

Circuit 3

100 g

Circuit 3

C f 2 < C f3

100 g

(B)

(D)

(F)

Fig. 6.25 Frost accumulations on the surface of three refrigerant circuits for FEC ¼ 80%. (A) 10:9:8, (B) 10:8:9, (C) 9:10:8, (D) 9:8:10, (E) 8:10:9, (F) 8:9:10.

Table 6.4 The mass of downward-flowing melted frost into down circuits Mass of melted frost

2(a)

2(b)

2(c)

2(d)

2(e)

2(f )

Unit

Into Circuit 2 Into Circuit 3 Total

100 190 290

100 180 280

90 190 280

90 170 260

80 180 260

80 170 250

g g g

6.4.1 Experimental cases As with the previous section, the experimental setup, procedures, and conditions in this section are totally the same as those introduced in Chapter 3. Therefore, the experimental case design process is directly given here. A series of experimental works using the experimental ASHP unit was carried out to study the defrosting performance when frost accumulated on the outdoor coil surface at different FECs with melted frost locally drained. In order to obtain meaningful experimental results, first, it was necessary to ensure that the frost that accumulated on the surface of the three circuits was at different FECs. For an ASHP unit with a multicircuit outdoor coil, it is hard to adjust

178

Defrosting for Air Source Heat Pump

Table 6.5 Results of the three experimental cases Item

Parameter

Case 1

Case 2

Case 3

1

Melted frost collected from Circuit 1 Melted frost collected from Circuit 2 Melted frost collected from Circuit 3 FEC Total mass of melted frost collected Results shown in

374 g

330 g

317 g

329 g

362 g

328 g

297 g

360 g

324 g

79.4% 1000 g

91.2% 1052 g

96.6% 969 g

Figs. 6.27, 6.30, 6.33, 6.34, 6.37–6.40

Figs. 6.28, 6.31, 6.33, 6.35, 6.37–6.40

Figs. 6.29, 6.32, 6.33, 6.36–6.40

2

3

4 5

6

the FEC, as many parameters affect frosting performance. However, in this section, modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to vary the refrigerant flow to each circuit, and thus the frost accumulation on each circuit is adjusted. Therefore, to adjust the refrigerant flow into each circuit, a series of trial-and-error manual adjustments of the opening degrees of the stop valves was carried out. Then, a set of fixed valve opening degrees was obtained, and frost accumulated on the surface of an outdoor coil at different FECs could be reached. Thereafter, the FEC was calculated by the mass of melted frost collected from water-collecting cylinders, with the water vaporized into the ambient air neglected. During the experiments, water-collecting trays were installed to collect the melted frost from its upside circuit. Finally, experimental work was carried out on three experimental cases, as listed in Table 6.5, so the system defrosting performances at different FECs with melted frost local drainage could be comparatively and quantitatively analyzed.

6.4.2 Experimental results Fig. 6.26 presents three photographs showing the airside surface conditions of the outdoor coil at the start of defrosting operations in the three cases. As observed from these pictures, the surface conditions at the start of defrosting for each circuit in the three cases were visually the same, which agreed well with the frost accumulated on the surface of each circuit, as listed in Table 6.5. Moreover, it could be found that it is hardly possible to visually distinguish the difference of FECs among the three cases, although their values are calculated totally differently. Therefore, the study on

Frosting evenness coefficient

179

Fig. 6.26 Airside surface conditions of the outdoor coil at the start of defrosting operations in the three cases. (A) Case 1, (B) Case 2, (C) Case 3.

exploring and concluding the defrosting performance of an ASHP unit with a vertically installed multicircuit outdoor coil when its defrosting operation starts at different FECs is meaningful and fundamental. As shown in Fig. 6.26, the water-collecting trays are installed between circuits, which take the melted frost away during defrosting, leading to the negative effects of downward-flowing melted frost being eliminated. As shown in Table 6.5, the results of the three experimental cases are listed in detail. The total mass of frost accumulated was weighed at 1000 g in Case 1, 1052 g in Case 2, and 969 g in Case 3, respectively. The three values are nearly the same, with the biggest difference of 83 g, or around 7.9% ( Tf

8

188 s 198 s

4

203 s

0 80

100

120

140

160

180

200

220

240

Time (s)

Fig. 6.33 Average measured tube and fin surface temperatures during defrosting in the three cases.

1.8 Circuit 1

Circuit 2

Circuit 3

1.5

o

Temperature ( C)

1.2 T a3 > T a2 > T a1

0.9 0.6

T w2 > T w3 ; T w2 > T w1

150 s

0.3

138 s 0.0

50 s

136 s

–0.3 0

30

60

90

120 Time (s)

150

180

210

240

Fig. 6.34 Measured temperatures of surrounding air and measured temperature of the melted frost collected in the three cylinders during defrosting in Case 1.

respectively. Therefore, the defrosting durations in the three cases are 197 s for Case 1, 187 s for Case 2, and 175 s for Case 3, respectively. Therefore, the experimental results first demonstrated that the defrosting duration could be shortened by about 22 s, or about 11.2%, when the FEC was increased from 79.4% to 96.6%.

184

Defrosting for Air Source Heat Pump

1.8 Circuit 1

Circuit 2

Circuit 3

1.5

o

Temperature ( C)

1.2 0.9 T w1 > Tw2 > T w3

T a3 > T a2 > Ta1

0.6

135 s

0.3

130 s 0.0

130 s

–0.3 0

30

60

90

120

150

180

210

240

Time (s)

Fig. 6.35 Measured temperatures of surrounding air and measured temperature of the melted frost collected in the three cylinders during defrosting in Case 2. 1.8 Circuit 1

Circuit 2

Circuit 3

1.5 T a3 > T a2 > T a1

o

Temperature ( C)

1.2 0.9 0.6

T w1 > T w3; T w2 > T w3

125 s

0.3

125 s 0.0

115 s

–0.3 0

30

60

90

120 Time (s)

150

180

210

240

Fig. 6.36 Measured temperatures of surrounding air and measured temperature of the melted frost collected in the three cylinders during defrosting in Case 3.

In addition, as shown in Figs. 6.27–6.29, the maximum deviations of temperature of three circuits are 1.87°C at 125 s for Case 1, 2.73°C at 122 s for Case 2, and 3.73°C at 120 s for Case 3, respectively. It is obvious that the order of the maximum deviation of temperature is at Case 1 < Case 2 < Case 3. However, the order of the FECs of the

Frosting evenness coefficient

185

1.8 Case 1 1.5

Case 2

Temperature of surrounding air

o

Temperature ( C)

1.2

Case 3 Temperature of melted frost collected

0.9 0.6 157 s

0.3

141 s 0.0

135 s

–0.3 0

30

60

90

120

150

180

210

240

Time (s)

Fig. 6.37 Average value of measured temperature of surrounding air and melted frost collected in the three cases.

Measured refrigerant volumetric flow rate (L/M)

1.8 Case 1 Case 2 Case 3

1.6

Steeply decrease

1.4 Steeply increase

1.2 1.0

Fluctuate

0.8 0.6

R Case 3 > R Case 2 > R Case 1

165 s 170 s

0.4 Fluctuate

0.2

175 s

0.0 0

40

80

120

160

200

240

Time (s)

Fig. 6.38 Measured refrigerant volumetric flow rate during defrosting in the three cases.

three cases is at Case 1 > Case 2 > Case 3. Moreover, as listed in Table 6.5, in Case 2 and 3, the mass of melted frost collected from the surface of Circuit 2 is the biggest among the three circuits. However, as shown in Figs. 6.28 and 6.29, the temperature of Circuit 2 is always the highest among the three circuits during defrosting. Therefore, it is contradictory between the maximum deviations and the FECs, and the mass of

186

Defrosting for Air Source Heat Pump

melted frost collected and temperature curves, for an ASHP unit with a multicircuit outdoor coil. This should result from the uneven refrigerant distribution into each circuit during defrosting because of their different inner tube resistances in the three refrigerant circuits or the gravity effects on refrigerant distribution. Figs. 6.30–6.32 show the measured fin surface temperatures at the center of the three refrigerant circuits during defrosting in the three cases. The same as the tube surface temperature, the fin surface temperature curves of the three circuits left 0°C at the same time, at 110 s into defrosting, and increased nearly at the same time, with the same trends as that shown in Fig. 3.16 in Chapter 3. The rise in fin surface temperature is about 10 s later than that in tube surface temperature. This is because the tube is in direct contact with the hot refrigerant, but the fin is indirectly in contact with the refrigerant via the tube. In Fig. 6.30, the temperatures of the three circuits reaching 24°C are orderly at 203, 207, and 202 s, respectively, from Circuit 1 to 3. In Fig. 6.31, their durations are orderly at 195, 201, and 198 s, respectively. In Fig. 6.32, the temperatures of the three circuits reaching 24°C are orderly at 182, 184, and 180 s, respectively. Therefore, the durations for the three circuits all reaching 24°C in the three cases are 207 s for Case 1, 201 s for Case 2, and 184 s for Case 3, respectively. Therefore, the experimental results indicated that the defrosting durations of the fin surface temperature all reaching 24°C could be shortened by about 23 s, or about 11.1%, when the FEC was increased from 79.4% to 96.6%. As shown in Figs. 6.30–6.32, the maximum deviations of fin surface temperature of three circuits are also calculated at 2.90°C at 170 s for Case 1, 3.01°C at 150 s for Case 2, and 1.93°C at 130 s for Case 3, respectively. It is obvious that the order of the maximum deviation of temperature is at Case 2 > Case 1 > Case 3. However, the order of the FECs of the three cases is at Case 1 > Case 2 > Case 3. Therefore, the fin surface temperature deviations are not totally affected by the FEC. Moreover, as listed in Table 6.5, the mass of melted frost collected from the surface of Circuit 2 is the biggest among the three circuits in Case 2. However, as shown in Fig. 6.31, the temperature of Circuit 2 is not always the lowest among the three circuits. This phenomenon should also result from the uneven refrigerant distribution into each circuit during defrosting, the same as the previously described contradictory phenomenon shown in tube surface temperature. Fig. 6.33 shows the average measured tube and fin surface temperatures during defrosting in the three cases. It is obvious that, from 80 to 125 s into defrosting, the tube surface temperatures are always higher than the fin surface temperatures. This is also because the tube is directly in contact with the hot refrigerant, but the fin is indirectly in contact with the refrigerant via the tube. Therefore, the rise in fin surface temperature is later than that in tube surface temperature. Second, as previously described, the tube surface temperature and the fin surface temperature are always kept at the order of TCase 1 > TCase 2 > TCase 3. This is because the FEC in Case 1 is the lowest, but the FEC in Case 3 is the highest. Third, the defrosting durations, calculated by the average measured values, are 188 s for Case 1, 186 s for Case 2, and 163 s for Case 3, respectively. The fin surface temperatures reaching 24°C from Case 1 to 3 are orderly at 203, 198, and 182 s, respectively. The difference between Case 1 and Case 3 is 25 s for tube surface temperature, or 13.3%, and 21 s for fin

Frosting evenness coefficient

187

surface temperature, or 10.3%. This further demonstrated that a better defrosting performance could be reached when the FEC increased. Finally, the maximum deviation between Case 1 and Case 3 for tube surface temperature is 5.4°C, and 9.8°C for fin surface temperature, both at 160 s into defrosting. Therefore, the fin surface temperature may be better to evaluate the defrosting performance when the FEC is changed. Figs. 6.34–6.36 show the measured temperature of the surrounding air and the measured temperature of the melted frost collected in the three cylinders during defrosting in the three cases. Before the melted frost flowed downward from the water-collecting trays into the relative cylinders, the temperatures of the surrounding air were measured and the data collected. When the melted frost flowed into the cylinder, the temperature of the melted frost started to be measured. Therefore, there are sharp decreasing stages in the three figures. In Case 1, as shown in Fig. 6.34, the melted frost starts flowing into the cylinder at 150 s for Circuit 1, 136 s for Circuit 2, and 138 s for Circuit 3, respectively. In Case 2, as shown in Fig. 6.35, the melted frost flowing into the cylinder from Circuit 1 to 3 is orderly at 135 s, 130 s, and 130 s, respectively. In Case 3, as shown in Fig. 6.36, the time when the melted frost started flowing into the cylinder was changed to 125 s for Circuit 1, 115 s for Circuit 2, and 125 s for Circuit 3, respectively. Therefore, the durations of melted frost reaching the cylinders are at Circuit 2 < Circuit 3 < Circuit 1 for Case 1, Circuit 2 < Circuit 1 < Circuit 3 for Case 2, and Circuit 2 < Circuit 1 < Circuit 3 for Case 3, respectively. However, the orders of their frost accumulations on each circuit are at Circuit 1 > Circuit 2 > Circuit 3 for Case 1, Circuit 2 > Circuit 3 > Circuit 1 for Case 2, and Circuit 2 > Circuit 3 > Circuit 1 for Case 3, respectively. Although the frost accumulations on Circuit 2 are not the least, they are always the first to be melted and collected. This must result because the refrigerant distributed into Circuit 2 is the maximum among the three refrigerant circuits. However, for Circuit 1 and Circuit 3, when the frost accumulation is more, the later the melted frost was collected. It is reasonable that more frost needs more energy to be melted. In addition, it is obvious that the temperature of the melted frost order is different from the temperature of the surrounding air order in the three cases. This phenomenon demonstrated that the melted frost temperature is mainly affected by the later melted frost flowing into the cylinder, without absorbing heat from the surrounding air. Fig. 6.37 shows the variations of the average value of the measured temperature of the surrounding air and melted frost collected in the three cases. The same as Figs. 6.34–6.36 before the water was collected, the temperature of the surrounding air was measured. When the frost was melted and flowed into the water-collecting cylinder, the temperature of the melted frost collected was measured. As shown in Fig. 6.37, the temperature of the melted frost collected reached its lowest value at 157 s in Case 1, at 141 s in Case 2, and at 135 s in Case 3, respectively. This phenomenon could also directly demonstrate the negative effects of a lower FEC on the RCD performance of an ASHP unit with a multicircuit outdoor coil. It was obvious that the temperature of the melted frost collected was very low, at about 0.2°C. The temperatures of the melted frost collected would increase sharply, with the heat from the surrounding air and the later high-temperature melted frost coming from water-collecting trays. The order of melted frost temperature is nearly totally different from the order of

188

Defrosting for Air Source Heat Pump

surrounding air temperature. This also demonstrated that the melted frost temperature is mainly affected by the later melted frost flowing into the cylinder. Fig. 6.38 presents the measured refrigerant volumetric flow rate during defrosting in the three cases. It is observed that the measured refrigerant volumetric flow rate keeps fluctuating severely from 0 to 80 s, especially during the first 40 s into defrosting. This is because the compressor discharge pressure increased suddenly at the start of an RCD operation, and the internal diameter of the EEV is very small. In addition, a lot of energy is consumed during defrosting at the frost-melting stage described in Chapter 4, with a lot of refrigerant changing phases from the gas state to the two-phase state. Therefore, the measured refrigerant volumetric flow rate fluctuates with severe pressure changes. When the defrosting process came into the water layer vaporizing stage described in Chapter 4, the pressures of the compressor suction and discharge both increase, leading to the refrigerant volumetric flow rate changing from increasing to decreasing. As shown in Fig. 6.38, from 80 to 165 s into defrosting, the order of the measured refrigerant volumetric flow rate during defrosting in the three cases is RCase 3 > RCase 2 > RCase 1. This results because the defrosting performance in Case 3 is better than in Case 1, which makes the refrigerant flow rate increase earlier. Finally, from Case 1 to Case 3, their peak values came out at 175, 170, and 165 s, respectively. Here, it is further confirmed that the defrosting performance would be improved with a higher FEC as a defrosting start for an ASHP unit with a vertically installed multicircuit outdoor coil.

6.4.3 Durations, energy analysis, and discussions Table 6.6 lists the durations for tubes and fins and their DECs in the three cases. It could be found that the DEC for tube surface temperature is 97.97% for Case 1, 97.86% for Case 2, and 98.26% for Case 3, respectively. The DEC for fin surface temperature is 97.58% for Case 1, 97.01% for Case 2, and 97.83% for Case 3, respectively. Obviously, the DEC orders of tube and fin surface temperature in the three cases are the same, at DEC3 > DEC2 > DEC1. But the order is different from that Table 6.6 Durations for tubes and fins, and the DECs in the three cases Item

Parameter

Case 1

Case 2

Case 3

1 2 3 4 5 6 7 8 9 10 11

Duration for tube of Circuit 1 Duration for tube of Circuit 2 Duration for tube of Circuit 3 DEC for tube surface temperature Defrosting duration (for tube) Duration for fin of Circuit 1 Duration for fin of Circuit 2 Duration for fin of Circuit 3 DEC for fin surface temperature Defrosting duration for fin Duration difference

193 s 197 s 193 s 97.97% 198 s 203 s 207 s 202 s 97.58% 207 s 9s

187 s 183 s 186 s 97.86% 188 s 195 s 198 s 201 s 97.01% 200 s 12 s

174 s 172 s 175 s 98.26% 175 s 182 s 184 s 180 s 97.83% 185 s 10 s

Frosting evenness coefficient

189

of their FECs. In addition, the duration differences between the fin surface temperature and the tube surface temperature are nearly the same, at 9 s for Case 1, 12 s for Case 2, and 10 s for Case 3, respectively. The energy used for the RCD comes from three sources: the power input to the compressor, the power input to the indoor air fan, and the thermal energy from indoor air. As shown in Table 6.7 and Fig. 6.39, the energy supplies for defrosting in the three cases were calculated, with the calculated relative standard errors listed in Table 3.3.2 in Chapter 3. In this experimental study, the total energy used for defrosting was calculated at 781.8 kJ in Case 1, 753.2 kJ in Case 2, or 3.7% less, and 678.8 kJ in Case 3, or 13.2% less than that in Case 1, respectively. The main difference came from the thermal energy from the indoor air, with a difference of 92.3 kJ between Case 1 and Case 3. However, the ratio of this part of the energy was kept at around 83%–85%, without obvious changes with the energy supply decreasing. Defrosting efficiency can be used to evaluate the performance of a defrosting operation. It is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation. In this section, the defrosting efficiencies were calculated at 45.0% in Case 1, 48.4% in Case 2, and 50.7% in Case 3, as shown in Table 6.7. The difference of defrosting efficiency between Case 1 and Case 3 was 5.7%. Moreover, Fig. 6.40 shows the defrosting durations and the durations of fin surface temperatures all reaching 24°C, the refrigerant volumetric flow rate reaching its peak value, and the temperature of the melted frost collected reaching its lowest value in the three cases, respectively. It could be found that the differences between Case 1 and Case 3 were 23 s for fin surface temperature all reaching 24°C, 22 s for defrosting duration, 22 s for the temperature of the melted frost collected reaching its lowest value, and 10 s for refrigerant volumetric flow rate reaching its peak value, from high to low. All the previous five parameters could demonstrate that the defrosting performance could be improved when an RCD operation starts at a higher FEC for an ASHP unit with a multicircuit outdoor coil. Table 6.7 Energy supply, energy consumption, and defrosting efficiency in three cases Item

Parameter

Case 1

Case 2

Case 3

Unit

1 2 3 4 5 6 7

The power input to compressor The power input to indoor air fan The energy from the indoor air The power input to outdoor air fan Total energy supply during defrosting Energy consumption on melting frost Energy consumption on vaporizing the retained water Total energy consumption for defrosting Defrosting efficiency

120.6 7.4 653.8 0 781.8 334.0 18.1

112.2 6.8 634.3 0 753.2 351.4 15.9

114.8 2.5 561.5 0 678.8 323.6 20.5

kJ kJ kJ kJ kJ kJ kJ

352.1

367.2

344.2

kJ

45.0%

48.8%

50.7%

–

8 9

190

Defrosting for Air Source Heat Pump The power input to compressor

The power input to indoor air fan

The energy from the indoor air

14.9%

15.42%

16.91% 0.9%

0.95% 120.6 kJ 7.4 kJ

2.5 kJ

6.8 kJ

561.5 kJ

634.3 kJ

653.8 kJ 83.63%

82.72%

84.2%

(A)

0.37%

114.8 kJ

112.2 kJ

(B)

(C)

Energy consumption on vaporizing the retained water Energy consumption on melting frost 4.33% 5.14% 5.96% 20.5 kJ 15.9 kJ 18.1 kJ

334.0 kJ

(D)

323.6 kJ

351.4 kJ

(E)

94.86%

(F)

95.67%

94.04%

Fig. 6.39 Energy supplies for defrosting and consumptions during defrosting. (A, D) Case 1, (B, E) Case 2, (C, F) Case 3.

Defrosting duration Melted frost collected

240

5.7%

22 s

200

52

48 46

180 10 s

44

160 22 s

Time (s)

54

50

23 s

220

Fin Refrigerant Defrosting efficiency

140

42 40

120 Case 1 (79.4%) Case 2 (91.2%)

Case 3 (96.6%)

100 78

80

82

84

86

88 90 FECs (%)

92

94

96

Fig. 6.40 Different durations and defrosting efficiency in the three cases.

38 98

Defrosting efficiency (%)

260

Frosting evenness coefficient

191

In this section, a comparative experimental study on the defrosting performance of an ASHP unit with a vertically installed multicircuit outdoor coil at different FECs with the melted frost locally drained was undertaken and the study results were concluded. The negative effects of uneven frosting at the start of a defrosting operation on defrosting performance were first confirmed, and the performance would be better when it starts at a higher FEC with melted frost locally drained. When the FEC increased from 79.4% to 96.6%, the defrosting duration would be shortened by 23, from 198 s to 175 s, or about 11.2%. At the same time, the total energy supply during defrosting could be decreased by about 3.7%, from 781.8 to 678.8 kJ, and the defrosting efficiency increased by about 5.7%, from 50.7% to 45.0%. In addition, there are many parameters that could be used to confirm the negative effects on defrosting performance for a lower FEC at the RCD start. These parameters could be used to regulate the termination of an RCD operation, especially the fin surface temperature reaching some preset value, and the temperature of the melted frost collected reaching its lowest value.

6.5

Concluding remarks

Uneven defrosting always results in a lower defrosting efficiency and a longer defrosting duration. In Chapter 3, the negative effects of downward-flowing melted frost along the surface of the outdoor coil were experimentally examined. Furthermore, the negative effects of downward-flowing melted frost on defrosting performance were quantitatively analyzed with two-circuit and three-circuit outdoor coils in an ASHP unit. However, frosting unevenly distributed on the surface of a multicircuit outdoor coil is also an important reason for uneven defrosting because the heating load for each circuit is different at the start of defrosting. In this chapter, to further quantitatively describe the uneven frosting, the frosting evenness coefficient was defined as the ratio of the minimum frost accumulation to the maximum one. Uneven frosting results in a lower COP. The higher the FEC for an ASHP unit with a multicircuit outdoor coil, the higher the COP could reach. For an ASHP unit with a three-circuit outdoor coil, when the FEC was increased from 75.7% to 90.5%, the COP could be increased from 4.10 to 4.26 at a 3600 s frosting process. Meanwhile, uneven frosting also downgrades the defrosting performance. The defrosting performance would be better when it starts at a higher FEC, irrespective of melted frost local drainage. For a traditional ASHP unit, when the FEC was increased from 82.6% to 96.6%, the defrosting duration would be shortened by 17 s, from 202 to 185 s, and the defrosting efficiency increased by about 6.7%, from 42.0% to 48.7%. For an ASHP unit with water-collecting trays installed between circuits, when the FEC was increased from 79.4% to 96.6%, the defrosting duration would be shortened by 23 s, from 198 to 175 s, and the defrosting efficiency increased by about 5.7%, from 50.7% to 45.0%. However, more conditions should be further examined. First, the FECs all are very high in this chapter, at the ranges of 80%–90% and 90%–100%. Hence, the defrosting performances for the FEC at lower ranges of 50%–60%, 60%–70%, and 70%–80%

192

Defrosting for Air Source Heat Pump

should be further studied. At the same time, the relationship between the FEC and defrosting performance evaluating indexes could be further analyzed. Second, when the FEC was calculated, the order of frost mass accumulation on the three circuits was not fixed. For example, the frost accumulation order in Section 6.3 was at Circuit 2 > Circuit 1 > Circuit 3 in Case 1, but at Circuit 2 > Circuit 3 > Circuit 1 in Case 2. Due to the negative effects of the downward flow of melted frost, the defrosting performance would be different for the two cases with the same FEC but different frost accumulation mass orders. Therefore, further studies should be carried out considering the frost accumulation mass order. Third, for an ASHP unit, its frosting duration is much longer than its defrosting duration. The FEC effects on a frosting-defrosting cycle should be further analyzed before the installation of valves in each circuit of the outdoor coil was considered in application. Fourth, uneven refrigerant distribution may also result in uneven defrosting. Further experimental investigation on the effects of uneven refrigerant distribution during defrosting in an ASHP unit with a multicircuit outdoor coil should be carried out.

References [1] Wang W, Feng YC, Zhu JH, Guo QC, Lu WP. Performances of air source heat pump system for a kind of mal-defrost phenomenon appearing in moderate climate conditions. Appl Energy 2013;112:1138–45. [2] Steiner A, Rieberer R. Parametric analysis of the defrosting process of a reversible heat pump system for electric vehicles. Appl Therm Eng 2013;61:393–400. [3] Yao Y, Jiang YQ, Deng SM, Ma ZL. A study on the performance on the airside heat exchanger under frosting in air-source heat pump water heater/chiller unit. Int J Heat Mass Transf 2004;47(17–18):3745–56. [4] Seker D, Kartas H, Egrican N. Frost formation on fin-and-tube heat exchanges, Part I: modeling of frost formation on fin-and-tube heat exchangers. Int J Refrig 2004;27(4): 367–74.

The influence of refrigerant distribution on defrosting 7.1

7

Introduction

For an ASHP unit having a vertically installed multicircuit outdoor coil, uneven defrosting, which means different circuits’ defrosting processes are terminated at different times, has been found and reported in limited previous experimental studies. O’Neal et al. [1] and Qu et al. [2] both investigated the transient defrosting performances of ASHP units, each with a vertically installed four-parallel refrigerant circuit outdoor coil. It was reported that when a defrosting process was terminated, the tube surface temperature at the exit of the lowest circuit was much lower than that at the exit of the highest circuit. Thereafter, a modeling study on the impact of allowing melted frost to flow down freely along the coil surface on defrosting performance was carried out by Qu et al. The negative effects of downward-flowing melted frost during defrosting were numerically assessed, and an increase of 18.3% in defrosting efficiency was predicted had the melted frost been drained away locally [3]. To further quantitatively and experimentally confirm the negative effects of downward-flowing melted frost, as introduced in Chapter 3, a series of experimental works using two-circuit and three-circuit outdoor coils was carried out, in which water-collecting trays were placed between circuits to take away the melted frost downward flowing from the up-circuits during RCD. Experimental results demonstrated the negative effects of melted frost on defrosting, and an increase of 13.2% in defrosting efficiency as compared to where no water-collecting trays were installed was reported. To examine the mechanism of uneven defrosting, two semiempirical mathematical models, corresponding to two settings of with and without the use of water-collecting trays between circuits, were also developed and are presented in Chapter 4. However, for an ASHP unit with a vertically installed multicircuit outdoor coil, uneven defrosting might be due to uneven refrigerant distribution. This is because uneven defrosting is fundamentally caused by different heat transfer rates between hot refrigerant and frost or water retained on the surface of an outdoor coil, when the supply of refrigerant to each circuit is the same. In Section 4.3, the control strategy by adjusting the refrigerant distribution into each circuit is also considered and evaluated. The distribution of refrigerant for each circuit is affected by the flow resistance on the refrigerant side when passing through an outdoor coil circuit. In the refrigerant distributor section, the gravity force may also have an influence on the uneven distribution. Therefore, it is hardly possible to make the refrigerant evenly distributed into each circuit. This may be the reason why none of the studies on the uneven refrigerant distribution due to the coupled forces of gravity and tube internal resistance over a

Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00007-2 © 2019 Elsevier Ltd. All rights reserved.

194

Defrosting for Air Source Heat Pump

vertically installed multicircuit outdoor coil in an ASHP unit during defrosting was reported in the open literature. For a multicircuit outdoor coil in an ASHP unit, the effects of uneven refrigerant distribution on defrosting performance is a fundamental problem. Therefore, in this chapter, experimental studies on system defrosting performance when refrigerant was evenly or unevenly distributed into each circuit have been carried out. Also, both with and without the melted frost being locally drained were considered, and a comparative and quantitative analysis was conducted using the experimental data.

7.2

Defrosting performance influenced by uneven refrigerant distribution

An experimental study on system defrosting performance when the refrigerant was evenly or unevenly distributed into each circuit was first carried out in a novel ASHP unit, with the melted frost locally drained by trays. The experimental setup, procedure, and conditions were introduced in previous chapters. In this section, the description of experimental cases is first reported. This is followed by presenting the experimental results, and finally the result analysis and conclusions are given.

7.2.1 Experimental cases A series of experimental works using the experimental ASHP unit has been carried out to study the effects of uneven refrigerant distribution due to gravity and tube internal resistance on defrosting performance. In order to obtain meaningful experimental results, it was necessary to ensure that the frost accumulated on the surface of the three circuits was even at first. For an ASHP unit with a multicircuit outdoor coil, it is hard to adjust the FEC, as many parameters affect frosting performance. However, in this section, modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to vary the refrigerant flow to each circuit, and thus the frost accumulations on each circuit are adjusted. Therefore, to adjust the refrigerant flow into each circuit, a series of trial-and-error manual adjustments of the opening degrees of the stop valves was carried out. Then, a set of fixed valve opening degrees was obtained, and the amount of frost accumulation on the three circuits was close to each other, with their differences smaller than 10%. As previously mentioned, the FEC was defined as the ratio of the minimum frost accumulation among the three circuits to the maximum one. The FEC could be calculated by the masses of melted frost collected from water-collecting cylinders, with the water vaporized into the ambient air neglected. Second, to comparatively study the effects of uneven refrigerant distribution, it was necessary to ensure that the evenness of the refrigerant distributed into the three circuits during defrosting was different. Due to gravity and tube internal resistance directly affecting the refrigerant distribution, it seems hardly possible to adjust the refrigerant for even distribution into the three circuits. However, in this section, modulating valves installed at an inlet refrigerant pipe to each circuit may be deployed to

The influence of refrigerant distribution on defrosting

195

vary the refrigerant flow to each circuit, according to the tube surface temperature at the exit of each circuit. Experimental work was then carried out at two experimental cases, as listed in Table 7.1, so that the system performances under different refrigerant distribution evenness values (RDEVs) can be comparatively and quantitatively analyzed. In this study, to easily distinguish the two experimental cases, similar with the FEC, the RDEV was defined as the ratio of the minimum refrigerant distribution into three circuits to the maximum one. In Case 1, all the stop valves on the three circuits were fully open; therefore, the opening degrees of each stop valve were kept constant. Refrigerant was kept at a fixed RDEV, lower than 100%, due to gravity and tube internal resistance. In Case 2, a series of trial-and-error manual adjustments of the opening degree of the stop valves was carried out to realize how to adjust the refrigerant flow into each circuit evenly, according to the tube surface temperature at the exit of each circuit. Finally, at the start of the defrosting experiment, a suite of suitable degrees was obtained and fixed for the three circuits. The refrigerant could be fixed at an RDEV around 100%. Figs. 7.1 and 7.2 show the measured tube surface temperatures at the exits of the three refrigerant circuits on defrosting mode without any frost accumulated on the surface of the outdoor coil in two cases. As shown in Fig. 7.1, when three stop valves were fully open in Case 1, the temperatures of the three circuits were always different from each other, from 0 to 110 s into the defrosting operation. It is demonstrated that the tube surface temperature of Circuit 1 was the highest, and that of Circuit 2 the lowest. This phenomenon suggests that the amount of refrigerant flowing into Circuit 1 was the highest due to the smallest circuit internal resistance, and that flowing into Circuit 2 was the least due to the biggest circuit internal resistance. In addition, it is clearly Table 7.1 Results of two experimental cases Item

Parameter

Case 1

Case 2

1 2 3 4 5

Stop valves state RDEV Trays during frosting Trays during defrosting Melted frost from Circuit 1 Melted frost from Circuit 2 Melted frost from Circuit 3 Total melted frost collected FEC Defrosting duration Results shown in

Fully open T3 > T2

8

110 s

15 s 4

0

20

40

60

80

100

120

Time (s)

Fig. 7.1 Measured tube surface temperatures at the exits of the three refrigerant circuits on defrosting mode without any frost accumulated in Case 1.

44 Circuit 1

Circuti 2

Circuit 3

36

o

Tube surface temperature ( C)

40

32 28

T1 = T 2 = T3

T2 > T3 > T1

24 20 16

10 s

12

115 s

40 s

8 0

20

40

60 Time (s)

80

100

120

Fig. 7.2 Measured tube surface temperatures at the exits of the three refrigerant circuits on defrosting mode without any frost accumulated in Case 2.

shown that the temperatures were observed at TCircuit1 > TCircuit3 > TCircuit2 from 15 to 110 s defrosting operation point. However, in Case 2, as shown in Fig. 7.2, the temperatures of the three circuits were nearly the same from 40 to 110 s. Only during the first 40 s were the three temperature curves not coincident while the system was

The influence of refrigerant distribution on defrosting

197

adjusting its refrigerant distribution. Before adjusting the opening degrees of the stop valves, three valves were fully open and the ASHP unit was working at frosting mode. Therefore, from 10 to 40 s in Case 2, the temperature of the three circuits was totally opposite to that in Case 1, at TCircuit2 > TCircuit3 > TCircuit1, due to the cold storage in the metal (tubes and fins) of the circuits. In addition, it could be found that the start temperatures in Figs. 7.1 and 7.2 are different. It is also the energy storage in the metal of the outdoor coil that makes the start temperature in Case 2 about 4°C higher than that in Case 1.

7.2.2 Experimental results As shown in Table 7.1, the results of the two experimental cases were listed. The total mass of melted frost collected was 949 g in Case 1 and 944 g in Case 2, which were nearly the same. In Case 1, frost accumulation weighed 325 g on Circuit 1, 322 g on Circuit 2, and 302 g on Circuit 3, respectively. And in Case 2, from Circuit 1 to Circuit 3, frost accumulation on each circuit weighed 311, 324, and 309 g, respectively. The FEC was 92.9% in Case 1 and 95.4% in Case 2, as listed in Table 7.1. Therefore, frost accumulations on the three circuits in the two cases were both close to each other (difference < 10%), which met the requirements described in the previous section. Fig. 7.3 presents two photographs showing the airside surface conditions of the outdoor coil at the start of defrosting in two cases. As observed from Figs. 7.3A and B, the surface conditions at the start of defrosting for each circuit in the two cases were virtually the same, which agreed well with the data listed in Table 7.1. Different from this section, in Case 1 in Section 3.3, without any water-collecting trays were installed between circuits. It is clear that the temperature orders of the three

Fig. 7.3 Airside surface conditions of the outdoor coil at the start of defrosting in two cases. (A) Case 1 (with trays) and (B) Case 2 (with trays).

198

Defrosting for Air Source Heat Pump

circuits for the tube surface and fin center at the same time point are TCircuit 1 > TCircuit 2 > TCircuit 3, from 80 s into the defrosting operation to termination, as shown in Figs. 3.13 and 3.16. In this section, the same parameters in the two cases were measured and presented in Figs. 7.4–7.7, but with water-collecting trays installed between circuits. In all these figures, for their time (horizontal) axis, 80 s is the chosen starting 28 Circuit 1

Circuit 2

Circuit 3

T 2 > T3 > T 1

20

o

Tube surface temperature ( C)

24

16 12 8 o

(T2 – T1)max = 2.8 C 4 125 s

0 80

100

120

140 Time (s)

185 s 160

180

200

Fig. 7.4 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 1. 28 Circuit 1

Circuit 2

Circuit 3

20

T2 > T3 > T1

o

Tube surface temperature ( C)

24

16 12 o

(T 2 – T 1)max = 7.5 C

8 4 125 s

0 80

100

120

140 Time (s)

185 s

172 s 160

180

200

Fig. 7.5 Measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting in Case 2.

The influence of refrigerant distribution on defrosting

199

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

20

T 3 > T 1; T 3 > T 2

o

Fin surface temperature ( C)

24

16 T1 > T3 > T2 12 8 4 217 s

210 s

0 80

100

120

140

160 180 Time (s)

200

220

240

Fig. 7.6 Measured fin surface temperature at the center of the three refrigerant circuits during defrosting in Case 1.

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

20

o

Fin surface temperature ( C)

24

16

T3 > T 1; T 3 > T 2

12 8 4 165 s

0 80

100

120

140

160 180 Time (s)

182 s 200

220

240

Fig. 7.7 Measured fin surface temperature at the center of the three refrigerant circuits during defrosting in Case 2.

200

Defrosting for Air Source Heat Pump

1.6

Refrigerant volumetric flow rate (L / min)

Case 1

Case 2

1.4 1.2 1.0 0.8 0.6 0.4

R Case 2 > R Case1

Fluctuating

0.2

70 s

90 s

155 s

170 s

0.0 0

20

40

60

80

100

120

140

160

180

200

Time (s)

Fig. 7.8 Measured refrigerant volumetric flow rate during defrosting in two cases.

time in order to clearly show the temperature rise during defrosting. It is obvious that all temperature curves’ orders of three refrigerant circuits are not the same as that shown in Figs. 3.13 and 3.16. This phenomenon shows that the negative effects of downward-flowing melted frost were eliminated when the melted frost was taken away locally by the water-collecting trays in this study. Fig. 7.8 shows the measured refrigerant volumetric flow rate during defrosting. Finally, variations of the measured temperatures of the surrounding air and the melted frost collected in the watercollecting Cylinder C in the two cases are shown in Fig. 7.9. As shown in Fig. 7.4, the measured tube surface temperatures at the exits of the three refrigerant circuits during defrosting all reached 24°C at 185 s, which means the defrosting duration for Case 1. It can be seen that the temperature of Circuit 2 was the highest and the temperature of Circuit 1 the lowest at the same time point. From 100 to 190 s into defrosting, the three temperature curves were observed with a relation of TCircuit2 > TCircuit3 > TCircuit1, which was different from those before 100 or after 190 s into the defrosting operation. The maximum temperature difference among the three circuits was 2.8°C, at 125 s into defrosting. The temperature curves’ order of three refrigerant circuits was opposite to that shown in Fig. 3.13 from 15 to 110 s, but the same as the order shown in Fig. 3.16 from 10 to 40 s. This phenomenon may result from the least cold storage in the metal of Circuit 2 among the three circuits during frosting, due to its internal resistance being the highest as shown in Fig. 3.13. As shown in Fig. 7.5, the defrosting duration was just 172 s in Case 2, or 13 s less than that in Case 1. The negative effects of uneven refrigerant distribution on the defrosting performance for an ASHP unit with a multicircuit outdoor coil are first demonstrated. From 80 to 185 s into defrosting, it is obvious that the temperature

The influence of refrigerant distribution on defrosting

201

1.4 Case 1

Case 2

1.2 1.0 Temperature of melted frost collected

o

Temperature ( C)

0.8 0.6 0.4 Temperature of surrounding air

0.2

140 s

0.0 120 s

–0.2 0

50

100

150 Time (s)

200

250

300

Fig. 7.9 Measured temperature of surrounding air and the melted frost collected in the watercollecting Cylinder C during defrosting in two cases.

of Circuit 2 was the highest among the three circuits, and Circuit 1 the lowest. This is because the gravity and tube internal resistance had been eliminated by the stop valve adjustment. However, the cold storage in the metal of the circuit played a more important role on the temperature curves’ order. In addition, the maximum temperature difference came out at 125 s into defrosting in Case 2, the same as the time in Case 1, with a value of 7.5°C, or 4.7°C higher than the temperature in Case 1. It is also the metal energy storage of Circuit 1 that delays the temperature rising in Case 2. Under this condition, the frost melting rate would be a little lower than that of the other two circuits. However, the total mass of melted frost collected in Circuit 1 in Case 2 would not be changed. Fig. 7.6 shows the measured fin surface temperature at the center of the three refrigerant circuits during defrosting in Case 1. It can be seen that all circuits’ temperatures reached 24°C at 217 s. From 80 to 210 s into defrosting, the temperature of Circuit 3 was always the highest. This may be mainly because the frost accumulation on the surface of Circuit 3 was the least among the three circuits. After 210 s into defrosting operation, the temperature curves’ order kept at TCircuit1 > TCircuit3 > TCircuit2, which agreed well with the order shown in Fig. 3.13 from 15 to 110 s. Fig. 7.7 presents the measured fin surface temperature at the center of the three refrigerant circuits during defrosting in Case 2. The duration of temperatures at the three circuits all reaching 24°C was 182, or 35 s less than that in Case 1. Therefore, the negative effects of uneven refrigerant distribution were further demonstrated. In addition, the same as that shown in Fig. 7.6, from 80 to 165 s into defrosting, the temperature of Circuit 3 was always the highest among the three circuits. Due to the refrigerant adjustments making it evenly distributed, this phenomenon might be because the

202

Defrosting for Air Source Heat Pump

frost accumulation on this circuit was obviously less, as listed in Table 7.1. Moreover, from Figs. 7.6 and 7.7, it could be concluded that the fin temperature was mainly affected by the frost accumulations on each circuit, especially when the refrigerant was evenly distributed. Fig. 7.8 presents the measured refrigerant volumetric flow rate during defrosting in two cases. At the start of the defrosting operation, the compressor discharge pressure increased suddenly, and the internal diameter of EEV was very small. Moreover, frost melting consumed a lot of energy, with a lot of refrigerant changing phases from the gas state to the liquid or two-phase state. This might be the reason why the measured refrigerant mass flow rate kept fluctuating severely from 0 to 70 s, especially during the first 40 s. When the defrosting process came into the water layer vaporizing stage described in Chapter 4, the compressor suction and discharge pressures increased, with the refrigerant volumetric flow rate changing from increasing to decreasing. As shown in Fig. 7.8, their peak values came out at 156 s into defrosting in Case 2 and 171 s in Case 1, respectively. This is another parameter to demonstrate the negative effects of uneven refrigerant distribution on defrosting performance for an ASHP unit with a vertically installed multicircuit outdoor coil. Fig. 7.9 shows the variation of the measured temperatures of the surrounding air and the melted frost collected in the water-collecting Cylinder C in two cases. Before the water was collected, the temperature of the surrounding air was measured. When the frost was melted and flowed into the water-collecting Cylinder C, the temperature of the melted frost was measured. Therefore, as shown in Fig. 7.9, the melted frost flowed into Cylinder C at 140 s into defrosting in Case 1, and 120 s in Case 2. It is about 20 s earlier in Case 2 than in Case 1. This phenomenon could also demonstrate the negative effects of uneven refrigerant distribution on system defrosting performance directly. It is obvious that the temperature of melted frost was very low, at around 0°C. Those temperature values measuring below 0°C are because of the measurement errors of the thermocouple, as listed in Table 3.5.

7.2.3 Energy analysis and discussions The energy used for RCD comes from three sources: the power input to the compressor, the power input to the indoor air fan, and the thermal energy from the indoor air. All the energy supplied is used to heat the outdoor coil metal, melt frost, heat the melted frost and the residual water, heat the cold ambient air, and evaporate the retained water on the surface of the outdoor coil. As shown in Table 7.2, the energy supply and consumption during RCD in two cases were calculated, with the calculated relative standard errors listed in Table 3.5. Here, the total energy used for defrosting was 648.9 kJ in Case 1, but 576.3 kJ in Case 2, or 11.2% less. The main difference comes from the energy from the indoor air, with a value of 72.6 kJ difference between two cases. Defrosting efficiency can be used to evaluate the performance of a defrosting operation. As described in Chapter 3, it is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil

The influence of refrigerant distribution on defrosting

203

Table 7.2 Defrosting efficiency calculations in the two cases Item

Parameter

Case 1

Case 2

Unit

1 2 3 4 5 6 7

The power input to compressor The power input to indoor air fan The energy from indoor air The power input to outdoor air fan Total energy supply for defrosting Energy consumption on melting frost Energy consumption on vaporizing the retained water Total energy consumption during defrosting Defrosting efficiency

111.2 7.6 648.9 0 767.7 317 26.9

104.7 5.1 576.3 0 686.1 315.3 39.1

kJ kJ kJ kJ kJ kJ kJ

343.9 44.8%

354.4 51.7%

kJ –

8 9

during an entire defrosting operation. In this section, the defrosting efficiencies calculated for the two cases were 44.8% and 51.7%, as listed in Table 7.2. Therefore, the negative effects of uneven distribution of refrigerant due to gravity and tube internal resistance on defrosting performance for vertically installed multicircuit outdoor coil were further demonstrated. Durations, defrosting efficiencies, and relative differences between the two cases are listed in Table 7.3. In this table, the duration of the tube surface temperatures at the exits of the three refrigerant circuits in the outdoor coil all reached 24°C. And the difference values are percentages, which are the ratios of the difference of Cases 1 and 2 to the value of Case 1. It is obvious that the difference of defrosting efficiencies in the two cases, 6.9%, is nearly the same as the difference of defrosting durations, 7.0%. Moreover, the duration difference of refrigerant volumetric flow rate reached its peak value in the two cases, 8.8%, approximates the difference of defrosting

Table 7.3 Duration, defrosting efficiency, and relative difference Item

Parameter

Case 1

Case 2

Difference

1 2

Defrosting duration Duration of all fin surface temperatures reaching 24°C Duration of refrigerant volumetric flow rate reaching its peak value Duration of melted frost reaching the water-collecting Cylinder C Total energy supply for defrosting Total energy consumption during defrosting Defrosting efficiency

185 s 217 s

172 s 182 s

7.0% 16.1%

171 s

156 s

8.8%

140 s

120 s

14.3%

767.7 kJ 343.9 kJ

686.1 kJ 354.4 kJ

10.6% 3.1%

44.8%

51.7%

6.9%

3 4 5 6 7

204

Defrosting for Air Source Heat Pump

efficiencies, too. However, the other durations’ differences are much bigger than the previous three differences, such as the duration of all fin surface temperatures reaching 24°C, the duration of the melted frost reached the water-collecting Cylinder C, and the total energy supply for defrosting. Therefore, it could be speculated that the defrosting duration and the duration of refrigerant volumetric flow rate reaching its peak value could be used to compare the defrosting performances of different system defrosting efficiencies. However, total energy consumption during defrosting could not be used as a parameter to evaluate the system defrosting performance because its difference value is 3.1%. In this section, the following conclusions could be reached. (1) An increase of 6.9% in defrosting efficiency if the refrigerant is evenly distributed into the three circuits was reported, as compared to the situation when all the stop valves were fully open. (2) The negative effects of uneven refrigerant distribution on system defrosting performance could be eliminated by adjusting the opening degrees of the stop valves, and thus the refrigerant flow into each circuit. (3) Besides the tube internal resistance, the refrigerant distribution should also be impacted by gravity. For an ASHP unit with a vertically installed multicircuit outdoor coil, the gravity impacts refrigerant distribution, and thus on system defrosting performance might be eliminated and comparatively studied by changing its placement method into horizontally installed. (4) To improve the defrosting efficiency for an ASHP unit with a multicircuit outdoor coil, the best refrigerant distribution plan may be not an even refrigerant distribution for each circuit, but distributing the refrigerant as the frost accumulates on each circuit. It means that more refrigerant should be distributed into the circuit on which frost accumulation is more.

7.3

The effect investigation of uneven refrigerant distribution and melted frost on uneven defrosting

Multiple parallel refrigerant circuits become commonly used for minimized pressure loss of flowing refrigerant and enhanced heat transfer efficiency for the refrigerant side of an outdoor coil used in an ASHP unit. For an ASHP unit having a vertically installed multicircuit outdoor coil, the phenomenon of uneven defrosting was widely found and reported. As indicated, the melted frost downward flowing (MFDF) due to gravity along the outdoor coil surface would have negative effects on system defrosting efficiency by prolonging the defrosting duration and increasing the energy consumption. To shorten the flowing path of melted frost, changing the vertically installed multicircuit outdoor coil into horizontally installed was carried out, indicating that the uneven defrosting phenomenon was avoided while the defrosting efficiency was improved. Furthermore, after the outdoor coil was horizontally installed, the residual water left on its downside surface due to surface tension also has negative effects on system defrosting performance. For a multicircuit outdoor coil in an ASHP unit, an uneven frosting phenomenon was found, and the FEC was also defined. By adjusting the refrigerant mass flow rate into each circuit, experimental studies on even frosting performance of an ASHP unit

The influence of refrigerant distribution on defrosting

205

were experimentally investigated, with results indicating that the system frosting COP improved. Furthermore, experimental studies demonstrated that improving the FEC of an RCD start could optimize the defrosting performance of an ASHP unit with a vertically installed three-circuit outdoor coil, no matter whether the downwardflowing melted frost was locally drained. In fact, refrigerant distribution not only affects its FEC in heating mode, but also results in an uneven defrosting phenomenon. The tube internal resistance and gravity both affect the refrigerant distributed into each circuit. In practical applications, it is impossible to make the refrigerant totally evenly distributed into each circuit for a vertically installed multicircuit outdoor coil in an ASHP unit. This may be the reason why studies on the URD effects due to tube internal resistance and gravity on RCD performance for an ASHP unit were scarce in the open literature. In the numerical study in Section 4.3, it is demonstrated that adjusting the refrigerant distributed into each circuit, at the ratio of 95.6%, 101.1%, and 103.3% for Circuits 1 to 3 in a vertically installed three-circuit outdoor coil, could alleviate the uneven defrosting phenomenon and thus improve defrosting performance. To fundamentally study the effects of URD on defrosting performance, as introduced in the previous section, another experimental study on system defrosting performance when the refrigerant was evenly and unevenly distributed into each circuit was carried out, with melted frost both locally drained, as shown in Figs. 7.10A and B. Experimental results demonstrated the negative effects of URD, and indicated that system defrosting efficiency was increased about 6.9% when the refrigerant was evenly distributed. However, in practical applications, there was always no water-collecting tray installed between circuits. Thus, the effects of URD and MFDF are always coupled, which is important and interactive. Also, the coupled effects of uneven heat supplying and viscous fluid disturbance on the heat and mass transfer performance for a multicircuit heat exchanger, with its two sides’ medium continuous changing phases, is a fundamental problem. Therefore, as a sequential study, an experimental investigation on system defrosting performance with refrigerant evenly or unevenly distributed into each circuit in an ASHP unit was carried out in this section, as shown in Figs. 7.10C and D. After comparative and quantitative analysis conducted using the experimental data, the coupled effects of MFDF and URD were finally given.

7.3.1 Experimental cases and RDEV adjustment In order to study the coupled effects of MFDF due to gravity and URD due to the tube internal resistance and gravity on system defrosting performance, a series of experimental works was carried out using the experimental ASHP unit. To obtain meaningful experimental results, at first it was necessary to ensure that the frost accumulated on the surface of the three circuits was even. Thus, a series of trial-and-error manual adjustments on the opening degrees of the stop valves work was conducted to adjust the FEC higher than 90%. Then, a set of fixed valve opening degrees was obtained, and the FECs of two cases could be kept the same. For easy reference in this work, this set of valves’ opening degrees was named State 1. Moreover, the FEC could be

206

Defrosting for Air Source Heat Pump

Circuit 3

R1 = R2 = R3

Refrigerant entrance (vapor)

(A)

(C)

R1

R2

Melted frost

Circuit 1

Melted Circuit 2 frost Circuit 3

R3

Melted R1 frost

Circuit 1

Melted R2 frost

Circuit 2

Melted R3 frost

Circuit 3

R1

(B)

R1 = R2 = R3

(D)

R1

R2

R2

Melted frost

R3

Circuit 1

Melted Circuit 2 frost

Refrigerant exit (liquid)

Melted R3 frost

Refrigerant entrance (vapor)

Circuit 2

Refrigerant entrance (vapor)

Melted R2 frost

Refrigerant exit (liquid)

Circuit 1

Refrigerant exit (liquid)

Refrigerant entrance (vapor)

Melted R1 frost

Refrigerant exit (liquid)

Water collecting tray

Water collecting tray

Circuit 3 R3 R1

R2

R3

Fig. 7.10 Four conditions of refrigerant distribution and melted frost flowing. (A) Refrigerant evenly distributed and melted frost locally drained. (B) Refrigerant unevenly distributed and melted frost locally drained. (C) Refrigerant evenly distributed and melted frost downwards flowing. (D) Refrigerant unevenly distributed and melted frost downwards flowing.

calculated, with water vaporized into the surrounding air neglected, by the masses of melted frost collected from three PVC cylinders in this work. Thereby, to comparatively study the coupled effects, the evenness of the refrigerant distributed into the three refrigerant circuits should be changed in different cases during defrosting as well as without any PVC trays installed between two circuits. Because the tube internal resistance and gravity directly affect the distribution of refrigerant, it seems hardly possible to adjust the refrigerant to make it evenly distributed into each circuit. However, according to each tube surface temperature at the circuit exit, these modulating valves located at an inlet refrigerant pipe to each circuit were used to vary the refrigerant flow into each circuit. Finally, as listed in Table 7.4, the experimental work was carried out on two experimental cases. With different refrigerant distribution evenness values (RDEVs) as well as melted frost downward flowing along the outdoor coil surface, the defrosting performances of the ASHP system can be comparatively and quantitatively analyzed. As with the previous section, the RDEV was defined as the ratio of the minimum refrigerant distribution into three circuits to the maximum one. In Case 1, the opening degrees of each stop valve were kept constant. It was ensured that all the stop valves

The influence of refrigerant distribution on defrosting

207

Table 7.4 Experimental conditions of two experimental cases Item

Parameter

Case 1

Case 2

1 2 3

FECs RDEVs Trays during frosting Trays during defrosting Stop valves’ state during frosting Stop valves’ state during defrosting Conditions shown in RDEV adjustment shown in Effects of URD

Higher than 90% Lower than 100% Without

Higher than 90% Nearly at 100% Without

Without

Without

State 1 (even frosting)

State 1 (even frosting)

State 2 (Fully open)

State 3 (Evenly adjusted)

Fig. 7.10D

Fig. 7.10C

Fig. 7.1

Fig. 7.2

Existence (at all stages of the defrosting process) Existence (only at the third stage of the defrosting process)

Inexistence (at any stage of the defrosting process) Existence (only at the third stage of the defrosting process)

4 5

6

7 8

9 10

Effects of MFDF

on the three circuits were fully open. Also, for easy reference in this study, the set of three stop valves fully open was named State 2. Due to gravity and tube internal resistance, the refrigerant flowing into each circuit was kept at a fixed RDEV lower than 100%. On the other hand, in Case 2, a series of trial-and-error manual adjustments of the opening degree of the stop valves was conducted. To adjust the refrigerant for even flow into each circuit, the tube surface temperature at the exit of each circuit was considered to guide the control strategy. Using this method, a suite of suitable degrees was obtained and fixed for three refrigerant circuits at the start of an RCD operation. Meanwhile, this set of valve opening degrees was named State 3. The refrigerant distribution condition could be fixed at an RDEV, nearly at 100%. For the negative effects of URD, it only exists at all stages of the defrosting process in Case 1. In two cases, the negative effects of MFDF only exist at the third stage of the defrosting process. At this stage described in Chapter 4, the frost was melting as well as the melted frost flowed away from a circuit. In order to obtain the opening degrees of the stop valves at State 3 in two cases during defrosting, the valves adjustment work was carried out by operating the ASHP unit in defrosting mode without any frost accumulating. With the heat transfer between the inside and outside of the tube eliminated, the refrigerant distribution could be directly reflected in the curve’s trends of tube surface temperatures at three

208

Defrosting for Air Source Heat Pump

refrigerant circuits. Meanwhile, when the stop valves are all fully open, the effects of gravity and tube internal resistance on refrigerant distribution for a vertically installed multicircuit outdoor coil could be found. This could be used for guiding to reach State 3 as well as analyzing the reason for the uneven frosting/defrosting phenomenon for an ASHP unit. The detailed method was introduced in the previous section.

7.3.2 Experimental results As shown in Table 7.5, the results of two experimental cases were listed, consisting of the total mass of melted frost collected, the FECs at the start of RCD, and two defrosting durations. The total masses of melted frost collected were nearly the same, at 875 g in Case 1 and at 872 g in Case 2. Their FECs were 91.7% in Case 1 and 90.6% in Case 2. Both of them were higher than 90%, with a difference of only 1.1%. Clearly, frost accumulations on the three circuits in the two cases were both close to each other. Their difference was smaller than 10%, which met the requirements described in the previous section. Also, Fig. 7.11 presents two photographs showing the airside surface conditions of the outdoor coil at the start of defrosting in the two cases. As observed Table 7.5 Experimental results of two experimental cases Item

Parameter

Case 1

Case 2

1 2 3 4

FECs Total mass of melted frost collected Defrosting durations Results shown in Figs. 7.11

91.7% 875 g 185 s A

90.6% 872 g 173 s B

Fig. 7.11 Airside surface conditions of the outdoor coil at the start of the defrosting operation. (A) Case 1 (without tray) and (B) Case 2 (without tray).

The influence of refrigerant distribution on defrosting

209

from Figs. 7.11A and B, the surface conditions at the start of defrosting for each circuit in the two cases were virtually the same, which agreed well with the data listed in Table 7.5. Therefore, a similar mass of frost accumulation, frost accumulated at the same FECs, both higher than 90%. Obviously, different RDEVs make this comparative study meaningful. Figs. 7.12–7.15 show the measured tube surface temperatures at the exits of the three refrigerant circuits and the measured fin surface temperatures at the centers 28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature ( C)

24 T1 > T2 > T3

20 T2 > T1 ; T2 > T3

16 12 8 4

181 s

185 s

0 80

100

120

140 Time (s)

160

180

200

Fig. 7.12 Measured tube surface temperatures during defrosting in Case 1. 28 Circuit 1

Circuit 2

Circuit 3

o

Tube surface temperature ( C)

24 20 16 T2

T 2 > T 3 >T 1

12

1

>T3

8 4

121 s

170 s

155 s

173 s

0 80

100

120

140 Time (s)

160

180

Fig. 7.13 Measured tube surface temperatures during defrosting in Case 2.

200

210

Defrosting for Air Source Heat Pump

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

Fin surface temperature ( C)

24 20 16 12

T 1 > T 2; T 3 > T 2

T 1 > T 3; T 2 > T 3

8

211 s 205 s

4

127 s

203 s

0 80

100

120

140 160 Time (s)

180

200

220

Fig. 7.14 Measured fin surface temperatures during defrosting in Case 1.

28 Fin in Circuit 1 Fin in Circuit 2 Fin in Circuit 3

o

Fin surface temperature ( C)

24 20

T1 = T2 > T3

16 T3 > T1 > T2

12 T3 > T2 > T1

8

129 s

4

193 s

168 s

200 s

0 80

100

120

140 160 Time (s)

180

200

220

Fig. 7.15 Measured fin surface temperatures during defrosting in Case 2.

of the three refrigerant circuits during defrosting in two cases. In all these figures, for their time (horizontal) axis, 80 s was the chosen starting time in order to clearly show the temperature rise during defrosting. As shown in Figs. 7.12 and 7.13, it is clear that the defrosting durations were 185 s for Case 1 and 173 s for Case 2. The 12 s earlier after refrigerant distribution evenly adjusted firstly shows that the defrosting performance was improved as the RDEV increased. At the same time, as shown in

The influence of refrigerant distribution on defrosting

211

Figs. 7.14 and 7.15, the durations of the fin surface temperature reaching 24°C were 211 s in Case 1 and 200 s in Case 2, with an 11 s difference. This also demonstrated the negative coupled effects of MFDF and URD on defrosting performance, when frost was evenly accumulated on each circuit. As presented in Fig. 7.12, from 80 to 140 s into defrosting, the temperature of Circuit 2 was kept the highest. This may be because the refrigerant distributed into Circuit 2 was more, or the frost accumulation on its surface was small. And then, after 140 s into defrosting, the temperature order was changed to at T1 > T2 > T3, which reflects the negative effects of MFDF. Finally, the tube surface temperatures at the exits of Circuits 2 and 3 reached 24°C at nearly the same time, at 181, 4 s earlier than the termination time of Circuit 3, 185 s. Also, as shown in Fig. 7.13, it is obvious that the three curves were not coincided, which is because the frost accumulations of the three circuits were not even. The same as in Case 1, the temperature of Circuit 2 was always kept the highest in Case 2. Consequently, the frost accumulation on Circuit 2 should be the least. From 80 to 121 s into defrosting, the temperature order was at T2 > T3 > T1, which reflects that the frost accumulated on Circuit 1 was the most. And then the order was changed to T2 > T1 > T3. The negative effects of MFDF also worked on it, with the temperature curves of Circuits 2 and 3 delayed. Also, the tube surface temperatures at the exits of Circuits 1 and 2 reached 24°C at nearly the same time, at 170, 3 s earlier than the defrosting termination, 173 s. Although there were some differences between each circuit on frost accumulation in the two cases, from 80 to 90 s into defrosting, the temperature differences between the three circuits were less than 2°C. This reflects that the frost accumulations on each circuit in the two cases were nearly the same, which meets the FECs listed in Table 7.5. After 90 s, there might be a bigger temperature difference between the circuits because the melted frost began downward flowing at this moment. As shown in Fig. 7.14, from 80 to 127 s, the temperature of Circuit 3 was the lowest. This shows the refrigerant distribution into this circuit was few, or the frost accumulation on its surface was more. And later, the temperature of Circuit 2 became the lowest. Considering the temperature curve’s trend of Circuit 2, it could be demonstrated that the refrigerant distributed into this circuit was the least. Finally, the durations of fin surface reached 24°C were 205 s for Circuit 1, 211 s for Circuit 2, and 203 s for Circuit 3, respectively. Considering of the negative effects of downwardflowing melted frost, the phenomenon that durations of Circuits 1 and 2 were shorter than that of Circuit 3 reflects the refrigerant distributed into Circuit 3 was the most. However, as shown in Fig. 7.15, it is obvious that the durations of the fin surface temperatures reaching 24°C were 193 s for Circuits 1 and 2 and 200 s for Circuit 3. Because the RDEV was 100%, the negative effects should make the temperature order T1 > T2 > T3. However, it was at T1 ¼ T2 > T3, different from that expected. The reason is that the frost that accumulated on Circuit 2 was less than the average value of the three circuits. From 80 to 90 s into defrosting, the order of the three temperature curves was at T3 > T2 > T1, which directly reflects the order of frost accumulation on the three circuits that was at T1 > T2 > T3. It was the negative effects of MFDF that made the curve of Circuit 3 the lowest after 168 s, which also reflects on the temperature

212

Defrosting for Air Source Heat Pump

order changing at the period of 129–168 s. It is obvious that there was suddenly a decrease in Circuit 2, which resulted from the melted frost flowing onto the thermocouple. Fig. 7.16 presents the measured temperatures of the tube surface at the outdoor coil entrance and exit during defrosting in two cases. Clearly, with a lower outdoor coil exit temperature and a higher outdoor coil entrance temperature in Case 2, the defrosting performance was improved after the refrigerant was evenly distributed into each circuit. Fig. 7.17 shows the temperature difference of the outdoor coil entrance and exit (TDOEE) during defrosting in the two cases. The maximum values were 42.8°C at 110 s into defrosting in Case 1, and 38.0°C at 105 s in Case 2. Obviously, the Δ T2, max was smaller than the Δ T1,max, and the former came out 5 s earlier than the latter. In addition, during defrosting, Δ T2 was always lower than Δ T1. This also reflects that the defrosting performance could be improved by evenly adjusting the refrigerant distributed into each circuit for an ASHP unit, with MFDF along the surface of its multicircuit outdoor coil. Fig. 7.18 presents the measured temperatures of the tube surface at the indoor coil entrance and exit during defrosting in two cases. The same as that shown in Fig. 7.16, with a lower indoor coil entrance temperature and a higher indoor coil exit temperature in Case 2, the defrosting performance was improved after the refrigerant was evenly distributed into each circuit. From 30 to 85 s, it is a short period of fluctuation for the temperature curves. This is because a lot of energy was consumed on frost melting at this period, as demonstrated in Chapter 4. However, at the same time, the energy transferred from the indoor air thermal energy and the compressor and air fan inputs could not cover this part of the energy consumption. Therefore, some of the energy stored in the metal of the indoor coil was taken away, and the tube surface temperature at its exit decreased to about 20°C, as shown in Fig. 7.18. 50

Entrance in Case 1 Exit in Case 1 Entrance in Case 2 Exit in Case 2

T Exit, 1 > TExit, 2

30

o

Temperature ( C)

40

20

10 T Entr, 2 > TEntr, 1

0 0

20

40

60

80

100 120 Time (s)

140

160

180

200

Fig. 7.16 Measured temperatures of the tube surface at the outdoor coil entrance and exit during defrosting.

The influence of refrigerant distribution on defrosting

213

50 Case 1 Case 2

45

o

ΔT1, max= 42.8 C

o

Temperature difference ( C)

40 35

ΔT1 > ΔT2

o

ΔT2, max= 38.0 C

30 25 20 15

105 s

10

110 s 5 0

20

40

60

80

100 120 Time (s)

140

160

180

200

Fig. 7.17 Temperature difference of the outdoor coil entrance and exit during defrosting.

40 Entrance in Case 1 Exit in Case 1 Entrance in Case 2 Exit in Case 2

30 T Entr, 1 > T Entr, 2

o

Temperature ( C)

20 10 0

TExit, 2 > TExit, 1 Fluctuation

–10

85 s

30 s –20 0

20

40

60

80

100 120 Time (s)

140

160

180

200

Fig. 7.18 Measured temperatures of the tube surface at the indoor coil entrance and exit during defrosting.

Fig. 7.19 shows the temperature difference of the indoor coil entrance and exit (TDIEE) during defrosting in the two cases. The maximum values were 43.2°C at 25 s into defrosting in Case 1, and 40.9°C at 30 s in Case 2. Obviously, the Δ T2, max was smaller than Δ T1, max, and the latter came out 5 s earlier than the former. In addition, during defrosting, Δ T2 was always lower than Δ T1, the same as that shown in Fig. 7.18. This further reflects that the defrosting performance could be

214

Defrosting for Air Source Heat Pump

45

o

T 1, max = 43.2 C

40

Case 1 Case 2

o

T 2, max = 40.9 C

o

Temperature difference ( C)

35 30 25 20 T1 > T2

15 10 5

25 s

0

30 s

–5 0

25

50

75

100 125 Time (s)

150

175

200

Fig. 7.19 Temperature difference of the indoor coil at the entrance and exit during defrosting.

improved by evenly adjusting the refrigerant distributed into each circuit for an ASHP unit, with MFDF along the surface of its multicircuit outdoor coil. Fig. 7.20 presents the measured refrigerant volumetric flow rate during RCD in two experimental cases. As shown, the compressor discharge pressure suddenly increased at the start of the defrosting operation. First, the internal diameter of the electronic expansion valve (EEV) was very small. Second, the system operation at the frost 1.6 Refrigerant volumetric flow rate (L / min)

Case 1

Case 2

1.4 1.2 1.0 0.8 0.6 0.4 R2 > R1

Fluctuation

0.2

74 s

152 s

Fluctuation 160 s

0.0 0

20

40

60

80

100 120 Time (s)

140

160

180

200

Fig. 7.20 Measured refrigerant volumetric flow rate during defrosting in two cases.

The influence of refrigerant distribution on defrosting

215

melting stage consumed a lot of energy, with a lot of refrigerant changing phases from gas to liquid or a two-phase state. This might be the reason why the measured refrigerant mass flow rate kept fluctuating severely from 0 to 74 s, especially during the first 50 s into defrosting. When the defrosting process came into the water layer vaporizing stage described in Chapter 4, the compressor suction and discharge pressures increased, with the refrigerant volumetric flow rate changing from increasing to decreasing. As shown in Fig. 7.20, their peak values nearly both came out at 160 s into defrosting in the two cases. From 74 to 152 s, the refrigerant volumetric flow rate order of the two cases was at R2 > R1. The peak value in Case 2 was slightly lower than that in Case 1. It is demonstrated that the defrosting operation in Case 2 was terminated a little quicker. For an ASHP unit with a vertically installed multicircuit outdoor coil, this is more proof to demonstrate the negative coupled effects of MFDF and URD on defrosting performance. Fig. 7.21 shows the variation of the measured temperatures of the surrounding air and the melted frost collected in the water-collecting Cylinder C in the two cases. Because there were no water-collecting trays installed between circuits, the melted frost collected in this cylinder was much more than that collected with trays installed between circuits, at 875 g for Case 1 and 872 g for Case 2, as listed in Table 7.4. Clearly, there were two suddenly decreasing points that existed in the two curves, at 120 s into defrosting for Case 1 and at 115 s into defrosting for Case 2. This means the frost was melted and flowed into the water-collecting Cylinder C, and the temperature of the melted frost was measured. Before the water was collected, during 0–120 s for Case 1 and 0–115 s for Case 2, the temperature of the coil surrounding the air was measured. It was about 5 s earlier in Case 2 than that in Case 1. This phenomenon

7 Case 1 Case 2

6

o

Temperature ( C)

5 4 T1 > T2

3

T2 > T1 T1

2 1

26 s 0

40

T2 > T1

T2

60 s 80

90 s 120 160 Time (s)

200

240

280

Fig. 7.21 Measured temperatures of the surrounding air and melted frost collected during defrosting.

216

Defrosting for Air Source Heat Pump

7 Case 1 Case 2

6

o

Temperature ( C)

5 4 T1 > T2

3

T2 > T1 T1

2 1

26 s 0

40

T2 > T1

T2

60 s 80

90 s 120 160 Time (s)

200

240

280

Fig. 7.22 Measured mean temperature of the air surrounding each circuit during defrosting.

could also directly demonstrate the negative coupled effects of MFDF and URD on system defrosting performance. It is obvious that the temperature of the melted frost was very low, at around 0.4°C. Before 148 s into defrosting, the temperature of the melted frost in Case 2 was lower than that in Case 1. This reflects that the melted frost took less heat from the coil in Case 2. After 148 s, the temperature order was changed to T2 > T1. This shows that the defrosting operation terminated earlier in Case 2, and thus the melted frost was heated by the surrounding air a lot. This phenomenon of defrosting operation terminating earlier in Case 2 is also shown in Fig. 7.22; from 26 s to the end of defrosting, the temperature of the air surrounding each circuit in Case 2 was always higher than that in Case 1. Before 90 s into defrosting, the difference between the two cases was very small, especially at 60 s into the defrosting operation. The temperatures were the same for the two cases because at the period of 0–90 s, the melted frost was melting without downward flowing into the down-circuits, and a few areas of the fin and coil contacted the surrounding air. This figure also shows that the defrosting performance could be improved by evenly adjusting the refrigerant distribution when frost evenly forms on the surface of each circuit for an ASHP unit.

7.3.3 Energy analysis and discussions During an RCD operation, the energy mainly comes from four sources: the thermal energy from the indoor air, the power input to the compressor, the power input to the indoor air fan, and the heat stored in the metal of the indoor coil. The last one is small, and thus always neglected. All the energy supplied is used to melt frost, heat the outdoor coil metal, heat the melted frost and the residual water, evaporate the

The influence of refrigerant distribution on defrosting

217

retained water on the surface of the outdoor coil, and heat the cold ambient air. Also, only the energy consumed in frost melting and water vaporing are big enough and calculated, with the sensible heat of the melted frost and surrounding air taken away from the outdoor coil neglected. As listed in Table 7.6, the energy supply and consumption during defrosting in the two cases were summarized, with their defrosting efficiencies calculated. As calculated, the total energy used for defrosting was 790.5 kJ in Case 1, but 691.6 kJ in Case 2, or 12.5% less. It could be found that in this experimental study, the main difference came from the thermal energy from the indoor air, with a value of 95 kJ difference between the two cases. Defrosting efficiency is always used to evaluate the performance of a defrosting operation for an ASHP unit. It is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation. In this section, the defrosting efficiencies calculated for the two cases were 40.5% for Case 1 and 47.9% for Case 2, as listed in Table 7.6. Therefore, the negative coupled effects of MFDF and URD on defrosting performance for a vertically installed multicircuit outdoor coil were further quantitatively demonstrated. Furthermore, the different durations, defrosting efficiencies, and relative differences in the two experimental cases were also calculated and summarized in Table 7.7. The temperature difference values of Items 4, 6, and 9 were directly presented. But the difference values of Items 1–3, 5, 7, 8, and 10–12 were shown with their ratios. The same as the previous section, they were also defined as the ratios of the difference of Cases 1 and 2 to the value of Case 1, and the results are percentages. Obviously, the difference of total energy supply in the two cases, 12.5%, was much bigger than the differences of the tube and fin defrosting durations, 6.5% and 5.2%. Also, the duration difference of TDOEE reached its peak value at 4.5%, and the duration difference of the melted frost collected was 4.2%. However, because the defrosting performance was better in Case 2, the defrosting efficiency and the duration difference of TDIEE reaching peak values and total energy consumption during defrosting were negative values, at 18.3%, 20% and 3.6%, respectively. Table 7.6 Energy supply, energy consumption, and defrosting efficiency calculation Item

Parameter

Case 1

Case 2

Unit

1 2 3 4 5 6 7

The energy from indoor air The power input to compressor The power input to indoor air fan The power input to outdoor air fan Total energy supply for defrosting Energy consumption on melting frost Energy consumption on vaporizing retained water Total energy consumption during defrosting Defrosting efficiency

674.8 108.9 6.8 0 790.5 292.3 27.6

579.8 106.1 5.7 0 691.6 291.2 40.3

kJ kJ kJ kJ kJ kJ kJ

319.9 40.5%

331.5 47.9%

kJ –

8 9

218

Defrosting for Air Source Heat Pump

Table 7.7 Different durations, defrosting efficiencies, and relative differences in two experimental cases Item

Parameter

Case 1

Case 2

Difference

1

Duration of tube surface temperature of Circuit 3 reaching 24°C Duration of fin surface temperature of 3 circuits all reaching 24°C Duration of TDOEE reaching its peak value Peak value of TDOEE Duration of TDIEE reaching its peak value Peak value of TDIEE Duration of refrigerant volumetric flow rate reaching its peak value Duration of melted frost reaching the water collecting cylinder Temperature of the lowest temperature of melted frost collected Total energy supply for defrosting Total energy consumption during defrosting Defrosting efficiency

185 s

173 s

6.5%

211 s

200 s

5.2%

110 s

105 s

4.5%

42.8°C 25 s

38.0°C 30 s

(4.8) 20%

43.2°C 160 s

40.9°C 160 s

(2.3) 0

120 s

115 s

4.2%

0.45°C

0.2°C

(0.25)

790.5 kJ 319.9 kJ

691.6 kJ 331.5 kJ

12.5% 3.6%

40.5%

47.9%

18.3%

2 3 4 5 6 7 8 9 10 11 12

Therefore, it could be speculated that the defrosting durations of the tube and fin could be used to compare the defrosting performances for systems with different defrosting efficiencies. In addition, for Items 4, 6, and 9, temperature differences were small, especially the value of melted frost at just 0.25°C. However, the peak value of the TDOEE difference, 4.8°C, as well as their durations’ difference, 4.5%, could also be considered comprehensively as one of the system defrosting performance evaluation parameters. In conclusion, for an ASHP unit with a vertically installed three-circuit outdoor coil, a comparative experimental study on system RCD performance when the refrigerant is evenly distributed into each circuit or not was undertaken, as well as the melted frost downward flowing during defrosting. After the study results were analyzed, the following conclusions could be drawn: (1) For an ASHP unit with a multicircuit outdoor coil, the refrigerant distributed into each circuit would be affected by the tube internal resistance and gravity during RCD. And the negative coupled effects of MFDF and URD on the system defrosting performance were qualitatively confirmed. (2) After the distribution of refrigerant was evenly adjusted, the duration of the tube surface temperature of Circuit 3 reaching 24°C decreased from 185 to 173 s, or a ratio of 6.5%. Meanwhile, the duration of the fin surface temperature of three circuits all reaching 24°C decreased from 211 to 200 s, or a ratio of 5.2%. (3) As compared with the refrigerant being unevenly distributed into each circuit, for an ASHP unit with a vertically installed multicircuit outdoor coil, the total energy

The influence of refrigerant distribution on defrosting

219

supply for defrosting could save 98.9 kJ, and the defrosting efficiency could be improved by 7.4% (from 40.5% to 47.9%), or at a ratio of 18.3%, when the refrigerant is evenly distributed. (4) As analyzed, it could be speculated that the defrosting durations of the tube and fin could be used to compare the defrosting performances for different defrosting systems. The peak value of TDOEE difference as well as their duration difference could also be considered comprehensively as a defrosting evaluation parameter.

7.4

Discussion on the effect of melted frost

To analyze the difference between with and without melted frost downward flowing during defrosting, as well as adjusting the refrigeration distribution into each circuit, the experimental results in four cases in previous sections were summarized in Table 7.8. In Cases 1 and 2, when the water-collecting trays were not installed, the defrosting durations were 185 s in Case 1 and 173 s in Case 2, respectively. However, in Cases 3 and 4, the trays were installed, and the defrosting durations were 185 and 172 s, respectively. It seems after the melted frost was taken away, the defrosting duration was not shortened. It is because the frost accumulations for the four cases were different. The accumulations in Cases 3 and 4 were more than those in Cases 1 and 2.

Table 7.8 Experimental cases designed in this study Item

Parameter

Case 1

Case 2

Case 3

Case 4

1 2 3

FEC Water-collecting trays Valves status at frosting mode Valves status at defrosting mode Duration of tube surface temperature of Circuit 3 reaching 24°C Duration of fin surface temperature of three circuits all reaching 24°C Total energy supply for defrosting Total energy consumption during defrosting Defrosting efficiency

>90% Without Even frosting Evenly adjusted 185 s

>90% Without Even frosting Fully open 172 s

>90% With Even frosting Evenly adjusted 185 s

>90% With Even frosting Fully open 173 s

217 s

182 s

211 s

200 s

767.7 kJ

686.1 kJ

790.5 kJ

691.6 kJ

343.9 kJ

354.4 kJ

319.9 kJ

331.5 kJ

44.8%

51.7%

40.5%

47.9%

4 5

6

7 8 9

220

Defrosting for Air Source Heat Pump

Therefore, the negative effects of DFMF were still shown here. Additionally, after the refrigerant was evenly adjusted in Cases 2 and 4, the defrosting durations were obviously shortened. Of course, the prerequisite is that the FECs in the four cases were all higher than 90%. The durations for all fin surface temperature reaching 24°C were 211 s in Case 1 and 200 s in Case 2, respectively. Although the trays were installed between circuits in Cases 3 and 4, the most delayed circuit for them is the same as that in Cases 1 and 2, at Circuit 2 and Circuit 2, respectively. The durations for all fin surface temperatures reaching 24°C were 217 s in Case 3 and 182 s in Case 4, respectively. Here, the effects of DFMF were not obvious, but the effects of URD were. As seen, the two durations in Cases 2 and 4 were much shorter than those in Cases 1 and 3, due to the refrigerant distribution being evenly adjusted during defrosting. As seen in Table 7.8, the total energy supply is 790.5 kJ in Case 1, 691.6 kJ in Case 2, 767.7 kJ in Case 3, and 686.1 kJ in Case 4, respectively. Clearly, after the melted frost was taken away during defrosting in Cases 3 and 4, the corresponding energy supplies were both reduced compared with those in Cases 1 and 2. Also, after the refrigerant was evenly adjusted during defrosting, the energy supplies were also reduced in Cases 2 and 4 compared with those in Cases 1 and 3. They both reflect the effects of eliminating the DFMT and URD. However, when it changes to energy consumption, the relations of the values in the four cases are totally different. As seen, the total energy consumption during defrosting is 319.9 kJ in Case 1, 331.5 kJ in Case 2, 343.9 kJ in Case 3, and 354.4 kJ in Case 4, respectively. Clearly, after the melted frost was taken away during defrosting in Cases 3 and 4, the corresponding energy consumptions were both increased compared with those in Cases 1 and 2. Also, after the refrigerant was evenly adjusted during defrosting, the energy consumptions were also increased in Cases 2 and 4, compared with those in Cases 1 and 3. But, they both reflect the effects of eliminating the DFMT and URD. Finally, the defrosting efficiencies were 40.5% in Case 1, 47.9% in Case 2, 44.8% in Case 3, and 51.7% in Case 4, respectively. The difference is 4.3% between Cases 1 and 3, and 3.6% between Cases 2 and 4. That means that after the melted frost was taken away during defrosting, the defrosting efficiency could be improved at least 3.6%–4.3%, no matter whether the refrigerant was evenly distributed. At the same time, the difference is 7.4% between Cases 1 and 2, and 6.9% between Cases 3 and 4. That means that after the refrigerant as evenly distributed, the defrosting efficiency could be improved around 6.9%–7.4%, with the DFMT locally drained or not. Therefore, the effect of ERD seems higher than the effect of DFMT.

7.5

Concluding remarks

In this chapter, experimental studies on refrigerant distribution during defrosting are presented, and the following conclusions may be drawn. (1) Tube internal resistance and gravity would affect the refrigerant distribution into each circuit for an ASHP unit with a multicircuit outdoor coil during RCD. The negative effects of uneven refrigerant distribution on system defrosting performance when the frost was evenly accumulated on the surface of each circuit at the start of a defrosting operation could be

The influence of refrigerant distribution on defrosting

221

reflected by a few different parameters, such as the duration of all fin surface temperatures, the duration for the refrigerant volumetric flow rate to reach its peak value, etc.; (2) The negative effects of uneven refrigerant distribution on system defrosting performance could be eliminated by adjusting the opening degrees of the stop valves, and thus the refrigerant flow into each circuit. The tube surface temperature at the exit of each circuit could be used as a control signal to regulate the valve opening degrees. To ensure even refrigerant distribution into each circuit during defrosting, the opening degrees of the stop valves should be adjusted first before the start of a defrosting operation. It could be reached by working at the defrosting mode under the condition of without any frost accumulation on the surface of the ASHP unit’s outdoor coil. (3) The mechanism of refrigerant distribution affected by the gravity is a fundamental problem for the research work on an ASHP unit with a vertically installed multicircuit outdoor coil. During defrosting, refrigerant density would be changeable as its state changes from gas to liquid or a two-phase state. Therefore, besides the tube internal resistance, the refrigerant distribution should also be impacted by gravity. For an ASHP unit with a vertically installed multicircuit outdoor coil, the gravity impacts refrigerant distribution, and thus on system defrosting performance might be eliminated and comparatively studied by changing its installation from vertical to horizontal. (4) For a multicircuit outdoor coil, the frost accumulations on the surface of each circuit are hardly the same. To improve the defrosting efficiency for an ASHP unit with a multicircuit outdoor coil, the best refrigerant distribution plan may be not even for each circuit, but according to the frost accumulation on each circuit. This means that more refrigerant should be supplied to the circuit with more frost accumulation. This kind of defrosting based on demand could be categorized as an intelligent defrosting method.

References [1] O’Neal DL, Peterson KT, Anand NK, Schliesing JS. Refrigeration system dynamics during the reverse cycle defrost. ASHRAE Trans 1989;95(2):689–98. [2] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil in an air source heat pump-Part I: experiments. Appl Energy 2012;91:122–9. [3] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil in an air source heat pump-Part II: modeling analysis. Appl Energy 2012;91:274–80.

Energy transfer during defrosting 8.1

8

Introduction

In recent years, various frost-retarding measures have been investigated to improve the operating performance of an ASHP unit. While the use of frost-retarding measures can only delay frost formation or growth, such measures are always expensive or consume additional energy, and there is still frost that must be removed even after applying the measures. Periodic defrosting therefore becomes necessary for guaranteeing the satisfactory operation of an ASHP unit. As mentioned in Chapter 2, defrosting may be realized by a number of methods, and the most widely used standard defrosting method for ASHP units is reverse cycle defrosting (RCD). As shown in Fig. 8.1, when an ASHP unit changes from heating mode to RCD mode, its outdoor coil changes to act as a condenser and its indoor coil as an evaporator. The energy that should have been used for space heating is consumed to melt frost and vaporize melted frost. Not only is the indoor space heating interrupted, but also the indoor thermal comfort level may be adversely affected. Usually, the ambient air temperature is low at night, therefore, sleep thermal comfort can be degraded due to frequent defrosting operations of an ASHP unit when it is used for space heating in a sleeping environment. As a transient and nonlinear heat and mass transfer process, the energy transfer during defrosting directly affects defrosting performance. Hence, to improve the defrosting performance of ASHP units, many experimental studies have been carried out and reported, such as (1) changing the installation method of the outdoor coil, (2) adjusting the refrigerant distribution, (3) eliminating uneven defrosting, (4) improving FECs, and (5) applying PCM-TES to defrosting. Among these studies, two indexes—defrosting duration and defrosting efficiency—were simply used to evaluate the defrosting performance. However, the dynamic and complicated energy transfer process has not been studied in great detail. It is easy to understand that the energy transfer process can be evaluated using a numerical method. Notably, Cole built a defrosting model for large commercial freezers, and reported the heat and mass transfer and fluid flow mechanisms. The resultant refrigeration loads due to defrosting were theoretically estimated [1]. Thereafter, Krakow et al. reported an idealized RCD model for an evaporator [2, 3]. Two evaluation indexes, system performance coefficient and defrosting efficiency (presented as coil efficiency), were separately defined. Furthermore, the fact that the refrigerant stored in a receiver for a transient system was acting as an energy source during defrosting was demonstrated [3]. Although many follow-up modeling studies were reported, attention was most paid to the system performance improvement by refrigerant distribution [4, 5] or component optimization [6, 7]. A detailed energy transfer process during a single RCD operation is not given. A previous experimental study tried to analyze the defrosting heat supplies and energy consumption during defrosting for an ASHP Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00008-4 © 2019 Elsevier Ltd. All rights reserved.

224

Defrosting for Air Source Heat Pump

Fig. 8.1 Energy transfer process during (A) heating mode and (B) defrosting mode.

unit [8]. It was indicated that the heat supply from the indoor air contributed to 71.8% of the total heat supplied for defrosting while 59.4% of the supplied energy was used for melting frost. In addition, the condition of even frosting at the start of a defrosting operation was neglected. Obviously, to optimize an ASHP system, the dynamic energy transfer during defrosting should be further comprehensively studied. During RCD, as shown in Fig. 8.1, the energy stored in the indoor and outdoor coil metals (MES), Ei, MESand Eo, MES, changes with the fluctuation of coil temperature. Although it has been reported that the energy consumption for heating the outdoor coil metal accounts for 16.5% of the total heat supplied [4], this part of the energy has always been neglected [7]. Hence, MES was first considered [9, 10] earlier when studying the effects of the downward flowing of melted frost. As a fundamental problem, the MES effects on defrosting performance for an ASHP unit are still not clear. Also, when the MES is changed, the energy transfer process in an ASHP unit, as expressed by the two equations shown in Fig. 8.1, should be qualitatively and quantitatively studied. The condition of melted frost being locally drained away should also be considered.

8.2

Energy transfer process during defrosting

Understanding the energy transfer process and the MES effect on defrosting performance are of importance for ASHP units’ application, but studies are scarce in the open literature. An experimental investigation on the energy transfer process in an ASHP unit and the effect of MES during its RCD has been carried out and a comparative and quantitative analysis was conducted using the experimental data. A tailor-made three-circuit outdoor coil is first introduced and two settings of the two-working-circuit and three-working-circuit cases are designed. Then, a description of the experimental setup and experimental procedures are given. This is followed by reporting the experimental results. Lastly, the evaluations and discussions on defrosting performance for this experimental ASHP unit are presented, followed by a conclusion.

Energy transfer during defrosting

225

8.2.1 Methodology To qualitatively and quantitatively study the energy transfer process in an ASHP unit and the effect of MES on RCD, a series of experiments should be carried out. Basing on these experimental results, all types of energy supplies and consumptions during defrosting could be calculated. First, an ASHP unit was selected and a three-circuit outdoor coil was specially made. And then, two typical experimental conditions were designed, with two-working-circuit and three-working-circuit outdoor coils used, respectively. To avoid the uneven frosting influence, frost was adjusted to be evenly accumulated on their surfaces during frosting. The FEC was kept at higher than 90%, which was defined as the ratio of the minimum frost accumulation among three circuits to the maximum one, and calculated by the melted frost collected from the watercollecting cylinders, with the water vaporized into the ambient air neglected. The tube surface temperature worked as the controlling index of frost accumulation distribution during frosting. Energy supplies and consumptions in different fields were calculated, including power consumptions for the compressor, the indoor and outdoor air fans, etc. Finally, the defrosting performance was evaluated by the two following indexes, (1) Defrosting efficiency,

ηd ¼

Qm + Qv  100% Ecomp + Ei, fan + Qi, air

(8.1)

and (2) MES effect on defrosting,

ηm ¼

Qi, MES  Qo, MES  100% Ecomp + Ei, fan + Qi, air

(8.2)

in which Qm is the energy consumed on melting the accumulated frost, and Qv on vaporizing the retained melted frost. Ecomp and Ei, fan are the power inputs to the compressor and indoor air fan during defrosting, respectively. Qi, fan is the thermal energy supply from the indoor air. Qi, MES and Qo, MES are the MES values of the indoor and outdoor coils, respectively. All the mentioned parameters are shown in Fig. 8.1. (1) Experimental setup

In this section, the experimental setup used is the same as that reported in Chapter 3. So, it is briefly introduced here. As shown in Fig. 8.2, the experimental ASHP unit was installed in an existing environmental chamber having a simulated heated indoor space (left) and a simulated outdoor frosting space (right). In each space, sensible and latent LGUs were installed and used to simulate thermal loads. The experimental conditions were jointly maintained by using a separate DX A/C system in the environmental chamber and the two LGUs. During experiments, to slightly adjust the latent load, two humidifiers were added in the outdoor frosting space. During normal heating (or frosting) operation, a frosting environment in the outdoor space was

226

Defrosting for Air Source Heat Pump

Fig. 8.2 Schematics of the ASHP unit installed in an environmental chamber.

maintained by running the experimental ASHP unit, humidifiers, and LGUs together while an indoor heated environment by the experimental ASHP unit and the existing DX A/C system. The outdoor coil was vertically installed, and in each circuit a solenoid modulating valve and a manual stop valve were used. There were three individual and parallel refrigerant circuits and the airside surface areas corresponding to each of the three circuits were the same. The indoor coil was original in the commercially ASHP unit, also totally consisting of three circuits. The number of tube rows in the two circuits was 6  2, and the other one 4  2. The specifications of the three-circuit outdoor coil and the original indoor coil are shown in Tables 8.1 and 8.2, respectively. The metal mass of each circuit in the two coils was calculated and listed in Table 8.3. For easy referencing in this paper, the tray installed under Circuit 1 was named Tray A, under Circuit 2 Tray B, and under Circuit 3 Tray C. As shown in Fig. 8.2, their connecting water-collecting cylinders are named Cylinder A, B, and C, respectively. As listed in Table 8.1, the volumes of water-collecting Cylinders A, B, and C were the same at 500 mL. When the water-collecting trays between circuits were

Energy transfer during defrosting

227

Table 8.1 Specifications of the tailor-made three-circuit outdoor coil Item

Parameter

Value

Unit

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Height of the outdoor coil Width of the outdoor coil Thickness of the outdoor coil Fin height Fin width Fin thickness Fin pitch Fin type Tube external diameter Tube thickness Tube spacing Circuit pitch Number of tube rows Number of circuits Number of water-collecting trays Number of water-collecting cylinders Number of wind boards Material of tube Material of fin Material of wind board Material of water-collecting tray Material of water-collecting cylinder Volume of cylinder A, B, and C Volume of cylinder D

500 590 44 152.4 44 0.115 2.1 Plate 10 0.5 20 22 2 3 3 3 2 Copper Aluminum Wood PVC PVC 500 2000

mm mm mm mm mm mm mm – mm mm mm mm – – – – – – – – mL mL

Table 8.2 Specifications of the prototypical three-circuit indoor coil Item

Parameter

Value

Unit

1 2

Number of circuits Number of tube rows in each circuit Tube spacing Length of indoor coil Fin type Fin length Fin width Fin pitch Fin thickness Tube external diameter Material of fin Material of tube

3 Circuits 1, 2: 6  2; Circuit 3: 42 5 103 Wavy Circuits 1, 2: 125; Circuit 3: 85 25 1.3 0.08 7 Aluminum Copper

/ /

3 4 5 6 7 8 9 10 11 12

mm cm / mm mm mm mm mm / /

228

Defrosting for Air Source Heat Pump

Table 8.3 Calculated metal mass of indoor coil and outdoor coil Item

Parameter

Value

Unit

1 2 3 4 5 6

Tube mass of indoor coil Tube mass of each circuit of outdoor coil Fin mass of indoor coil Fin mass of each circuit of outdoor coil Total mass of indoor coil Total mass of each circuit of outdoor coil

2116 533 380 477 2496 1010

g g g g g g

Water collecting tray

Circuit 1 Refrigerant entrance (Vapor)

R1 Circuit 2

Refrigerant exit (Liquid)

Melted frost

R2 Circuit 3 R3

Air temperature sensor

Fig. 8.3 Schematics of the outdoor coil and locations of air temperature sensors.

removed during defrosting, only the 2000 mL Cylinder D was placed, connecting with water-collecting Tray C. >As shown in Fig. 8.3, three water-collecting trays made of PVC were placed under each circuit of the outdoor unit. They can be taken away when they are unnecessary. Three water-collecting cylinders made of PVC were connected to these trays, so that the melted frost from each circuit during defrosting could be collected and weighed. After defrosting was terminated, the retained water on the fins was absorbed by preweighed cotton tissues. And thus, the total mass of melted frost, mm, could be obtained. In this section, the total mass of frost accumulation during the experiment, mf, could be calculated with the density of air in the outdoor frosting space and the volumetric flow rate of air passing through the outdoor coil, as shown in the following equations. Z mf ¼

tf

Δmf  dt ¼

X Δmf  Δt

(8.3)

0

Δmf ¼ ρo, a  Vo, a  Δt  ðωo, o  ωo, i Þ

(8.4)

In Eqs. (8.3), (8.4), ρo, a and Vo, a are the density and volume of outdoor air, ωo, o and ωo, i the moisture content of air at the outlet and inlet of the outdoor coil. Δt and tf are the measuring time interval and frosting duration, respectively.

Energy transfer during defrosting

229

Then, the total mass of vaporized water, mv, was expressed by, mv ¼ mf  mm ¼ mf  mcf  mrw

(8.5)

in which mcf is the total mass of the melted frost collected in the cylinders, and mrw the total mass of retained water. In Eqs. (8.1), (8.2), the energy consumed on melting frost and vaporizing retained water, Qm and Qv, were evaluated by: Qm ¼ mf  Lsf

(8.6)

Qv ¼ mv  Lv

(8.7)

where mf and mv are the total mass of the frost accumulated over the outdoor coil surface and the total mass of vaporized water, and Lsf and Lv the latent heat of frost melting and water vaporization, respectively. Air dry-bulb temperatures upstream of the outdoor coil were measured at six points using precalibrated K-type thermocouples and air wet-bulb temperatures were also measured at six points using temperature sensors (PT100, class A), as shown in Fig. 8.4. The average values from these measurements were used as the inlet air dry-bulb temperature and the wet-bulb temperature in the follow-up calculation. On the other hand, the air flow rate passing through the outdoor coil was measured

Circuit 1

Circuit 1 m f,1

Circuit 2

m f,1 Circuit 2

m f,2

m f,2

Circuit 3 m f,3

Frosting in Case 1

Circuit 1

Defrosting in Case 1

Circuit 1 m f,1

m f,1 Circuit 2

Circuit 2 m f,2

Circuit 3

m f,2 Circuit 3

m f,3

Frosting in Case 2

m f,3

Defrosting in Case 2

Fig. 8.4 Difference of working circuits for the outdoor coil in two experimental cases.

230

Defrosting for Air Source Heat Pump

using a flow hood that had a 16-point velocity grid located at the center of a 400  400 mm air duct that was 600 mm long. Air temperature and humidity downstream of the outdoor coil were measured by a hygrosensor located inside an air duct 900 mm downstream of the outdoor coil. In Eqs. (8.1), (8.2), Qi, a is the thermal energy transferred from the indoor air, which was evaluated by: Z

td

Qi, a ¼

ci, a mi, a  dT ¼

X ci, a ρi, a Vi, a Δt  ðTind, in  Tind, out Þ

(8.8)

0

where ci, a is the specific heat of the indoor air and mi, a the mass rate of the indoor air. ρi, a is the density of air in the indoor heated space, and Vi, a the volumetric flow rate of air passing through the indoor coil. Tind, in and Tind, out are the measured air temperature at the inlet and outlet of the indoor coil, respectively. All the system operating parameters, such as temperature, pressure, relative humidity, refrigerant mass flow rate, voltage, current, etc., were measured in real time. All sensors and measuring devices were able to output direct current signals of 4–20 mA or 1–5 V to a DAS for logging and recording. All the measured data throughout both the frosting and defrosting periods were collected and recorded by the DAS at an interval of 5 s. During defrosting, photos of surface conditions of the outdoor coil were also taken at an interval of 10 s. The measuring accuracy for various sensors/instruments used in the experimental ASHP unit was summarized in Chapter 3. Experimental procedures and conditions, the calculated relative standard errors for the four calculated parameters, the total energy supply for defrosting, and the total energy consumption during defrosting were also detailed. (2) Experimental cases

A series of experimental works using the experimental ASHP unit was carried out to investigate the energy transfer process and the effect of MES on system defrosting performance. In order to obtain meaningful experimental results, first it was necessary to ensure that the MES was different during RCD. MES is decided by the metal temperature difference, specific heat, and total mass. In this study, frosting/defrosting modes fixed the lowest/highest metal temperature of the outdoor coil and the highest/lowest value of the indoor coil, respectively. That means the metal temperature difference is unchangeable. Specific heat is also constant, decided by the type of material. Therefore, only the total metal mass of the indoor or outdoor coil could be adjusted. For an ASHP unit with a multicircuit outdoor coil, it could be reached by changing the working circuit number, with the help of the solenoid valves installed at the outlet refrigerant pipe of each circuit. Total refrigerant mass flow quality is constant when different numbers of circuits are working at defrosting mode. Second, for each circuit, frost accumulation over their surfaces should be similar at different experimental cases. In this study, this was carried out by adjusting the opening degrees of the manual stop valves installed at the inlet of each refrigerant pipe, and thus adjusting the refrigerant mass flow rate into each circuit. With this operational method, the FEC was controlled at higher than 90%. For each circuit, the frost accumulation difference was less than 5%.

Energy transfer during defrosting

231

Finally, experimental work was carried out at the two experimental cases. At frosting mode, there were three working circuits, Circuits 1–3, in the two cases. However, at defrosting mode, only Circuits 1 and 2 worked in Case 1, but three circuits all worked in Case 2. The differences of the working circuits for the outdoor coil in the two experimental cases are illustrated in Fig. 8.4, and the system operation differences are summarized in Table 8.3. The total metal masses of the outdoor coil in the two cases are 2020 and 3030 g, respectively. Consequently, the system defrosting performances at different MES could be comparatively and quantitatively analyzed. (3) MES calculations

The metal temperatures of the indoor and outdoor coils change with the temperature variations of the inner flowing refrigerant. At the beginning of a defrosting operation, the metal temperature of the indoor coil was high, where a lot of heat was stored in the metal. At defrosting, the warm indoor air transferred thermal energy to the cold refrigerant across the tube and fin of the indoor coil, which resulted in the metal temperature decreasing. Thus, the heat stored in the metal was taken away by the cold refrigerant. Meanwhile, some energy taken by the hot refrigerant was used to heat the cold outdoor coil metal, then to heat the frost and the retained water on its surface. The heat transferred around the metal of the indoor and outdoor coils in the system, QMe, was evaluated by: Z QMe ¼

td

PMe  dt ¼

X PMe  Δt

(8.9)

0

where td is the defrosting duration. The rate of heat supply from the indoor coil metal, PMe, was evaluated by: PMe ¼ cPMe  ðmCu + mAl Þ 

ΔTMe Δt

(8.10)

where mCu and mAl are the total masses of copper and aluminum of the coils. ΔTMe is the average temperature difference of the indoor coil metal, and was evaluated by: ΔTMe ¼ T0  Tt

(8.11)

1 T0 ¼ ðTin,0  Tout,0 Þ 2

(8.12)

1 Tt ¼ ðTin,t  Tout, t Þ 2

(8.13)

T0 and Tt are the average temperatures of the indoor coil metal at the start and end of the defrosting operation, Tin, 0 and Tout, 0 are the inlet and outlet tube surface temperatures of the indoor coil at the start of the defrosting operation, and Tin, t and Tout, t are the inlet and outlet tube surface temperatures of the indoor coil at the end of the defrosting operation, respectively. In Eq. (8.8), cPMe, the average specific heat of copper and aluminum, could be evaluated by:

232

Defrosting for Air Source Heat Pump

cPMe ¼

mCu cCu + mAl cAl mCu + mAl

(8.14)

Using the Eqs. (8.9)–(8.14), the MES values of the indoor coil and outdoor coil during RCD were obtained. In Eq. (8.2), Qi, MES is the energy discharged from the metal of the indoor coil during defrosting, and Qo, MES is the energy stored in the metal of the outdoor coil. The two parameters were calculated by the following equations, Qi, MES ¼ Qi, MES0  Qi, MESt

(8.15)

Qo, MES ¼ Qo, MESt  Qo, MES0

(8.16)

where Qi, MES0 and Qo, MES0 are the energy stored at the metal of the indoor and outdoor coils at the start of defrosting, and Qi, MESt and Qo, MESt are the energy stored at the metal of the indoor and outdoor coils at the end of defrosting, respectively. The four parameters could be calculated with the Eq. (8.9). When the energy discharged from the metal of the indoor coil is higher than the energy stored in the metal of the outdoor coil during defrosting, there is ηm > 0. It means that the metal energy storage has a positive effect on system defrosting performance. On the contrary, the metal energy storage has a negative effect on system defrosting performance when the energy stored in the metal of the indoor coil is less than that of the outdoor coil.

8.2.2 Results and analysis Experimental results in the two cases are listed in Table 8.5. At the start of defrosting, the total masses of frost accumulation were 710 g in Case 1 and 952 g in Case 2, with their average values at 355 and 317.3 g for each circuit in the two cases, respectively. That means that more frost accumulated on a circuit’s surface during the frosting stage in Case 1. During defrosting, the total masses of melted frost collected were 620 g in Case 1 and 921 g in Case 2, with their average values at 310 and 307 g for each circuit, respectively. It is easy to understand that larger frost accumulation means more melted frost is collected. As weighted and calculated, the total masses of retained water were 81 g in Case 1 and 88 g in Case 2, with their average values at 40.5 and 29.3 g for each circuit, respectively. As calculated, more frost was vaporized for each circuit in Case 1. Therefore, it is demonstrated that the defrosting efficiency in Case 1 is higher in the two cases due to more vaporization energy for a circuit being efficiently used. Fig. 8.5 presents eight photographs showing the airside surface conditions of the outdoor coil during defrosting in the two cases. Obviously, there are no watercollecting trays installed between circuits. Therefore, the two cases can reflect the defrosting conditions in a traditional ASHP unit, with the melted frost flowing downward freely along the outdoor coil surface. Fig. 8.5(1A–1D) are in Case 1, with the two working circuit, and Fig. 8.5(2A–2D) are in Case 2, with the three working circuit, respectively. As observed from Fig. 8.5(1A and 2A), the surface conditions at the start of defrosting for each circuit in the two cases were visually the same, which agreed

Energy transfer during defrosting

233

Fig. 8.5 Airside surface conditions of the outdoor coil during defrosting in two cases. (1A) 0 s in Case 1. (1B) 60 s in Case 1. (1C) 100 s in Case 1. (1D) 120 s in Case 1. (2A) 0 s in Case 2. (2B) 100 s in Case 2. (2C) 120 s in Case 2. (2D) 150 s in Case 2.

well with the fixed high FEC as mentioned in the previous section. It is obvious that the frost accumulated more in Case 1 than in Case 2, which also met the 355 and 317.3 g listed in Table 8.5. In Case 1, the preheating defrosting stage ended about 60 s into defrosting, as shown in Fig. 8.5(1B). After that, the tube and fin started to contact with the ambient air. As shown in Fig. 8.5(1C), the melted frost started flowing downward away from the circuits at about 100 s into defrosting in Case 1. When the defrosting came to 120 s, most of the frost was melted off, as presented in Fig. 8.5(1D). However, at 100 s into defrosting in Case 2, the preheating stage had ended, and the tube and fin contacted with the ambient air for a long time, as shown in Fig. 8.5(2B). At 120 s into defrosting, there was still much frost left, as shown in Fig. 8.5(2C). Most of the frost was melted off at about 150 s in Case 2, as presented in Fig. 8.5(2D). Comparing the last two pictures in the two cases, it was demonstrated that the defrosting duration in Case 1 would be at least 30 s earlier than that in Case 2. Using Eqs. (8.1)–(8.13), the experimental data of the defrosting heat supplies and energy consumptions for the experimental ASHP unit in the two cases were calculated and presented in Figs 8.6–8.14. In Figs. 8.6–8.9 and 8.14, for their time (horizontal) axis, 0 s is the actual start time for the defrosting operation. From the experiments, the defrosting durations for Case 1 and Case 2 were obtained as 152 and 185 s, respectively. It is also reflected in Fig. 8.6. Therefore, in Figs. 8.6–8.9 and 8.14, all the data during defrosting are shown. As to Figs. 8.10–8.13, the melted frost began to flow down from the water-collecting tray to the cylinder at 110 and 125 s for Case 1 and Case 2, respectively. As the defrosting terminated, the heat transfer between the melted frost collected and the ambient air around the outdoor coil continued. Therefore, the data in the two figures showed the results until 155 s for Case 1 and 200 s for Case 2 into the defrosting operation.

234

Defrosting for Air Source Heat Pump

Fig. 8.6 Measured tube surface temperature of the lowest outdoor coil circuit.

Fig. 8.6 shows the measured tube surface temperature of the lowest outdoor coil circuit in the two cases. In this study, the lowest circuit was Circuit 2 in Case 1 and Circuit 3 in Case 2, as shown in Fig. 8.4 and Table 8.4. Clearly, two curves reached 24°C at 152 s in Case 1 and 185 s in Case 2, respectively. That means the time difference of the two defrosting durations was 33 s, which met the demonstration given in Fig. 8.5. As seen in Fig. 8.6, at 60 s into defrosting in Case 1, the tube surface of Circuit 2 was at about 0°C. Also, the tube surface of Circuit 2 in Case 2 was also at 0°C at 100 s into defrosting. This met Fig. 8.5(1B and 2B). In addition, at 120 s in Case 1 and 150 s in Case 2, the temperature values were also nearly the same, at 11.5°C and 10.7°C, respectively. This was also reflected in Fig. 8.5(1D and 2D). However, from 0 to 60 s, the temperature in Case 1 was always lower than that in Case 2. This might be because of different frost accumulations, at 355 g in Case 1 and 317.3 g at Case 2, as listed in Table 8.5. After 60 s into defrosting, the delay of defrosting due to more frost made the temperature in Case 1 obviously higher than that in Case 2. The average values of the indoor and outdoor coils’ measured metal temperatures in the two cases are given in Fig. 8.7. Obviously, the trends of the indoor coils in the two cases are nearly the same. However, the temperature differences between the two outdoor coils were always big, from 5.1°C at 0 s to 20.0°C at 150 s into defrosting. This phenomenon results from different frost accumulations on each circuit in the two cases. The biggest temperature differences between the indoor and outdoor coils were 39.2°C at 150 s in Case 1 and 66.5°C at 185 s in Case 2. From 0 to 130 s, the trends of the two outdoor coils in the two cases were nearly the same. However, after 130 s, the trend of the outdoor coil temperature in Case 2 was changed to suddenly increase. The rate of temperature increase was changed from 40.9°C at the first

Energy transfer during defrosting

235

o

Indoor coil (Case 1) Outdoor coil (Case 1) Indoor coil (Case 2) Outdoor coil (Case 2)

70

o

Measured mean metal temperature ( C)

80

60

22.0 C/55 s

o

50

o

40.9 C/130 s

20.0 C

40 o

66.5 C 30 20

o

39.2 C

10 0

o

5.1 C 0

25

50

75

100 Time (s)

125

150

175

200

Fig. 8.7 Measured mean metal temperatures of indoor and outdoor coils in two cases.

Measured air temperature difference (⬚C)

16 o

13.8 C

Case 1 Case 2

14 12 10 8 6 4 2

135 s

0 0

20

40

60

80

100 120 Time (s)

140 s 140

160

180

200

Fig. 8.8 Measured air temperature differences between the inlet and outlet of the indoor coil.

130 s to 22.0°C at the later 55 s. That means the MES effect might be changed to negative. The measured air temperature differences between the inlet and outlet of the indoor coil in the two cases are presented in Fig. 8.8. As seen, their trends are similar. Therefore, during defrosting, the processes of thermal energy being taken from the indoor air were nearly the same. At 135 and 140 s into defrosting in the two cases, the curves

236

Defrosting for Air Source Heat Pump

Fig. 8.9 Energy input to the compressor during defrosting in the two cases.

Case 1 Case 2

2.1

o

Temperature of melted frost collected ( C)

2.4

1.8 1.5 1.2 0.9 0.6 0.3

125 s 110

120

130

140

155 s 150 160 170 Time (s)

180

190

200

210

Fig. 8.10 Temperature variations of melted frost collected in the two cases.

reached their peaks at the same value of 13.8°C. The curve of Case 2 was also few later than that of Case 1. These reflect that the energy-taken process in Case 1 was much quicker than that in Case 2. Fig. 8.9 shows the energy input to the compressor during defrosting in the two cases. The two curves’ trends are similar, with the curve of Case 1 increasing much quicker after 80 s into defrosting. At the end of defrosting, they reached their peak values at 4.12 kJ in Case 1 and 3.69 kJ in Case 2, respectively.

Energy transfer during defrosting

10

Circuit 1 in Case 1 Circuit 2 in Case 1 Circuit 1 in Case 2 Circuit 2 in Case 2 Circuit 3 in Case 2

9 8

o

Temperature of ambient air ( C)

237

7 6 5 4 3 2 110

120

130

140

150 160 170 Time (s)

180

190

200

210

Fig. 8.11 Temperature variations of ambient air around each circuit of the outdoor coil.

900

Energy from system outside (kJ)

800 700

Thermal energy of indoor air MES of indoor coil Input to indoor air fan Input to compressor

600 500 400

77.94%

80.10%

6.04% 0.40% 15.61%

4.50% 0.82%

300 200 100 0 Case 1

14.58% Case 2

Fig. 8.12 Heat supplies during defrosting in the two cases.

The data in Case 2 were always lower than that in Case 1. It looks contradictory because the total frost accumulated in Case 1 was much more than that in Case 2, as listed in Table 8.5. However, it mainly resulted from the higher frost density in Case 1, which increased the energy-taken rate during defrosting. More power input to the compressor was transferred by the refrigerant to the frost via the tube and fins of the outdoor coil. In addition, the energy input to the compressor was kept small, at lower

238

Defrosting for Air Source Heat Pump

900 800

Energy consumptions (kJ)

700

Heating ambient air Heating melted frost Heating outdoor coil metal Vaporizing retained water Melting frost

600

44.97%

500 52.04%

1.17% 6.68%

1.33%

5.13%

400 300

5.71% 3.59%

200

43.21%

38.67%

100 0 Case 1

Case 2

Fig. 8.13 Heat consumptions during defrosting in two cases. Table 8.4 System operation differences in the two experimental cases Item

Parameters

Case 1

Case 2

1

Circuit number of outdoor coil working during frosting Circuit number of outdoor coil working during defrosting Circuit number of indoor coil working during frosting Circuit number of indoor coil working during defrosting Number of water-collecting trays installed during frosting Number of water-collecting trays installed during defrosting Total metal mass of indoor coil Total metal mass of outdoor coil The lowest circuit detecting the defrosting termination temperature Results shown in

3 (Circuit 1–3)

3 (Circuit 1–3)

2 (Circuit 1–2)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

1 (Tray B)

1 (Tray C)

1 (Tray B)

1 (Tray C)

2496 g 2020 g Circuit 2

2496 g 3030 g Circuit 3

Figs. 8.5–8.14; Tables 8.5 and 8.6

Figs. 8.5–8.14; Tables 8.5 and 8.6

2 3 4 5 6 7 8 9 10

3.0 kJ, at 0–80 s in Case 1 and 0–120 s in Case 2. It is demonstrated that at the start of defrosting, little energy for defrosting came from the energy input, and most of the energy came from the indoor air thermal energy, or the MES. This also met the increase of air temperature differences between the inlet and outlet of the indoor coil, as shown in Fig. 8.8.

Energy transfer during defrosting

239

Fig. 8.14 Heat transferred around the metal of the indoor coil and outdoor coil in the two cases.

Measured heat transferred (kJ)

6 From indoor coil metal To outdoor coil metal

5 4

S3

3

S4

2

S2 1

S1

0

(A)

0

20

40

60

80 100 Time (s)

120

140

160

Measured heat transferred (kJ)

8 From indoor coil metal To outdoor coil metal 6

S3 4

S2 2 0

(B)

S4

S1 0

40

80

120 Time (s)

160

200

Table 8.5 Experimental results in two cases Item

Parameter

Case 1

Case 2

Unit

1 2 3 4 5 6

Total mass of frost accumulation Mass of frost accumulation for each circuit Total mass of melted frost collected Mass of melted frost collected for each circuit Total mass of retained water Mass of retained water for each circuit

710 355 620 310 81 40.5

952 317.3 921 307 88 29.3

g g g g g g

Fig. 8.10 shows the temperature variations of the melted frost collected. As seen, the trends of the two curves are similar, increasing quickly at first and keeping steady later. The temperature values kept at the range of 0.3°C to 2.0°C. Therefore, little energy was taken away by the melted frost during defrosting. In addition, the temperature of the melted frost in Case 2 was always later and lower than that in Case 1. This is because more frost accumulated on each circuit in Case 1 as compared with Case 2.

240

Defrosting for Air Source Heat Pump

But the total value of frost accumulation in Case 2 was more than that in Case 1. Therefore, during defrosting, more melted frost was downward flowing along the surface of the outdoor coil in Case 2. This increased the flowing rate of the melted frost, and thus the temperature was lower. However, as shown in Fig. 8.11, the temperature variations of the ambient air around each circuit of the outdoor coil in the two cases are totally different. The temperature in Case 1 was always lower than 4.5°C, but the peak value was nearly at 9.5°C in Case 2. It is obvious that the ambient air was warmed a lot in Case 2. Therefore, more energy would be wasted during defrosting in Case 2, which also prolonged the defrosting duration and degraded the defrosting efficiency. Fig. 8.12 summarizes the heat supplies from the four fields during defrosting: (1) thermal energy of the indoor air, (2) MES of the indoor coil, (3) input to the indoor air fan, and (4) input to the compressor. The total heat supply was about 613 kJ for Case 1 and 761 kJ for Case 2, 24% higher. This results from different total frost accumulations in the two cases. As illustrated, the indoor air accounts for the highest ratio in the two cases, at 77.9% in Case 1 and 80.1% in Case 2, respectively. Energy supplies from the indoor coil metal for the two cases were 37 and 34 kJ, accounting for 6.04% and 4.50%, respectively. The ratio differences of energy input to the compressor and indoor air fan were very small, less than 1%. This is obvious and easily understood because the two values increased as time progressed. This may be the reason why it was neglected in previous calculations. In conclusion, most of the energy for defrosting came from the indoor air, and the ratio of MES would be different when the number of working circuits was changed. Fig. 8.13 shows the heat consumption during defrosting in the two cases. As seen, there are the following five consumptions: (1) heating the ambient air, (2) heating the melted frost, (3) heating the outdoor coil metal, (4) vaporizing the retained water, and (5) melting the frost. Obviously, the energy consumed on heating the ambient air took the biggest percentage, at 52.04% in Case 1 and 44.97% in Case 2, respectively. Their differences mainly result from different total areas of the outdoor coil. The percentage of energy consumed on melting frost took 38.7% in Case 1 and 43.2% in Case 2, due to different frost accumulations at the start of defrosting. Around 20% of the energy was consumed on heating the retained water and the outdoor coil metal and vaporizing the retained water. Compared with Case 1, the energy consumed on heating the outdoor coil metal in Case 2 was increased. It would degrade the MES effect on defrosting performance. However, the ratios of energy consumed on melting frost and vaporizing retained water were also increased in Case 1. Therefore, a higher defrosting efficiency was expected. To further quantitatively study the effect of MES, Fig. 8.14 shows the MES of the indoor coil and outdoor coil during defrosting. Defrosting efficiency and MES effects were also calculated and listed in Table 8.6. As seen in Fig. 8.14, the energy is represented by the area of shadow. The total energy used for defrosting can be expressed by, EMES ¼ ES1  ES2 + ES3  ES4

(8.17)

in which ES1 and ES3 are the net energy transferred from the indoor coil to the outdoor coil, and ES2 and ES4 are the net energy transferred from the outdoor coil to the indoor

Energy transfer during defrosting

241

coil, respectively. It is obvious that the EMES in Case 1 was positive. It was calculated at 0.33%, as listed in Table 8.6. Meanwhile, the value of the MES effect in Case 2 was calculated at 2.18%, which also agreed well with Fig. 8.14. That means the MES had a negative effect in Case 2, after the working circuit number increased 50% from two to three circuits. Basing on the two cases with frost evenly accumulated on the surface of twoworking-circuit and three-working-circuit outdoor coils, an experimental study on the energy transfer process during RCD in an air source heat pump unit has been carried out and reported, with the following conclusions: (1) Four types of heating supply and five types of energy consumption were quantitatively analyzed. As observed, the heating supply of the indoor air thermal energy contributed about 80% of the total energy usage for defrosting, with more than 40% of the energy wasted in heating the ambient air. (2) The effect of metal energy storage on defrosting performance was quantitatively evaluated. As concluded, after the outdoor coil was enlarged 50% from two working circuits to three working circuits, the metal energy storage effects changed from positive (0.33%) to negative (2.18%). (3) The percentages of energy consumed on vaporizing the retained water and melting the frost were both increased. Defrosting efficiency was increased about 6.08%, from 42.26% in the twoworking-circuit case to 48.34% in the three-working-circuit case. (4) The law of energy transfer process and the effect of metal energy storage could guide the design optimization of two coils and promote energy saving for air source heat pump units. The effect of metal energy storage on frosting performance should also be further quantitatively evaluated.

8.3

Defrosting with local drainage of the melted frost

The space heating used energy is changed to be consumed on melting frost and vaporizing water. Not only was the indoor space heating interrupted, but also the thermal comfort level was adversely affected [11]. The energy conversion process directly affects the defrosting performance, which is a key problem for the application of ASHP units. Therefore, to improve the defrosting performance, various experimental studies were conducted. Noticeably, the MES values of the indoor and outdoor coils Table 8.6 Defrosting efficiency and MES effect in two cases Item

Parameter

Case 1

Case 2

Unit

1 2

Energy consumed on melting frost Energy consumed on vaporizing retained water Energy from indoor coil Energy consumed in outdoor coil Total energy supply Defrosting efficiency MES effect

237.1 22.0

329 39.1

kJ kJ

37.1 35.0 613.2 42.26% 0.33%

34.2 50.9 761.4 48.34% 2.18%

kJ kJ kJ – –

3 4 5 6 7

242

Defrosting for Air Source Heat Pump

were quantitatively analyzed. As indicated, the energy consumptions for heating the outdoor coil metal accounted for 16.5% of the total defrosting energy consumption [12]. And thus, the MES was considered in the previous defrosting model development, as introduced in Chapter 4. The MES effects on defrosting performance for an ASHP unit should be identified. It is meaningful to quantitatively study the energy transfer mechanism shown in Fig. 8.1. On the other hand, the negative effects of melted frost on defrosting were demonstrated in the aforementioned experimental studies. It was demonstrated that much thermal energy was consumed during defrosting when the melted frost flowed downward along the surface of the multicircuit outdoor coil, or was kept on the downside of the outdoor coil due to surface tension. After water-collecting trays were installed between circuits in two-circuit and three-circuit outdoor coils, the total defrosting energy consumption could be decreased by 10.3% and 10.4%, respectively. Clearly, when the energy conversion process was analyzed, the condition of the melted frost locally drained should not be neglected. It is a fundamental parameter that affects the optimization of ASHP units. Understanding the melted frost and MES effects on defrosting performance are of importance for the application ASHP units, but energy conversion studies are scarce in the open literature. An experimental investigation on the energy conversion process in an ASHP unit during defrosting with the melted frost locally drained has been carried out. In this section, the experimental setup used is totally the same as that used in the previous section. All information about the environmental chamber, the outdoor coil and indoor coil, and the DAS system are detailed in Chapter 3. Here, two settings of the two-working-circuit and the three-working-circuit cases are first designed. Basing on the experimental results, the evaluations and discussions on defrosting performance of this experimental ASHP unit are presented, and the effect of MES is comparatively and quantitatively analyzed. The conclusions of this study are finally given.

8.3.1 Experimental cases To qualitatively and quantitatively investigate the energy conversion process in an ASHP unit during RCD with melted frost locally drained, a series of experiments should be carried out. Basing on these experimental results, all types of energy supplies and consumptions during defrosting could be calculated. Meanwhile, the effect of MES on defrosting performance could be studied. First, an ASHP unit was selected in which a three-circuit outdoor coil was tailor-made. Then, two typical experimental conditions were designed, with two and three circuits of the outdoor coil used. To avoid uneven frosting influence, the frost was adjusted to be evenly accumulated on the surface of each circuit. Their FECs were kept at higher than 90%. Here, the tube surface temperature worked as the controlling index of frost accumulation distribution during frosting. Then, the energy supplies and consumptions in different fields were calculated. Finally, the defrosting performance was evaluated by the defrosting efficiency and MES effect on defrosting. The equations were listed in the previous section. A series of experimental works using the experimental ASHP unit was carried out to investigate the effect of MES on its RCD in an ASHP unit with melted frost

Energy transfer during defrosting

Circuit 1

Circuit 2

m f,1 m f,2

243

Circuit 1

Circuit 1 m f,1

Circuit 2

Circuit 1 m f,1

m f,1

Tray A Circuit 2

m f,2

m f,2

Circuit 2

m f,2 Tray B

Tray B Circuit 3

Circuit 3 m f,3

Frosting in Case 1

Circuit 3 m f,3

Defrosting in Case 1

Tray A

Frosting in Case 2

m f,3

Tray C

Defrosting in Case 2

Fig. 8.15 Difference of working circuits for the outdoor coil in two experimental cases.

locally drained. In order to obtain meaningful experimental results, first it was necessary to ensure the MES was different during RCD. MES is decided by the metal temperature difference, specific heat, and total mass. In this study, the frosting/ defrosting modes fixed the lowest/highest metal temperature of the outdoor coil and the highest/lowest value of the indoor coil, respectively. That means the metal temperature difference is unchangeable. Specific heat is also constant, decided by the type of material. Therefore, only the total metal mass of the indoor or outdoor coil could be adjusted. For an ASHP unit with a multicircuit outdoor coil, it could be reached by changing the working circuit number with the help of the SVs installed at the outlet refrigerant pipe of each circuit. Total refrigerant mass flow quality is constant when different numbers of circuits are working at defrosting mode. Second, for each circuit, frost accumulation over the surface should be similar at different experimental cases. In this study, it was carried out by adjusting the opening degrees of the MVs installed at the inlet of each refrigerant pipe, and thus adjusting the refrigerant mass flow rate into each circuit. With this operational method, the FEC was controlled at higher than 90%. For each circuit, the frost accumulation difference was less than 5%. In addition, the melted frost should be locally drained during defrosting by installing water-collecting trays between circuits. Finally, experimental work was carried out at the two experimental cases. At frosting mode, there were three working circuits, Circuits 1–3, in two cases. However, at defrosting mode, only Circuits 1 and 2 worked in Case 1, but three circuits all worked in Case 2. The differences of the outdoor coils in the two experimental cases are illustrated in Fig. 8.15, and the system operation differences are summarized in Table 8.7. The total metal masses of the outdoor coils in the two cases are 2020 and 3030 g, respectively. Consequently, the system defrosting performances at different MESs could be comparatively and quantitatively analyzed.

8.3.2 Results and analysis The experimental results in the two cases are listed in Table 8.8. At the start of defrosting, the total masses of frost accumulation were 717 g in Case 1 and 1080 g in Case 2, with their average values at 358.5 and 360 g for each circuit in the two cases, respectively. The difference is 1.5 g, or about 0.41%, which is very small. That means that the comparative analysis work on the two frosting cases is meaningful. During

244

Defrosting for Air Source Heat Pump

Table 8.7 System operation differences in the two experimental cases Item

Parameters

Case 1

Case 2

1

Circuit number of indoor coil working during frosting Circuit number of indoor coil working during defrosting Circuit number of outdoor coil working during frosting Circuit number of outdoor coil working during defrosting Number of water-collecting trays installed during frosting Number of water-collecting trays installed during defrosting Total metal mass of indoor coil Total metal mass of outdoor coil The lowest circuit detecting the defrosting termination temperature Results shown in

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

3 (Circuit 1–3)

2 (Circuit 1–2)

3 (Circuit 1–3)

1 (Tray B)

1 (Tray C)

2 (Trays A, B)

3 (Trays A, B, C)

2496 g 2020 g Circuit 2

2496 g 3030 g Circuit 3

Figs. 8.15–8.24; Tables 8.8 and 8.9

Figs. 8.15–8.24; Tables 8.8 and 8.9

2 3 4 5 6 7 8 9 10

Table 8.8 Experimental results in the two cases Item

Parameter

Case 1

Case 2

Unit

1 2 3 4 5 6 7

Total mass of frost accumulation Mass of frost accumulation for each circuit Total mass of melted frost collected Mass of melted frost collected for each circuit Total mass of retained water Mass of retained water for each circuit Mass of vaporized water calculated for each circuit

717 358.5 655 327.5 55 18.3 12.7

1080 360 969 323 97 32.3 4.7

g g g g g g g

defrosting, the total masses of the melted frost collected were 655 g in Case 1 and 969 g in Case 2, with their average values at 327.5 and 323 g for each circuit, respectively. The difference is 4.5 g, or about 1.39%, with a little more melted frost collected in Case 1. As weighted and calculated, the total masses of retained water were 55 g in Case 1 and 97 g in Case 2, with their average values at 18.3 and 32.3 g for each circuit, respectively. As calculated, more frost was vaporized for each circuit in Case 1. Therefore, it is demonstrated that the defrosting efficiency in Case 1 is higher in the two cases, due to more vaporization energy for a circuit being efficiently used.

Energy transfer during defrosting

245

Fig. 8.16 presents eight photographs showing the airside surface conditions of the outdoor coil during defrosting in the two cases. Obviously, there are water-collecting trays installed under the circuits. Therefore, the two cases can reflect the defrosting conditions in an ASHP unit, with the melted frost locally drained. Fig. 8.16(1A–1D) are in Case 1, with two working circuits, and Fig. 8.16(2A–2D) are in Case 2, with three working circuits, respectively. As observed from Fig. 8.16(1A and 2A), the surface conditions at the start of defrosting for each circuit in the two cases were visually the same, which agreed well with the fixed high FEC. It is obvious that the frost accumulations in the two cases are similar, which also met the 358.5 and 360 g listed in Table 8.8. In Case 1, the preheating defrosting stage ended before 60 s into defrosting, as shown in Fig. 8.16(1B). As shown, the upside tube and fin started to contact with the ambient air. As shown in Fig. 8.16(1C), the melted frost was downward flowing away from the circuits at about 100 s into defrosting in Case 1. When the defrosting came to 120 s, most frost was melted off, as presented in Fig. 8.16(1D). However, at 100 s into defrosting in Case 2, the preheating stage had ended, and the tube and fin contacted with the ambient air for a long time, as shown in Fig. 8.16(2B). At 130 s into defrosting, there was still much frost left, as shown in Fig. 8.16(2C). Most frost was melted off at about 150 s in Case 2, as presented in Fig. 8.16(2D). Comparing the two last pictures in the two cases, it was demonstrated that the defrosting duration in Case 1 would be earlier by at least 30 s than that in Case 2. Using the previously listed equations, the experimental data of the defrosting heat supplies and energy consumptions for the experimental ASHP unit in the two cases were calculated and presented in Figs. 8.17–8.25. In Figs. 8.17–8.20 and

Fig. 8.16 Airside surface conditions of the outdoor coil during defrosting in two cases. (1A) 0 s in Case 1. (1B) 60 s in Case 1. (1C) 100 s in Case 1. (1D) 120 s in Case 1. (2A) 0 s in Case 2. (2B) 100 s in Case 2. (2C) 130 s in Case 2. (2D) 150 s in Case 2.

246

Defrosting for Air Source Heat Pump

8.25, for their time (horizontal) axis, 0 s is the actual start time for defrosting operation. From the experiments, the defrosting durations for Case 1 and Case 2 were obtained as 136 and 164 s, respectively. It is also reflected in Fig. 8.18. Therefore, in Figs. 8.18–8.21 and 8.25, all the data during defrosting are shown. As to Figs. 8.22 and 8.23, the melted frost began to flow down from the water-collecting tray to the cylinder at 110 and 125 s for Case 1 and Case 2, respectively. As the defrosting terminated, the heat transfer between the melted frost collected and the ambient air around the outdoor coil continued. Therefore, the data in the two figures showed the results until 140 s for Case 1 and 165 s for Case 2 into the defrosting operation. Fig. 8.17 shows the mean measured tube surface temperature of all outdoor coil circuits in the two cases. Clearly, the two curves reached 24°C at 136 s in Case 1 and 164 s in Case 2, respectively. That means the time difference of the two defrosting durations was 28 s, which met the demonstration given in Fig. 8.16. As seen in Fig. 8.17, at 60 s into defrosting in Case 1, the tube surface temperature of Circuit 2 was at about 0.8°C. That means the preheating stage was just over. The tube surface temperature of Circuit 2 in Case 2 was at 16°C at 100 s into defrosting. This met Fig. 8.16(1B and 2B). At 100 s in Case 1 and 130 s in Case 2, the temperature values were at 8.1°C and 12.5°C, respectively. Therefore, in Fig. 8.16(1C and 2C), it was at the status of the melted frost flowing downward along the circuits. When it came to 120 s in Case 1 and 150 s in Case 2, the temperature values were at 16.1°C and 20.3°C, respectively. That means Fig. 8.16(2D) is more close to the defrosting termination than that in Fig. 8.16(1D). However, from 0 to 20 s, the temperature in Case 1 was always lower than that in Case 2. This might be because of different frost accumulations, at 358.5 g in Case 1 and 360 g at Case 2, as listed in Table 8.5. After 40 s into defrosting, the delay of defrosting due to more frost made the temperature in Case 1 obviously higher than that in Case 2. The average values of the indoor and outdoor coils’ measured metal temperatures in the two cases are presented in Fig. 8.18. Obviously, at the start of defrosting, the two differences were different, at 18.5°C in Case 1 and 7.6°C in Case 2, respectively. This might result from their different frost accumulations at the start of defrosting. However, at the termination of defrosting, the temperature differences between the indoor and outdoor coils in the two cases were nearly the same, at 33.1°C in Case 1, and 32.2° C in Case 2. This also confirmed that this comparative study was meaningful. From 0 to 60 s into defrosting, the two outdoor coil temperatures were nearly kept the same. As shown in Fig. 8.18, there is a parallelogram. This is because after 100 s into defrosting in Case 1 and 132 s in Case 2, their rates of temperature increase were nearly the same. Therefore, the difference of temperature increasing rates came out at the melted frost downward flowing stage as described in Chapter 4. It is the melted frost effects that make the temperature increasing rate in Case 2 lower. The measured air temperature differences between the inlet and outlet of the indoor coil in the two cases are presented in Fig. 8.19. As seen, their trends are similar. Therefore, during defrosting, the processes of thermal energy taken from the indoor air were nearly the same. At 135 and 140 s into defrosting in the two cases, the curves reached their peaks at the values of 13.8°C and 14.0°C. The curve of Case 2 was also few later

Energy transfer during defrosting

247

Fig.8.16(2B)

Fig. 8.17 Mean measured tube surface temperatures of all outdoor coil circuits.

40

Indoor coil (Case 1) Outdoor coil (Case 1) Indoor coil (Case 2) Outdoor coil (Case 2)

o

Measured mean metal temperature ( C)

35 30

32 s

25

32 s

20 15

Melted frost downwards flowing stage

18.5 C

10

33.1 C 32.2 C

5 7.6 C

0

60 s

132 s

–5 0

25

50

75

100

125

150

175

200

Time (s)

Fig. 8.18 Measured mean metal temperatures of the indoor and outdoor coils in the two cases.

than that of Case 1. This reflects that the energy taken process in Case 1 was much quicker than that in Case 2. Fig. 8.20 shows the energy input to the compressor during defrosting in the two cases. The two curves’ trends are similar, with the curve of Case 1 increasing much quicker after 80 s into defrosting. At the end of defrosting, they reached their peak values at 3.72 kJ in Case 1 and 3.66 kJ in Case 2, respectively.

248

Defrosting for Air Source Heat Pump

16

o

Measured air temperature difference ( C)

o

14.0 C

14

o

Case 1 Case 2

12

13.8 C

10 8 6 4 2 135 s

0 0

20

40

60

80

100 120 Time (s)

140 s 140

160

180

200

Fig. 8.19 Measured air temperature differences between the inlet and outlet of the indoor coil.

Fig. 8.20 Energy input to the compressor during defrosting in the two cases.

The data in Case 2 were always lower than those in Case 1. It looks contradictory because the total frost that accumulated in Case 1 was much more than that in Case 2, as listed in Table 8.8. However, it mainly resulted from the higher frost density in Case 1, which increased the energy taken rate during defrosting. In addition, the energy input to the compressor was kept small, at lower 3.0 kJ, at 0–86 s in Case 1 and 0–100 s in Case 2. It is demonstrated that at the start of defrosting, little energy

Energy transfer during defrosting

249

0.4 Tray A in Case 1 Tray B in Case 1 Tray A in Case 2 Tray B in Case 2 Tray C in Case 2

0.3

Case 1

Case 2

o

Temperature ( C)

0.2

0.1

0.0

–0.1

–0.2 100

110

120

130

140

150

160

170

180

Time (s)

Fig. 8.21 Temperature variations of the melted frost collected in the two cases.

5.0 Circuit 1 in Case 1 Circuit 2 in Case 1 Circuit 1 in Case 2 Circuit 2 in Case 2 Circuit 3 in Case 2

o

Temperature of ambient air ( C)

4.5 4.0 3.5

Case 1

3.0 2.5 2.0 1.5 100

Case 2 110

120

130

140 Time (s)

150

160

170

180

Fig. 8.22 Temperature variations of ambient air around each circuit of the outdoor coil.

for defrosting came from the energy input, and most of the energy came from the indoor air thermal energy, or the MES. This also met the increase of air temperature differences between the inlet and outlet of the indoor coil, as shown in Fig. 8.20. Fig. 8.21 shows the temperature variations of the melted frost collected. As seen, the trends of the two curves are both increasing quickly. The temperature values were

250

Defrosting for Air Source Heat Pump

900

Energy from system outside (kJ)

800 700

Thermal energy of indoor air MES of indoor coil Input to indoor air fan Input to compressor

600 500 400

79.26%

81.22%

5.09% 0.39% 15.26%

2.98% 0.63%

300 200 100 0 Case 1

15.17% Case 2

Fig. 8.23 Heat supplies during defrosting in the two cases.

900 800

Energy consumptions (kJ)

700

Heating ambient air Heating melted frost Heating outdoor coil metal Vaporizing retained water Melting frost

600

33.10%

500 46.33%

1.46% 6.65% 5.09%

400 1.02%

300

5.53% 3.14%

200

53.70%

43.99%

100 0 Case 1

Case 2

Fig. 8.24 Heat consumptions during defrosting in the two cases.

kept at lower than 0.3°C. The low temperature means that little energy was taken away by the melted frost during defrosting. In addition, the temperature of the melted frost in Case 2 was always later and lower than that in Case 1. This minor difference might result from more frost accumulating on each circuit in Case 2, as compared with that in Case 1. During defrosting, more melted frost was downward flowing along the surface of the outdoor coil in Case 2. This increased the flowing rate of the melted frost, and

Energy transfer during defrosting

251

Fig. 8.25 Heat transferred around the metal of the indoor coil and the outdoor coil.

Measured heat transferred (kJ)

6 From indoor coil metal To outdoor coil metal

5

S3

4 3 2

S2

1

S1

0

(A)

0

20

40

60

80 100 Time (s)

120

140

160

Measured heat transferred (kJ)

8 From indoor coil metal To outdoor coil metal

S2

6 4 2

S3 S1

0 0

(B)

40

80

120 Time (s)

160

200

thus the temperature was lower. As shown in Fig. 8.22, the temperature variations of the ambient air around each circuit of the outdoor coil in the two cases showed similar trends. First, the temperature in Case 1 was always higher than that in Case 2. Second, the air temperature around the lowest circuit was always lower, Circuit 1 in Case 1 and Circuit 3 in Case 2, respectively. Figs. 8.21 and 8.22 demonstrate that the defrosting efficiency in Case 1 was lower than that in Case 2, due to more energy consumed on heating melted frost and the surrounding air. Fig. 8.23 summarizes the heat supplies from the four fields during defrosting: (1) thermal energy of the indoor air, (2) MES of the indoor coil, (3) input to the indoor air fan, and (4) input to the compressor. The total heat supply was about 544 kJ for Case 1 and 672 kJ for Case 2, 23.4% higher. This results from different total frost accumulations in the two cases. As illustrated, the indoor air accounts for the highest ratio in the two cases, at 79.26% in Case 1 and 81.22% in Case 2, respectively. The energy supplies from the indoor coil metal for the two cases were 28 and 20 kJ, accounting for 5.09% and 2.98%, respectively. The ratio differences of the energy input to the compressor and the indoor air fan were very small, less than 1%. This is easily understood because the two values increased as time. This may be the reason why it was

252

Defrosting for Air Source Heat Pump

neglected in previous calculations. In conclusion, most of the energy for defrosting came from the indoor air, and the ratio of MES would be different when the number of working circuits was changed. Fig. 8.24 shows the heat consumption during defrosting in the two cases. As seen, there are the following five consumptions: (1) heating the ambient air, (2) heating the melted frost, (3) heating the outdoor coil metal, (4) vaporizing the retained water, and (5) melting the frost. Obviously, the energy consumed on heating the ambient air took a big percentage, at 46.33% in Case 1 and 33.10% in Case 2, respectively. Their differences mainly result from different total areas of the outdoor coils. The percentage of energy consumed on melting frost took 43.99% in Case 1 and 53.70% in Case 2, due to different frost accumulations at the start of defrosting. Around 10% of the energy was consumed on heating the retained water and outdoor coil metal and vaporizing the retained water. Compared with Case 1, the energy consumed on heating the outdoor coil metal in Case 2 was increased. It would degrade the MES effect on defrosting performance. However, the ratios of energy consumed on melting frost and vaporizing retained water were also increased in Case 1. Therefore, a higher defrosting efficiency was expected. To further quantitatively study the effect of MES, Fig. 8.25 shows the MES of the indoor coil and outdoor coil during defrosting. The defrosting efficiency and MES effects were also calculated and listed in Table 8.9. As seen in Fig. 8.25, the EMES values in the two cases were negative. They were calculated at 0.44% in Case 1 and 3.67% in Case 2, respectively. This agreed well with Fig. 8.25, in which the area shadow in Fig. 8.25B was much bigger than that in Fig. 8.25A. Therefore, after the two-working-circuit outdoor coil was changed to a three-working-circuit coil, the negative effects of MES on defrosting performance increased. In conclusion, the following conclusions could be reached from this section: (1) Four types of heating supply were quantitatively analyzed. As indicated, the heating supply of the indoor air thermal energy contributed about 80% of the total energy usage for defrosting, with 15% energy from the compressor inputs. The total energy from the metal energy storage of the indoor coil and the input to the indoor air fan costs only about 5%. (2) Five types of energy consumption were divided and listed. During defrosting, nearly 90% of the energy was consumed on melting frost and heating Table 8.9 Defrosting efficiency and MES effect in the two cases Item

Parameter

Case 1

Case 2

Unit

1 2

Energy consumed on melting frost Energy consumed on vaporizing retained water Energy from indoor coil Energy consumed in outdoor coil Total energy supply Defrosting efficiency MES effect

239.5 17.1

360.7 34.2

kJ kJ

27.7 30.1 516.7 47.13% 0.44%

20.0 44.7 651.7 58.79% 3.67%

kJ kJ kJ – –

3 4 5 6 7

Energy transfer during defrosting

253

ambient air. The metal energy storage of the outdoor coil accounts for about 5%. Decreasing the proportion of energy consumed on heating the ambient air could effectively improve defrosting efficiency. (3) The effect of metal energy storage on defrosting performance was quantitatively evaluated. As concluded, after the outdoor coil was enlarged by 50%, from a two working circuit to a three working circuit, the metal energy storage effects changed from 0.44% to 3.67%. The percentages of energy consumed on vaporizing retained water and melting frost were both increased. (4) A higher defrosting efficiency was reached, from 47.13% in a two-working-circuit case to 58.79% in a three-working-circuit case, with 11.66% improvement. The law of energy conversion and the effect of metal energy storage could guide the design optimization of the two coils and save energy for ASHP units.

8.4

Discussion on effect of melted frost and thermal comfort

8.4.1 Effect of melted frost To analyze the difference between with and without melted frost downward flowing during defrosting as well as adjusting the number of working circuits in the outdoor coil, the experimental results in the four cases in the previous sections were summarized in Table 8.10. As seen in this table, the total energy supplies in the four cases are different. After the water-collecting trays were installed between circuits in Cases 3 and 4, the energy consumptions were clearly reduced. For two-working-circuit cases, Cases 1 and 3, the energy consumption decreased from 613.2 to 516.7 kJ, with a reduction of 96.5 kJ. When it comes to three-working-circuit cases, Cases 2 and 4, the energy consumption reduced about 109.7 kJ, from 761.4 to 651.7 kJ. Their reduction ratios are similar at 15.7% and 14.4%, respectively. The defrosting efficiency for two-working-circuit cases, Cases 1 and 3, improved from 42.26% to 47.13%, with an increase of 4.87% after the trays were installed. For three-working-circuit cases, Cases 2 and 4, the defrosting efficiency improved from 48.34% to 58.79%, with an increase of 10.45% after the trays were installed. That means that the negative effects of melted frost on a two working circuit and a three working circuit are different, with the latter higher. These results also meet the conclusions in Chapter 3. As indicated, after the working-circuit number increased from two to three, the negative effects of melted frost improved. The MES effects in Cases 1–4 were calculated at 0.33%, 0.44%, 2.18%, and 3.67%, respectively. Finally, the MES effects on defrosting performance were calculated at 2.51% by Cases 1 and 3, and at 3.23% by Cases 2 and 4. That means that the MES effects on defrosting performance for an ASHP unit would be increased, from 2.51% to 3.23%, after the melted frost was taken away during defrosting by installing water-collecting trays under each circuit in the multicircuit outdoor coil. It is meaningful for the structural optimization of ASHP units as well as their energy savings with a heating mode at a low-temperature and high-humidity environment.

254

Defrosting for Air Source Heat Pump

Table 8.10 Experimental cases designed in this study Item

Parameter

Case 1

Case 2

Case 3

Case 4

1 2 3

FEC Water-collecting trays Circuit number of indoor coil working during frosting Circuit number of indoor coil working during defrosting Circuit number of outdoor coil working during frosting Circuit number of outdoor coil working during defrosting Number of water-collecting trays installed during frosting Number of water-collecting trays installed during defrosting Total energy supply Defrosting efficiency MES effect

>90% Without 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 2 (Circuit 1–2) 1 (Tray B)

>90% Without 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 1 (Tray C)

>90% With 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 2 (Circuit 1–2) 1 (Tray B)

>90% With 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 3 (Circuit 1–3) 1 (Tray C)

1 (Tray B)

1 (Tray C)

2 (Trays A, B)

3 (Trays A, B, C)

613.2 kJ 42.26% 0.33%

761.4 kJ 48.34% 2.18%

516.7 kJ 47.13% 0.44%

651.7 kJ 58.79% 3.67%

4

5

6

7

8

9 10 11

8.4.2 Thermal comfort The ASHP unit plays a critical role in the indoor thermal comfort level. The RCD process always results in the deterioration of indoor thermal comfort. The nature of the problem is that the indoor air thermal energy is taken away for defrosting. Clearly, understanding the energy transfer mechanism in the ASHP system is fundamental for avoiding the deterioration of indoor thermal comfort. The first problem is that the defrosting energy should be preprepared, which should not be taken from the indoor air. As mentioned in Chapter 2, there are some publications about using a PCM-TES unit to avoid the indoor thermal energy being taken away during defrosting. In fact, as shown in Fig. 8.26, the function of MES is the same as a PCM-TES unit. First, the two coils could be considered as two coils and two TES units, from Fig. 8.26A and B. Then, the two TES units in the system could be shortened to only one. For the condition of the MES having a positive effect, as shown in Fig. 8.26C, only one TES unit is located at the indoor side. The net energy stored is the difference of TES-in and TES-out in Fig. 8.26B. Moreover, the TES-in-out is in series in Fig. 8.26C, which could be considered in parallel in the ASHP unit, as shown in Fig. 8.26D.

Energy transfer during defrosting

Outdoor coil

Compressor

Outdoor space

Indoor space

(D)

TES-in-out

Compr essor

Indoor coil

EEV

Outdoor coil

Indoor space

EEV

Compr essor

Indoor coil

(B) Outdoor space

Outdoor coil

Compr essor

TES-in-out

(A)

EEV

TES-in

EEV

Indoor space

Indoor coil

Outdoor space

TES-out

Indoor space

Indoor coil

Outdoor coil

Outdoor space

255

(C)

Fig. 8.26 Illustrations of two coils and the TES unit in an ASHP unit. (A) Two coils. (B) Two coils and two TES units. (C) One TES unit in series. (D) One TES unit in parallel.

Here, the design method for a PCM-TES could be introduced. As an ASHP unit takes Q from the outside air, the total energy input to the TES is COP  Q. The efficiencies of energy storage and release are n1 and n2, respectively. The two values are percentages, in the range of 0%–100%. During defrosting, the total energy consumed from the indoor coil is Qd. There is the balance of Qd ¼ COP  Q  η1  η2. So, there are four directions to limit the volume of TES: (1) improve the energy storage efficiency, (2) improve the energy release efficiency, (3) decrease the defrosting energy consumption, or improve defrosting efficiency, and (4) increase the energy storage per unit volume. Clearly, when we consider optimizing the ASHP unit, the fundamental study could guide us to find the solution.

8.5

Concluding remarks

In this chapter, experimental studies on the energy transfer mechanism in an ASHP unit and the effects of indoor and outdoor coils’ MES on defrosting performance were carried out, with the following conclusions: (1) Four types of heat supply and five types of energy consumption were quantitatively analyzed. The heat supplies during defrosting included the thermal energy of the indoor air, the MES of the indoor coil, the power input to the indoor air fan and compressor, and the heat consumption during defrosting, including heating the ambient air, heating the melted frost, heating the outdoor coil metal, and vaporizing the retained melting frost; (2) The MES effects on

256

Defrosting for Air Source Heat Pump

defrosting performance were experimentally examined and quantitatively evaluated. For a traditional ASHP unit, if the outdoor coil was enlarged by 50%, the MES effects were changed from positive (0.33%) to negative (2.18%). For a novel ASHP unit, the MES effects changed from 0.44% to 3.67%. The negative value suggested more cold energy stored in the outdoor coil metal, and thus certain energy ta the ken from indoor air should be first used to balance the stored energy; (3) The melted frost effect on MES was analyzed. After the melted frost was taken away during defrosting by installing water-collecting trays under each circuit in an multicircuit outdoor coil, the MES effects on defrosting performance for an ASHP unit could be increased; and (4) Indoor thermal comfort during defrosting was discussed when using an ASHP unit for space heating and the discussion results could help further optimize a defrosting operation.

References [1] Cole RA. Refrigeration loads in a freezer due to hot gas defrost and their associated costs. ASHRAE Trans 1989;95(1):1149–54. [2] Krakow KI, Lin S, Yan L. An idealized model of reversed-cycle hot gas defrosting of evaporators, Part 1: Theory. ASHRAE Trans 1993;99(2):317–28. [3] Krakow KI, Lin S, Yan L. An idealized model of reversed-cycle hot gas defrosting of evaporators, Part 2: Experimental analysis and validation. ASHRAE Trans 1993;99 (2):329–38. [4] Kim JH, Braun JE, Groll EA. A hybrid method for refrigerant flow balancing in multicircuit evaporators: upstream versus downstream flow control. Int J Refrig 2009;32 (6):1271–82. [5] Kim JH, Braun JE, Groll EA. Evaluation of a hybrid method for refrigerant flow balancing in multi-circuit evaporators. Int J Refrig 2009;32(6):1283–92. [6] Wang ZH, Wang FH, Ma ZJ. Numerical study on the operating performances of a novel frost-free air-source heat pump unit using three different types of refrigerant. Appl Therm Eng 2017;112:248–58. [7] Qu ML, Xia L, Deng SM, Jiang YQ. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil unit in an air source heat pump-Part II: Modeling analysis. Appl Energ 2012;91:274–80. [8] Dong JK, Deng SM, Jiang YQ, Xia L, Yao Y. An experimental study on defrosting heat supplies and energy consumptions during a reverse cycle defrost operation for an air source heat pump. Appl Therm Eng 2012;37:380–7. [9] Song MJ, Deng SM, Xia L. A semi-empirical modeling study on the defrosting performance for an air source heat pump unit with local drainage of melted frost from its three-circuit outdoor coil. Appl Energ 2014;136:537–47. [10] Song MJ, Xia L, Deng SM. A modeling study on alleviating uneven defrosting for a vertical three-circuit outdoor coil in an air source heat pump unit during reverse cycle defrosting. Appl Energ 2016;161:268–78. [11] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse-cycle defrosting method for air source heat pumps. Build Environ 2010;45 (11):2354–61. [12] Dong JK, Deng SM, Jiang YQ, Xia L, Yao Y. An experimental study on defrosting heat supplies and energy consumptions during a reverse cycle defrost operation for an air source heat pump. Appl Therm Eng 2012;37:380–7.

Defrosting control strategy 9.1

9

Introduction

For an ASHP unit during defrosting, it is easy to understand that its energy efficiency is always affected by the defrosting initiation and termination control strategies [1]. The recent advancements in artificial intelligence and the Internet of Things also promote the developments of advanced control strategies for various equipment and processes [2]. For an ASHP unit, two types of defrosting initiation control strategies are in place, time-based and demand-based. For the first type, defrosting initiation is always simply controlled by using a preset timer, at 60–90 min of frosting duration. However, it is hard to give an exact and fixed frosting duration or defrosting starting point, due to the complicated and changeable ASHP operating conditions. Therefore, the use of a time-based initiation strategy always results in two typical mal-defrosting problems: unnecessary defrosting when no or little frost is accumulated on the outdoor coil’s surface, and no defrosting when it is actually needed. As shown by Wang et al., these mal-defrosting phenomena adversely downgrade system frosting performance. For example, after the ASHP system was operated for 5 days, the aforementioned second type of mal-defrosting was found, with a frosted area of more than 60% on the multicircuit outdoor coil while no defrosting was initiated. Meanwhile, the system COP was significantly reduced to only 2.3 while the ambient air temperature was at 7.9°C. As analyzed with the testing results before and after the frosting operation, maldefrosting would decrease the COP and the heating capacity by 40.4% and 43.4%, respectively [1, 3]. Demand-based defrosting initiation control strategy could start a defrosting operation only when sufficient frost is formed to adversely affect the operational performance of ASHP units. Thus, it relies on accurately detecting the presence and growth of frost by using direct and indirect frost accumulation-sensing technologies. These technologies could be classified as: (1) measuring the frost thickness using a holographic interferometry technique; (2) measuring the frost surface temperature by an infrared thermometer [4]; (3) detecting the refrigerant flow instability [5]; (4) detecting the frost accumulation using a photocoupler or photo- or fiberoptic sensors; (5) simulating the frost amount by applying neural networks [6]; and (6) calculating the effective mass-flow fraction [7]. The frost accumulation on the surface of the outdoor coils was fully investigated with both experimental and modeling approaches. For example, in Yao’s distributed mathematical frosting model, the frost thickness could reach 0.5 mm at most with ambient air temperature at 1.5°C and relative humidity at 85%. Based on the first principles, a frosting model was developed and validated by Da Silva, and the effects of progressive frost clogging and the low conductivity of the frost layer on the overall thermal resistance were assessed, suggesting that the former is the main cause of heating capacity reduction under frosting conditions [8]. In Ye’s study, the errors of numerical heat-transfer rate and frost mass with the Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00009-6 © 2019 Elsevier Ltd. All rights reserved.

258

Defrosting for Air Source Heat Pump

experimental values were limited to 7% and 9%, respectively [9]. However, it is still a difficult problem to accurately detect the frost accumulation for an ASHP unit. With respect to the time-based defrosting initiation control strategy in application, the frost distribution on the surface of a multicircuit outdoor coil is always neglected. In fact, frost is always unevenly distributed due to the complicated and variable ambient conditions. Two typical types of uneven frosting phenomena for ASHP units were found, as illustrated in Fig. 9.1. Type 1 is the unequal frost accumulation on the surfaces of different circuits. For example, in an experimental study on the performance of an ASHP unit reported by Wang et al., for a kind of mal-defrost phenomenon appearing in moderate climate conditions, the frost accumulation on the downside is much more that on the upside of its vertically installed multicircuit outdoor coil [3]. As shown in Fig. 9.1A, a similar phenomenon could also be found in experimental studies reported by Qu [10], Song [11], and Steiner [12]. Type 2 is the unequal frost accumulation on the windward and leeward sides of the outdoor coil. As known for Type 1 uneven frosting, frost evenness could be quantitatively calculated with the melted frost collected by water-collecting trays during defrosting. Type 2 results from the obvious relative humidity difference between the inlet air and outlet air. However, this type is hardly possible to be further quantitatively described. Furthermore, it is the total frost accumulation on the circuit that directly affects its defrosting performance. Therefore, only the Type 1 uneven frosting phenomenon was widely considered in previous works. This may also be the reason why nearly no attention was paid to the Type 2 uneven frosting phenomenon. This type of uneven frosting phenomenon was only mentioned in a recent experimental work reported by Zhang et al. [13], as shown in Fig. 9.1B. It is reported that the ratios of the frost mass accumulated on the edge of the windward fins to that on the entire surface were 13.7% (60 min frosting duration) and 12.5% (120 min frosting duration) for the two heat exchangers used in this study, respectively. Although the frost accumulations on the two sides of an outdoor coil seem different, their masses were not clearly given. Therefore, only the Type 1 uneven frosting influence on the time-based defrosting initiation control strategy was investigated here.

9.2

Time-based initiation of defrosting control

As demonstrated, a higher FEC would improve the frosting and defrosting performances for an ASHP unit, but how to influence the defrosting performance for different frost accumulations at high FECs is still unknown. The uneven frosting phenomenon should be fully considered when we further develop or optimize the time-based defrosting initiation control strategy. After this fundamental problem was qualitatively and quantitatively solved, the preset frosting duration could be given in a time-based defrosting initiation control strategy [7]. The mal-defrosting problems could be effectively avoided, and thus potentially large amounts of energy are saved [14]. Consequently, an experimental study on the optimization of the time-based defrosting initiation control strategy for an ASHP unit with frost evenly distributed was designed and carried out. First, the experimental setup was introduced as well

Defrosting control strategy

Fig. 9.1 Two types of uneven frosting phenomena for ASHP units. (A) Three examples of Type 1 uneven frosting. (B) Example of Type 2 uneven frosting. (C) Two types of uneven frosting. 259

260

Defrosting for Air Source Heat Pump

as the experimental procedures. Then, five designed continuous cases were listed. After these observed, measured and calculated results were presented, the system energy and stability performances as well as the defrosting starting indexes were comparatively discussed and analyzed. The conclusions of this work were finally given.

9.2.1 Experimental cases To investigate the frost accumulation influence on RCD performance, experimental works were carried out using the described experimental ASHP unit. First, it was necessary to ensure that the frost accumulations on the surface of the outdoor coil were different to reach meaningful experimental results. It could be reached by two methods. One is changing the relative humidity of the outdoor air in the experiment because water vapor in the outdoor air is the source of frost. But the relative humidity was always kept at a fixed value during frosting for comparison. Finally, another method is changing the duration of the frosting mode, which was taken in this work. In a real application, a time-based RCD is always started after frosting for 60–90 min. However, the relative humidity of the ambient air is always at 40%–80%, which is much lower than 90%. Therefore, in this study, the frosting durations were designed to be a little shorter, at 50–70 min. Second, for each circuit, frost accumulations over their surfaces should be similar at different experimental cases. The FEC should be nearly the same in the different experimental cases. For an ASHP unit with a multicircuit outdoor coil, the FEC is hard to adjust due to many parameters affecting frosting performance, including the structure of the heat exchanger, the type of fin and its surface, the distribution of air and refrigerant, etc. However, the refrigerant flow to each circuit could be varied by adjusting the opening degree of the modulating valves, and thus adjusting the frost accumulations and FEC. Therefore, in this study, a set of fixed valve opening degrees was finally obtained by trial-and-error manual adjustments of the opening degrees of the valves. With this operational method, the FEC was successfully controlled at higher than 90%. For each circuit, the frost accumulation difference was less than 5%. After the opening degree was fixed, the water-collecting Cylinder D was placed under Circuit 3 during defrosting. Then, a meaningful and effective experimental case was conducted. Finally, the entire procedure of an experimental case is clearly illustrated in Fig. 9.2. As shown in Fig. 9.2, the experimental ASHP unit was first operated at frosting mode, at an air dry-bulb temperature of 0.5
0.2°C and a relative humidity of 90
3%. This ambient condition was jointly maintained by the use of experimental ASHP unit, the LGUs, and humidifiers placed in the outdoor frosting space. Meanwhile, the indoor air temperature was maintained at 20
0.2°C by jointly using the ASHP unit and the existing DX A/C system. Before frosting was terminated, the opening degrees of valves could be adjusted as required. Before initiating the defrosting operation, the FEC should be controlled at a value of higher than 90%. In order to keep the system stable and safe, the four-way valve was switched to defrosting mode after compressor shutdown for 1 min. It cost about 4 s before the compressor was powered on again, and then a defrosting operation started. In this

Defrosting control strategy

261

Fig. 9.2 Flow chart of the entire procedure in an experimental case.

study, the defrosting operation was manually terminated when the tube surface temperature at the exit of Circuit 3 reached 24°C. At the beginning of a frosting operation, the outdoor air fan during frosting was at 1.2 m s1, which would decrease slowly during frost growth. The indoor air fan was changed from 3.68 to 2.31 m s1, to supply enough energy by continuously operating at the lower speed. During frosting, no water-collecting tray was placed under the circuit. But, one tray or three trays were installed during defrosting, depending on the requirements of different cases. Finally, five experimental cases were designed and carried out in this study, as listed in Table 9.1. The frosting durations were at an interval of 5 min in five cases. That means the frost accumulations were steadily increased from Case 1 to Case 5. The FECs in the five cases were all higher than 90%, which meets the basic

Table 9.1 Five experimental cases designed in this study Case no.

Frosting duration (min)

FEC (%)

Trays during defrosting

Results shown in

Case 1 Case 2 Case 3 Case 4 Case 5

50

92.3

Without

Figs. 9.4–9.16, Table 9.2

55

95.6

Without

Figs. 9.4–9.16, Table 9.2

60

96.6

Without

Figs. 9.4–9.16, Table 9.2

65

95.6

Without

Figs. 9.4–9.16, Table 9.2

70

95.6

Without

Figs. 9.4–9.16, Table 9.2

262

Defrosting for Air Source Heat Pump

requirements of this study. Consequently, the system defrosting performances at different frost accumulations and with frost evenly distributed could be fully analyzed.

9.2.2 Experimental results and discussions Fig. 9.3 presents six photographs showing the outdoor coil airside surface conditions during frosting in the five cases. Fig. 9.3A shows the airside of the outdoor coil surface at the initiation of the frosting stage. Figs. 9.3B–F show the surface conditions at the end of the frosting operation for each circuit in the five cases. As observed, they were visually the same, which agreed well with the FECs listed in Table 9.1. Moreover, it was easy to visually distinguish the frost accumulation differences among the five cases, as the surface of the outdoor coil was becoming whiter and whiter from Case 1 to Case 5 corresponding to the increasing frost accumulations. However, it is hardly possible to visually distinguish the difference of FECs. That means, in this study,

Fig. 9.3 Outdoor coil airside surface conditions at the initiation and end of frosting in five experimental cases. (A) Frosting start for 5 cases, (B) FEC ¼ 92.3% for Case 1, (C) FEC ¼ 95.6% for Case 2, (D) FEC ¼ 96.6% for Case 3, (E) FEC ¼ 95.6% for Case 4, and (F) FEC ¼ 95.6% for Case 5.

Defrosting control strategy

263

calculating the FEC by adding water-collecting trays under each circuit is feasible and meaningful. Fifteen pictures of outdoor coil airside surface conditions during defrosting in five cases are shown in Fig. 9.4. Obviously, there is no water-collecting tray installed under the circuits. Therefore, the five cases can reflect the defrosting conditions in a traditional ASHP unit, with different frost accumulating on the surface of the outdoor coil. As seen from the five pictures in the first line, Fig. 9.4(1A)–(5A), they are conditions at 15 s into defrosting at the preheating stage. The color of the picture is not as white as that shown in Fig. 9.4. And on the surface of the tube at the right side of each picture, the frost was melted and the tube is bare. For the five pictures in the second line, Fig. 9.4(1B)–(5B), they are conditions at the end of the preheating stage. The places in Circuit 1 where white arrows point are where the fin begins to contact the ambient air. As seen, the time points are at 39, 40, 43, 87, and 102 s in Cases 1–5, respectively. The big gaps in the five cases show the main frost accumulation influence on the preheating stage of defrosting. The five pictures in the third line show the conditions

Fig. 9.4 Outdoor coil airside surface conditions at different stages of defrosting in the five experimental cases. (1A) 15 s in Case 1, (2A) 15 s in Case 2, (3A) 15 s in Case 3, (4A) 15 s in Case 4, (5A) 15 s in Case 5; (1B) 39 s in Case 1, (2B) 40 s in Case 2, (3B) 43 s in Case 3, (4B) 87 s in Case 4, (5B) 102 s in Case 5; (1C) 89 s in Case 1, (2C) 101 s in Case 2, (3C) 108 s in Case 3, (4C) 118 s in Case 4, (5C) 127 s in Case 5.

264

Defrosting for Air Source Heat Pump

Table 9.2 Experimental results in the five experimental cases

No.

Time of preheating stage terminated (s)

Time of the melted frost flowing from water-collecting tray (s)

Defrosting duration (s)

Mass of the melted frost collected (g)

Case 1 Case 2 Case 3 Case 4 Case 5

39 40 43 87 102

89 101 108 118 127

160 179 193 199 208

763 930 962 993 1051

when the melted frost started flowing away from the water-collecting Cylinder D. As observed in Fig. 9.4(1C)–(5C), the white dash lines show the conditions of frost melting. They are at 89, 101, 108, 118, and 127 s into defrosting in Cases 1–5, respectively. These differences all result from different frost accumulations in the five cases. Finally, the total masses of melted frost collected were 763, 930, 962, 993, and 1051 g in Cases 1–5, respectively. These experimental results are summarized in Table 9.2. It could be demonstrated that the frost accumulation is not proportional to the frosting duration. This also reflects the inaccuracy of a time-based defrosting initiation control strategy. Therefore, experimentally investigating the frost accumulation influence on the RCD performance for an ASHP unit is meaningful. Fig. 9.5 shows the measured tube surface temperature at the exit of the lowest circuit and outdoor coil during defrosting in the five cases. It could be found that from Fig. 9.5,

Fig. 9.5 The fluctuation of the tube surface temperature at Circuit 3 in the five experimental cases.

Defrosting control strategy

265

Fig. 9.6 The fluctuation of the tube surface temperature at the entrance of the outdoor coil in the five experimental cases.

the temperature curves of the lowest circuit, Circuit 3, in Case 1 and Case 2 started increasing at 80 s into defrosting, but the curves of the other three cases at 100 s into defrosting. Obviously, the delay results from different frost accumulations in the five cases. Finally, this figure shows that the defrosting terminations in Cases 1–5 were at 163, 179, 193, 199, and 208 s, respectively. Obviously, the less frost accumulation, the shorter the defrosting duration. Meanwhile, as shown in Fig. 9.6, the duration for each curve reaching 24°C shows the same status. It is 160, 174, 183, 193, and 203 s in Cases 1–5, respectively. The durations for temperatures at the exit of the outdoor coil are about 3–5 s earlier than those temperatures at the exit of Circuit 3. This is reasonable because the temperatures of the up-circuits always increased much quicker than the lowest circuit due to the downward flowing of the melted frost. In addition, part of temperature curve is lower than 0°C in Fig. 9.5, especially in Fig. 9.6. It results because the refrigerant temperature is much lower than 0°C during frosting. The exit of the outdoor coil during defrosting is just the entrance during frosting. The frost preheating stage is also prolonged by more frost accumulations. This also meets Fig. 9.4. Fig. 9.7 presents the fluctuation of the melted frost temperature collected in Cylinder D in the five experimental cases. As seen, the temperature data were collected at 110–145 s into defrosting, which is about 20 s later than the time of the melted frost flowing from water-collecting Tray C. This is because the flowing of melted frost costs some time. Also, it needs more water before the temperature sensor contacts with the water. Finally, after 200 s into defrosting, the temperature of the melted frost collected became steady at 2.09–2.61oC. However, in this figure, the gap between each curve is not as even as that shown in Fig. 9.6. The gaps between the first two curves and the later three ones are much bigger. Therefore, it is demonstrated that the most suitable frost accumulation might be between the data of Case 2 and Case 3. Fig. 9.8

266

Defrosting for Air Source Heat Pump

Fig. 9.7 The fluctuation of melted frost temperature collected in Cylinder D in the five experimental cases.

Fig. 9.8 The fluctuation of refrigerant volumetric flow rate at defrosting mode in the five experimental cases.

shows the fluctuation of refrigerant volumetric flow rate at defrosting mode in the five experimental cases. Obviously, these curves reached different peaks at different times due to different frost accumulations. It is 1.39 L/min at 145 s in Case 1, 1.40 L/min at 155 s in Case 2, 1.35 L/min at 160 s in Case 3, 1.34 L/min at 170 s in Case 4, and 1.32 L/min at 175 s in Case 5, respectively. The five curves, the same as the previous

Defrosting control strategy

267

Fig. 9.9 The fluctuation of refrigerant pressures at the inlet and outlet of the compressor in the five experimental cases.

curves, show as orderly from 80 to 180 s into defrosting. That means the more frost that is accumulated, the later the refrigerant flow rate peaks are reached. In addition, as seen from the figure, the highest peak occurs in Case 2. This also demonstrated that the most suitable frost accumulation might be around the data of Case 2. Figs. 9.9 and 9.10 show the fluctuations of the refrigerant pressures at the inlet and outlet of the compressor and their calculated differences in the five experimental

Fig. 9.10 The fluctuation of the refrigerant pressure difference in the five experimental cases.

268

Defrosting for Air Source Heat Pump

cases. As seen in Fig. 9.9, the discharge pressure of the compressor increased quickly when little frost accumulated. In addition, the pressure differences show the same status in Fig. 9.10. That means the less frost that accumulates on the outdoor coil’s surface, the quicker the measured compressor discharge pressor increased during defrosting. And thus, the system stability is better. Consequently, to avoid the system stability problem, frost accumulation should be controlled in a limited status. The discharge pressure curves in Case 1 and Case 2 are much quicker than those in the other three cases. This phenomenon is also obviously in the curves of suction pressure, especially from 80 to 160 s into defrosting. Although the gap between the frost accumulation data in Case 1 and Case 2 is big, the two curves nearly merged from 60 to 130 s. It further demonstrated that when frost accumulation is around that in Case 2, the system defrosting performance in stability is the best. Moreover, as shown in Fig. 9.10, from 85 to 120 s, the pressure differences in Case 4 and Case 5 are nearly lower than 2 bars. This is very dangerous for the system. Therefore, in view of system stability, frost accumulation in Case 3 may be the maximum for this experimental ASHP unit. Finally, it could be found that there are some 5 s-increasing stops in the pressure difference curves. This phenomenon might result from the end of the frost melting stage. When the defrosting status is changed to the vaporizing stage, there is a sudden decrease in energy consumption. Figs. 9.11 and 9.12 show the measured air temperature at the inlet and outlet of the indoor coil and the temperature differences. As seen in Fig. 9.11, at the first 20 s into defrosting, the air temperature at the outlet of the indoor coil is still increasing. This is the delay of frosting or heating mode. Then, the air temperature at the outlet of the indoor coil decreased quickly. From 35 to 65 s into defrosting, there is a pause for

Fig. 9.11 The fluctuation of air temperature at the inlet and outlet of the indoor coil in the five experimental cases.

Defrosting control strategy

269

Fig. 9.12 The fluctuation of the indoor coil air temperature difference in the five experimental cases.

decreasing curves for temperatures and their differences. This is because the defrosting is at the preheating stage. After the frost contacting with the tube and fins was melted, thin water films exist between the hot fins and the outside cold frost. Therefore, the rate of energy consumption decreased. But, the melted frost will flow downward when the maximum water surface tension is destroyed. Therefore, after about 30 s of melting, the temperature differences increased quickly. Then, as seen from Fig. 9.12, the curves of Case 1 and Case 2 increased much quicker than the other three curves. That means the total heat transfer between the fins and melted frost during defrosting was enhanced due to the downward flowing of the melted frost. Finally, because the frost accumulation in Case 5 is the most, the air temperature difference curve in Case 5 became the highest. In addition, after it reached the vaporizing stage, the temperature differences were nearly the same, and kept steady. However, it could be found from Fig. 9.11 that there are 1, 4, 7, 11, and 9 points lower than 1°C from Case 1 to Case 5, respectively. That means the deterioration of indoor thermal comfort is nearly proportional to the frost accumulation. Clearly, it will be better for Case 1 and Case 2. During defrosting, the energy mainly comes from the thermal energy taken from the indoor air and the electricity inputs on the compressor and fans. The two energy sources are separately presented in Figs. 9.13 and 9.14. As seen from Fig. 9.13, the peaks of thermal energy taken from the indoor air came at 135, 145, 155, 160, and 165 s in Cases 1–5. It is directly influenced by the frost accumulation. However, in Fig. 9.14, the curves fluctuated at the first 20 s into defrosting due to the initial start up of the system. Then, from 20 to 105 s, the electricity consumption in the five cases is nearly the same, and kept steady at 2.8 kJ. After 105 s, the curves started increasing. This might be because the energy requirements are increasing while it is hard to take more thermal energy from the indoor air at this defrosting stage. At the termination of

270

Defrosting for Air Source Heat Pump

Fig. 9.13 Thermal energy taken from the indoor air during defrosting in the five experimental cases.

Fig. 9.14 Electricity consumption on the compressor and fans during defrosting in the five experimental cases.

defrosting, the highest values in the five cases are nearly the same, at the range of 3.59–3.68 kJ. Fig. 9.15 shows the energy supply and effective energy consumption during defrosting in the five cases. As seen, the total energy supplies are 648.1 kJ in Case 1, 745.8 kJ in Case 2, 808.5 kJ in Case 3, 854.5 kJ in Case 4, and 890.8 kJ in Case 5,

Defrosting control strategy

271

Fig. 9.15 Energy supply and effective energy consumption in the five experimental cases.

respectively. And from Case 1 to Case 5, the total effective energy consumptions are 294.7, 343.4, 356.7, 368.1, and 389.1 kJ. Most energy comes from the indoor air thermal energy, which accounts for more than 80%. And most of the energy is consumed in melting the frost, which takes more than 80%. As the frost accumulation increased, the energy supply during defrosting increased correspondingly. However, the effective energy consumption is not changed as the frost accumulation is increasing. Therefore, it is demonstrated that the highest defrosting efficiency in the five cases is not Case 1 or Case 5. That means there is the most-suitable frost accumulation for this experimental ASHP unit, which can best improve the defrosting performance. In order to analyze the relationship between frost accumulation and defrosting performance, the defrosting efficiency was calculated in the five cases. As shown in Fig. 9.16, the frost accumulation increases as the increase of the frosting duration, and reached its highest value at 1051 g at 70 min in Case 5. Clearly, the increasing rate is not constant as time, which is much higher at 50–55 min. On the other hand, although the defrosting duration is prolonged as the frost accumulation increases, the defrosting efficiency reached its highest value at 46.05% in Case 2. That means that after the frost accumulation is more than 930 g, the defrosting performance becomes worse and worse as the frost accumulation increases. Consequently, the most suitable frost accumulation for good defrosting performance comes out around Case 2. In addition, compared with the results in Case 2, the system defrosting duration was prolonged by 29 s, or 16.2%, and the defrosting efficiency was decreased by 2.36%, or 5.13%, in Case 5. Furthermore, when the frosting performance decrease is considered in Case 5, the frost accumulation is highly recommended at 930 g. Therefore, it is suggested as the reference value for frosting termination. In conclusion, an experimental study on the optimization of a time-based defrosting initiation control strategy for an ASHP unit with different frost accumulations that were evenly distributed on the surface of the outdoor coil was conducted. The

272

Defrosting for Air Source Heat Pump

Fig. 9.16 Analysis on frost accumulation and defrosting efficiency in the five experimental cases.

following four conclusions were reached: (1) An experimental setup was designed and carried out as well as a three-circuit outdoor coil. To make the frost evenly accumulate on the surface of the outdoor coil, three water-collecting trays were installed under each circuit during defrosting in the frost evenness value experimental stage. (2) Frost accumulation increased as the frosting duration was prolonged, although it is not at a positive proportional relationship. After 50 min into frosting, the frost accumulation increased quickly at first, and then slowly. In the same interval of 5 min, frost accumulation increased 167 g, or 21.89%, from Case 1 to Case 2. However, the increase of frost accumulation was only 58 g, or 5.84%, from Case 4 to Case 5. (3) Defrosting duration is also not at a positive proportional relationship with frost accumulation. The increased frost accumulation prolonged the defrosting duration, but the main difference comes at the preheating stage during defrosting. When frost accumulation is 763 g in Case 1, the preheating stage costs 39 s while it takes 102 s for the frost accumulation at 1051 g in Case 5. (4) In view of system stability and indoor thermal comfort, the system performance would be degraded when the frost accumulation was more than 930 g in this study. In view of defrosting performance, the defrosting efficiency reached its peak at 46.05% when the frost accumulation was at 930 g. As demonstrated, frost accumulation is the most fundamental reference parameter for frosting termination in an ASHP unit.

9.3

Defrosting control with melted frost locally drained

For an ASHP unit having a multicircuit outdoor coil, the downward-flowing melted frost over the surface of the outdoor coil always has a negative effect on system defrosting performance. In addition, frost is always unevenly accumulated on the surface of the multicircuit outdoor coil. However, an experimental study on defrosting

Defrosting control strategy

273

performance for an ASHP unit at different FECs with melted frost locally drained was carried out [15]. After the melted frost was taken away by the water-collecting trays during defrosting, the defrosting duration was shortened by 11.2% and the defrosting efficiency increased by 5.7%, as the FEC increased from 79.4% to 96.6%. Clearly, it is meaningful to improve the FEC during the frosting stage and locally drain the melted frost during the defrosting stage. To further clarify the negative effect of melted frost, a semiempirical modeling study on the defrosting performance for an ASHP with local drainage of melted frost from its three-circuit outdoor coil was developed [16]. Based on the validated model, three study cases of varying heat supply to the outdoor unit were further investigated, demonstrating the optimization of the ASHP unit by decreasing the defrosting energy to 96.4% and reducing the defrosting duration by 7 s [17]. Clearly, when we optimize the time-based initiation defrosting control strategy, the melted frost effects could not be neglected. Although it was demonstrated that a higher FEC would improve both the frosting and defrosting performances for an ASHP unit, how to influence the defrosting performance for different frost accumulations at high FECs with the melted frost locally drained is still unknown. After this fundamental problem was qualitatively and quantitatively solved, the preset frosting duration could be given in a time-based initiation defrosting control strategy. The mal-defrosting problems could also be effectively avoided, and thus potentially large amounts of energy saved. Therefore, in this paper, an experimental study on time-based defrosting initiation control strategy optimization for an ASHP unit with frost evenly distributed and melted frost locally drained is carried out. First, the experimental setup was introduced. Then, the experimental procedures and five continuous cases were designed. After these were observed, the measured and calculated results were given and the system energy and stability performances and initiation defrosting indexes were comparatively analyzed and discussed. Finally, a conclusion was given. It is expected that this work will be used to optimize the control strategy for intelligent heat pumps as well as the ultimate goal of building energy savings.

9.3.1 Experimental cases To optimize the time-based initiation defrosting control strategy, a series of experimental works using the experimental ASHP unit was carried out. In order to obtain meaningful experimental results, first it was necessary to ensure that the frost accumulations on the surface of the outdoor coil were different. It could be reached with two methods. One is changing the relative humidity of the outdoor air in the experiment because water vapor in the outdoor air is the source of frost. The other one is changing the duration of the frosting process. Here, to shorten the frosting duration in the experiments, the second one was finally taken, with the air relative humidity kept at 90
3% during frosting. In real application, a time-based RCD is always started after frosting for 60–90 min, due to the relative humidity of the ambient air always being much lower, at 40%–80%. In addition, when the frosting duration is longer, the COP of the system would suddenly decrease. Therefore, in this study, the frosting durations were designed to be a little shorter, at 50–70 min.

274

Defrosting for Air Source Heat Pump

Fig. 9.17 Flow chart of the entire procedure in an experimental case.

Second, for each circuit, frost accumulation over their surfaces should be similar at different experimental cases. Moreover, the FECs should be nearly the same in different experimental cases. For an ASHP unit with a multicircuit outdoor coil, it is hard to adjust the FEC as many parameters affect the frosting performance, including the structure of the heat exchanger, the type of fin and its surface, the distribution of air and refrigerant, etc. However, it was demonstrated that modulating valves installed at an inlet refrigerant pipe to each circuit could be deployed to vary the refrigerant flow to each circuit, and thus the frost accumulations on each circuit are adjusted. Therefore, in this study, a series of trial-and-error manual adjustments of the opening degrees of the stop valves was carried out. Then, a set of fixed valve opening degrees was obtained, and frost accumulating on the surface of an outdoor coil at different FECs was reached. With this operational method, the FEC was controlled at higher than 90%. For each circuit, the frost accumulation difference was less than 5%. After the opening degree was fixed, the water-collecting Cylinder D was placed under Circuit 3 during defrosting. Then, a meaningful and effective experimental case was conducted. Finally, experimental work was carried out at five experimental cases, with the entire procedure of an experimental case shown in Fig. 9.17. As listed in Table 9.3, the frosting durations were at an interval of 5 min in the five cases. That means the frost accumulations steadily increased from Case 1 to Case 5. The FECs in the five cases were all higher than 90%, which meets the basic requirements of this study. Consequently, the system defrosting performances at different frost accumulations could be comparatively and quantitatively analyzed.

9.3.2 Experimental results and discussions Fig. 9.18 presents 20 photographs showing the airside surface conditions of the outdoor coil at the defrosting operation in the five cases. As observed from Fig. 9.18(1A)– (5A), the surface conditions at the start of defrosting for each circuit in the five cases

Defrosting control strategy

275

Table 9.3 Five experimental cases in this study Case no.

Frosting duration (min)

Melted frost collected

FEC (%)

Case 1

50

93.6

Case 2

55

Case 3

60

Case 4

65

Case 5

70

805 g (279 g/265 g/261 g) 933 g (318 g/311 g/304 g) 969 g (317 g/328 g/324 g) 1001 g (340 g/343 g/319 g) 1074 g (359 g/362 g/353 g)

95.6 96.6 93.8 97.5

Results shown in Figs. 9.18–9.27, Tables 9.3 and 9.4 Figs. 9.18–9.27, Tables 9.3 and 9.4 Figs. 9.18–9.27, Tables 9.3 and 9.4 Figs. 9.18–9.27, Tables 9.3 and 9.4 Figs. 9.18–9.27, Tables 9.3 and 9.4

were visually the same, which agreed well with the frost accumulation on the surface of each circuit, as listed in Table 9.3. The FECs from Case 1 to Case 5 are at 93.6%, 95.6%, 96.6%, 93.8%, and 97.5%, respectively. Moreover, it could be found that it is easy to visually distinguish the difference of frost accumulations among the five cases, as the surface of the outdoor coil is becoming whiter and whiter from Case 1 to Case 5, corresponding to the increasing frost accumulation. However, it is hardly possible to visually distinguish the difference of FECs. That means that in this study, calculating the FEC by adding water-collecting trays under each circuit is feasible and meaningful. Fifteen more pictures of the outdoor coil airside surface conditions during defrosting in the five cases are shown in Fig. 9.18(1B)–(5D). Obviously, there are water-collecting trays installed under the circuits. Therefore, the melted frost could be locally drained during defrosting. As seen from the five pictures in the second row, Fig. 9.18(1B)–(5B), they are the conditions at 15 s into defrosting. It is at the preheating stage, with all fins covered with frost. The color of the picture is not as white as that shown in Fig. 9.18(1A)–(5A). And on the surface of the tube at the right side of each picture, the frost was melted and the tube was bare. For the five pictures in the third row, Fig. 9.18(1C)–(5C), they are the conditions at the end of the preheating stage. At this moment, some water layer on the fin’s surface started directly connecting with the ambient air. The places in Circuit 1 where the white arrows point are where the fin surface was beginning to contact the ambient air. As seen, the time points are at 35 s in Case 1, 40 s in Case 2, 45 s in Case 3, 64 s in Case 4, and 96 s in Case 5, respectively. The big gaps in the five cases show the main frost accumulation influence on the preheating stage of defrosting. The five pictures in the fourth row show the conditions when the melted frost starts flowing away from their corresponding water-collecting trays. As observed in Fig. 9.18(1D)–(5D), the white dash lines show the conditions of frost melting. They are at 86, 93, 95, 112, and 114 s into defrosting in Cases 1–5, respectively. These differences all result from different frost accumulations in the five cases because the melted frost in the five cases was all locally taken away by the trays. Finally, the total masses of the melted frost

276

Defrosting for Air Source Heat Pump

1A

2A

3A

4A

5A

1B

2B

3B

4B

5B

1C

2C

3C

4C

5C

1D

2D

3D

4D

5D

Fig. 9.18 Outdoor coil airside surface conditions during defrosting in the five cases. (1A) FEC ¼ 93.6% in Case 1, (2A) FEC ¼ 95.6% in Case 2, (3A) FEC ¼ 96.6% in Case 3, (4A) FEC ¼ 93.8% in Case 4, (5A) FEC ¼ 97.5% in Case 5; (1B) 15 s in Case 1, (2B) 15 s in Case 2, (3B) 15 s in Case 3, (4B) 15 s in Case 4, (5B) 15 s in Case 5; (1C) 35 s in Case 1, (2C) 40 s in Case 2, (3C) 45 s in Case 3, (4C) 64 s in Case 4, (5C) 96 s in Case 5; (1D) 86 s in Case 1, (2D) 93 s in Case 2, (3D) 95 s in Case 3, (4D) 112 s in Case 4, (5D) 114 s in Case 5.

collected were 805 g in Case 1, 933 g in Case 2, 969 g in Case 3, 1001 g in Case 4, and 1074 g in Case 5, respectively. These experimental results are summarized in Tables 9.3 and 9.4. It could be demonstrated that the frost accumulation is not proportional to the frosting duration. This also reflects the inaccuracy of a time-based initiation defrosting control strategy. Therefore, experimentally investigating the frost accumulation influence on RCD performance for an ASHP unit is meaningful. Fig. 9.19 shows the measured tube surface temperature at the exit of each circuit and their average values during defrosting in the five cases. It could be found that from

Defrosting control strategy

277

Table 9.4 Experimental results in the five cases Item

Durations

Case 1 (s)

Case 2 (s)

Case 3 (s)

Case 4 (s)

Case 5 (s)

1

Time of preheating stage terminated Time of the melted frost flowing away from tray Defrosting duration Duration calculated by average value Duration calculated by outdoor coil temperature

35

40

45

64

96

86

93

95

112

114

153 150

160 159

172 170

180 179

189 188

149

155

171

174

186

2

3 4 5

Fig. 9.19A–E, the temperature curves of the lowest circuit, Circuit 3, in the five cases started increasing at 85, 100, 100, 105, and 110 s into defrosting. Obviously, the delay results from their different frost accumulations. This figure also shows the defrosting terminations in the five cases, at 153 s in Case 1, 160 s in Case 2, 172 s in Case 3, 177 s in Case 4, and 189 s in Case 5, respectively. The less the frost accumulation is, the shorter the defrosting duration. Meanwhile, as shown in Fig. 9.19F, the duration for the average temperature values of each circuit in the five cases reaching 24°C shows the same status. It is 150 s in Case 1, 159 s in Case 2, 170 s in Case 3, 179 s in Case 4, and 188 s in Case 5, respectively. The durations for their average temperatures are about 1–3 s earlier than those temperatures at the exit of Circuit 3. This is reasonable because the temperatures of the up-circuits always increase a little quicker than the lowest circuit due to the downward flowing of the melted frost. In addition, part of temperature curve is lower than 0°C in Fig. 9.19, especially in Fig. 9.19E. It results from the refrigerant temperature being much lower than 0°C during frosting. The exit of the outdoor coil during defrosting is just the entrance during frosting. That means the defrosting preheating stage was also prolonged by more frost accumulation. This also meets Fig. 9.18. As shown in Fig. 9.20, the durations of reaching 24°C for the measured tube surface temperature at the entrance of the outdoor coil during defrosting in the five cases are at 149, 155, 171, 174, and 186 s, respectively. Clearly, the differences of the five curves are unequal. The curve of Case 4 was much earlier than that of Case 3. This is because of the lower FEC in Case 4, at 93.8%. It was much lower than the FECs in Cases 3 and 5. In fact, when the FEC is lower, the temperature curve should be delayed and experimentally investigated. However, it is the opposite phenomenon shown in Fig. 9.20. Authors demonstrated that it results from the uneven distribution of the refrigerant during defrosting in this three-circuit outdoor coil. The same as the previous results, as frost accumulation increases from Case 1 to Case 5, the start curve becomes lower, and lower than 0°C for Cases 4 and 5 before 100 s into defrosting. Clearly, Case 2 has the best performance in the temperature increase process during defrosting. Additionally, in Table 9.4, the defrosting duration is calculated with the tube surface of the

278

Defrosting for Air Source Heat Pump

(A)

(B)

(C)

(D)

(E)

(F)

Fig. 9.19 Measured tube surface temperature at the exit of each circuit and their average values. (A) Case 1, (B) Case 2, (C) Case 3, (D) Case 4, (E) Case 5, and (F) 5 Cases.

lowest circuit, which is different from the other two durations calculated with the average temperature of the three circuits, and with the temperature of the outdoor coil. The three typical durations are shown in Figs. 9.19 and 9.20. Although their differences are small, their changing trends demonstrated that the best defrosting performance occurred in Case 2. Fig. 9.21 presents the measured temperatures of the melted frost collected in three cylinders during defrosting in the five cases. As seen, the temperature data were collected at 90–135 s into defrosting, which is about 20 s later than the time of the melted frost flowing away from the water-collecting tray. The temperature of Circuit 2 is

Defrosting control strategy

279

Fig. 9.20 Measured tube surface temperature at the entrance of the outdoor coil during defrosting.

always the first to fall down, which is because the refrigerant distribution resistance in this circuit is the lowest. After the melted frost flows into the cylinder, the temperature increase is because of two reasons. One is the heating by ambient air, and the other is the continued melted frost at higher temperature. Therefore, as shown in Fig. 9.21C, from 165 to 215 s into defrosting, the curve of Circuit 2 increased to be higher than that of Circuit 1. This is similar to Fig. 9.21E, but different from Fig. 9.21A, B, and D. Table 9.5 further lists the durations of melted frost downward flowing away from the trays. As seen, the average durations for the five cases are 90, 95, 110, 120, and 130, respectively. If the linear analysis was carried out with the data from all five cases, it could be found that the duration of Case 2 is much earlier than the theoretical value, 100 s. It was calculated with the average value of 90 and 110 s. This also demonstrated that Case 2 has the best defrosting performance. The most suitable frost accumulation should be around the data of Case 2, 933 g. Figs. 9.22 and 9.23 show the measured pressures of compressor suction and discharge and their calculated differences. As seen in Fig. 9.22, the discharge pressure of the compressor in Case 1 and Case 2 increased much more quickly due to less frost accumulation. The pressure in Case 2 was always the highest, and even its frost accumulation was not the lowest in the five cases. That means that the less frost that accumulates on the surface of the outdoor coil in an ASHP unit, the quicker the compressor discharge pressure will increase. And thus, the system stability is better. Consequently, to avoid the system stability problem, frost accumulation should be controlled in a limited status. Also, from 60 to 200 s into defrosting, the compressor suction pressures in Case 2 were always the highest. These phenomena demonstrated that the frost accumulation in Case 2 had the best defrosting performance in the five cases. In addition, the pressure differences show the same status in Fig. 9.23. From 85 to 120 s into

280

Defrosting for Air Source Heat Pump

(A)

(B)

(C)

(D)

(E)

(F)

Fig. 9.21 Measured temperature of the melted frost collected in cylinders during defrosting. (A) Case 1, (B) Case 2, (C) Case 3, (D) Case 4, (E) Case 5, and (F) 5 Cases.

defrosting, the pressure differences were obviously lower. It further demonstrated that when frost accumulation is around that in Case 2, the system defrosting performance in stability is the best. During defrosting, the energy mainly comes from two sources, the indoor air thermal and the electricity consumed by the compressor and fans. Here, the two energy sources are separately presented in Figs. 9.24 and 9.25. As seen from Fig. 9.24, the peaks of thermal energy taken from the indoor air came at 120 s in Case 1, 115 s

Defrosting control strategy

281

Table 9.5 Durations of melted frost downward flowing away from trays Case no.

Circuit 1 (s)

Circuit 2 (s)

Circuit 3 (s)

Average (s)

Case 1 Case 2 Case 3 Case 4 Case 5

105 115 120 130 135

90 95 110 120 125

95 105 120 120 135

90 95 110 120 130

Fig. 9.22 Measured pressures of compressor suction and discharge during defrosting.

in Case 2, 135 s in Case 3, 130 s in Case 4, and 165 s in Case 5, respectively. The different durations result from their frost accumulations. The peak values are at the range of 25.8–26.8 kJ, with the peak in Case 2 the lowest. In Fig. 9.25, the curves fluctuated at the first 80 s into defrosting due to the initial start up of the system. Then, from 80 to 180 s, the electricity consumption in the five cases all steadily increased. This might be because the energy requirements are increasing, although it is hard to take more thermal energy from the indoor air at this defrosting stage. At the termination of defrosting, the highest values in the five cases are nearly the same, at the range of 3.68–3.85 kJ. Obviously, the curve of Case 2 is always much lower than that of Case 1, or even lower than that of Case 3. This means the electricity consumption in Case 2 is much less, and thus a higher defrosting efficiency is expected. Fig. 9.26 shows the energy supply and effective energy consumption during defrosting in the five experimental cases. These values were calculated with the recorded data, which were considered as the average values in the interval of 5 s during defrosting. As seen, the total energy supplies are 616.8 kJ in Case 1, 692.2 kJ in

282

Defrosting for Air Source Heat Pump

Fig. 9.23 Pressure difference between suction and discharge during defrosting.

Fig. 9.24 Thermal energy taken from the indoor air during defrosting.

Case 2, 743.4 kJ in Case 3, 799.5 kJ in Case 4, and 840.6 kJ in Case 5, respectively. And from Case 1 to Case 5, the total effective energy consumptions are 313.7, 358.5, 374.1, 387.1, and 414.1 kJ. This energy was consumed on melting the frost and vaporizing the residual water. As calculated, the energy coming from the indoor air thermal energy accounts for more than 80% of the total energy supply during defrosting. And the preparation of the energy consumed in melting the frost accounts for more than

Defrosting control strategy

283

Fig. 9.25 Electricity inputs on the compressor and fans during defrosting.

Fig. 9.26 Energy supply and effective energy consumption during defrosting in the five cases.

80% of the total energy consumption during defrosting. Clearly, as the frost accumulation increased, the total energy supply increased correspondingly, but the effective energy consumption was different in the five cases. It is demonstrated that the highest defrosting efficiency in the five cases is not in Case 1 or Case 5. That means there is the most suitable frost accumulation for this experimental ASHP unit, which can improve the defrosting performance the best in the five cases.

284

Defrosting for Air Source Heat Pump

Fig. 9.27 Analysis on frost accumulation and defrosting efficiency.

Fig. 9.27 presents the analysis results on frost accumulation and defrosting efficiency. The frost accumulation increased with the frosting duration, and reached its highest value at 1074 g at 70 min in Case 5. Clearly, the increasing rate was not constant as time, which was much higher at 50–55 min. Although the defrosting duration was prolonged as the increase of frost accumulation, the defrosting efficiency reached its highest value at 51.80% in Case 2. That means that after the frost accumulation is more than 933 g, the defrosting performance becomes worse and worse as the frost accumulation increases. Consequently, the most suitable frost accumulation for good defrosting performance comes out around Case 2. Compared with Case 2, the defrosting duration was prolonged by 29 s, or 16.2%, and the defrosting efficiency decreased by 2.53%, or 5.14%, in Case 5. Although the defrosting efficiency goes up when the frost duration is longer, as shown in Fig. 9.25, the COP of the system during the frosting stage would adversely degrade due to more frost accumulation. Therefore, in this study, the frost accumulation is highly recommended at 933 g for defrosting initiation control. It is also suggested as the reference value for the frosting termination control strategy. In conclusion, an experimental study on time-based defrosting initiation control strategy optimization for an ASHP unit with frost evenly distributed and melted frost locally drained was conducted, and the study results were analyzed and conclusions were reached. To make the frost evenly accumulate on the surface of the multicircuit outdoor coil, an experimental setup was designed and carried out. In the three-circuit outdoor coil, three water-collecting trays were installed under each circuit during defrosting to experimentally validate the frost distribution status. Second, frost accumulation increased as the frosting duration was prolonged, although it is not at a positive proportional relationship. After 50 min into frosting, the frost accumulation increased quickly at first, and then much more slowly. In the same interval of

Defrosting control strategy

285

5 min, the frost accumulation increased 128 g, or 15.9%, from Case 1 to Case 2. However, the increase of frost accumulation was only 73 g, or 7.3%, from Case 4 to Case 5. Additionally, the defrosting duration is also not at a positive proportional relationship with the total frost accumulation. The increased frost accumulation prolonged the defrosting duration, but the main difference comes at the preheating stage during defrosting. When the frost accumulation is 805 g in Case 1, the preheating stage cost 35 s while it takes 96 s for the frost accumulation to reach 1074 g in Case 5. Finally, in view of system stability and indoor thermal comfort, the system performance would be degraded when frost accumulation was more than 933 g. In view of defrosting performance, the defrosting efficiency also reached its peak at 51.80% when the frost accumulation was at 933 g. As demonstrated, frost accumulation is the most fundamental reference parameter for initiation defrosting control in an ASHP unit with the frost evenly distributed and the melted frost locally drained. Meanwhile, the time-based initiation defrosting control was optimized with this method, and thus the potential energy waste was expected to be saved.

9.4

Termination of defrosting control

In practical applications, an RCD operation can be started based on the surface temperature of an outdoor coil, the pressure difference across an outdoor coil, or time. Among them, terminating a defrosting operation based on the surface temperature of an outdoor coil is currently the mostly widely used method [11]. For an outdoor coil with one circuit, the cold refrigerant flows from the downside to the upside along the tube. And thus, it is easy to understand that more frost accumulates on the downside surface. In order to make sure the melted frost was totally vaporized, the temperature sensor is always located at the circuit exit for a one-circuit outdoor coil, as shown in Fig. 9.28A. However, for an outdoor coil used in an ASHP unit, on its refrigerant side, multiple parallel circuits are commonly used for minimized refrigerant pressure loss and enhanced heat transfer efficiency. Also, to save more floor space, the multicircuit outdoor coil is always vertically installed on the outside wall [18]. During RCD, the melted frost would flow downward along the tube wall and fin, which

(A)

(B)

Fig. 9.28 Location of temperature sensors at the two typical outdoor coils. (A) One circuit outdoor coil and (B) Multicircuit outdoor coil.

286

Defrosting for Air Source Heat Pump

always delays the temperature increase for the lowest circuit. Consequently, the temperature sensor is usually placed on the lowest liquid-line circuit of a vertically installed multicircuit outdoor coil, as shown in Fig. 9.28B. In the open literature, an RCD operation will be terminated once a preset temperature, or DTT, is reached for Te, as shown in Fig. 9.28. It was thought that when the tube surface temperature at the exit of the lowest circuit reaches DTT, all the frost has been melted off and the retained water vaporized. It is obvious that when the DTT is set higher, the defrosting duration would be prolonged. Not only is more energy consumed on heating the ambient air, but also the occupants’ thermal comfort is adversely affected [19]. Because the low temperature for the ambient air always comes out at night, sleep thermal comfort is always degraded due to frequent defrosting operations of an ASHP unit. On the contrary, if the DTT is set lower, the frost may be melted off from the outdoor coil surface, but much residual water would be left there. As previously mentioned, the residual water would adversely degrade the system performance during the next frosting/defrosting cycle [20]. However, in the reported RCD experimental studies, temperatures at a wide range of 10–35°C were used as the preset DTTs. Without fixed temperature value or any relatively detailed explanation is given. Although researchers tried their best to improve the defrosting performance of ASHP units at serve cold climate, previous attention was paid to improving the frost evenness values of the outdoor coil, the fin type, the structure, the installation style optimization, the fin surface coating or microtreatment, the refrigerant distribution control strategy exploring, additional heating source searching, etc. With respect to RCD termination, especially for DTT, a scarcity of related research is reported. However, for an ASHP unit with a multicircuit outdoor coil, it is a fundamental and meaningful problem to find a suitable DTT or its range. Consequently, in this paper, an experimental investigation on RCD operation for an ASHP unit with a three-circuit outdoor coil was conducted. First, the detailed description of an experimental ASHP unit and the experimental conditions are introduced. This is followed by presenting the experimental method and data analysis results as well as giving conclusions. This study reports on a method to find a suitable DTT for RCD, and thereby makes contributions for the control strategy optimization for ASHP units.

9.4.1 Methodology As shown in Fig. 9.29, the methodology used in this study is illustrated in the flow chart. In order to carry out this study, collecting the RCD experimental data is the first step. Clearly, an ASHP unit should be first selected, with a vertically installed multicircuit outdoor coil included. Because the frost formation is essentially a transient process with heat and mass transfer that depends on six primary parameters, air temperature, air relative humidity, air velocity, air cleanliness, metal surface temperature, and its property, it is difficult to monitor all those parameters for practical air conditioners or refrigerators. Therefore, the defrosting operation is often executed inaccurately. In addition, the mal-distribution of air and refrigerant into the circuits would make the frost accumulations on the surface of each circuit unequal. The phenomenon

Defrosting control strategy

287

Fig. 9.29 Flow chart for the methodology used in this study.

that the frost accumulation on each circuit’s surface of a multicircuit outdoor coil is different was defined as the uneven frosting phenomenon, which is easily found in the open literature. To clearly describe the uneven frosting phenomenon, a frosting evenness value (FEC) was further named as the ratio of the minimum frost accumulation among the three circuits to the maximum one. In this experimental study, to minimize the effects of uneven frosting on the RCD process and thus on the DTT setting, the FEC should be higher than 90% by adjusting the refrigerant distribution. To solve this problem, manual stop valves should be installed at each circuit. In addition, to make sure of the FEC value, water-collecting trays should be placed under each circuit, collecting and weighing the melted frost flowing downward from each circuit. That means that the multicircuit outdoor coil should be specially made for this study. After the defrosting experimental results were obtained from Step 1, the tube surface temperature at the exit of the lowest circuit was chosen and analyzed in Step 2. With regard to DTTs at the range of 10–35°C in the reported RCD experimental studies, the temperature curve is divided into seven periods with eight nodes. As shown in Fig. 9.30, the seven periods and relative parameters are further listed in Table 9.6. In this study, the T1 is set at 10°C, and the temperature difference at a period is 5°C. At Step 3, the experimental results will be analyzed at each period. All the special time points during defrosting will be collected. Clearly, in Fig. 9.30, each time point will correspond to a temperature value. It is demonstrated that the best RCD termination temperature would be reflected at each operating parameter. In Step 4, after all the special time points are collected, their reflected DTTs will be validated with defrosting efficiency. The time point that corresponds with the highest defrosting efficiency value would be the most suitable DTT for RCD operation for an ASHP unit with a vertically installed multicircuit outdoor coil. Therefore, the study on exploring and concluding a DTT range or value for an ASHP unit with a vertically installed multicircuit outdoor coil is very meaningful and fundamental.

288

Defrosting for Air Source Heat Pump

Fig. 9.30 Description of the divided periods in the tube surface temperature curve.

Table 9.6 The seven divided periods and relative parameters Period

Segment

P1 P2 P3 P4 P5 P6 P7

From From From From From From From

N1 N2 N3 N4 N5 N6 N7

Temperature difference to to to to to to to

N2 N3 N4 N5 N6 N7 N8

T2  T1 T3  T2 T4  T3 T5  T4 T6  T5 T7  T6 T8  T7

¼ 15  10 ¼ 20  15 ¼ 25  20 ¼ 30  25 ¼ 35  30 ¼ 40  35 ¼ 45  40

¼ 5°C ¼ 5°C ¼ 5°C ¼ 5°C ¼ 5°C ¼ 5°C ¼ 5°C

Time difference t2  t1 t3  t2 t4  t3 t5  t4 t6  t5 t7  t6 t8  t7

9.4.2 Experimental cases First, a series of experimental works using the experimental ASHP unit has been carried out to ensure that frost evenly accumulates on the surface of the three circuits. As previously reported, modulating valves installed at an inlet refrigerant pipe to each circuit were deployed to vary the refrigerant flow to each circuit, and thus the frost accumulations on each circuit were adjusted. Using this method, a set of fixed valve opening degrees was obtained after many trial-and-error experimental works. As listed in Table 9.7, this experimental case is set as Case 1, with the FEC calculated higher than 90%. The airside surface conditions at the initiation and end of the frost

Defrosting control strategy

289

Table 9.7 Two frosting experimental cases Item

Parameters

Case 1

Case 2

1 2 3 4 5

Circuit number Frosting duration Water-collecting trays FEC Purpose

3 60 min With >90% (calculated) Adjusting FEC

3 60 min Without >90% (adjusted) Analyzing DTT

Fig. 9.31 Airside surface conditions at the initiation and end of the frost experiment. (A) Outdoor coil, (B) start of frost experiment, and (C) end of frost experiment.

experiment as well as the three-circuit outdoor coil are shown in Fig. 9.31. As compared Fig. 9.31A with Fig. 9.31B, there is few frost formed at the initiation of frosting experiment, which formed during cooling the outdoor frosting space in the environmental chamber. In Fig. 9.31C, much frost accumulated, which is visually evenly distributed for each circuit. To clearly present the previous three photos’ taken time, in Fig. 9.32, different durations at a frosting/defrosting experimental cycle are further illustrated. In this experimental cycle, about 180 min were consumed on cooling the outdoor frosting space to reach the outdoor experimental conditions (0.5
0.2°C, 90
3% RH). The outdoor coil photos shown in Fig. 9.31A and B were taken at the initiation and end of this period, as shown in Fig. 9.32. After the frosting experiment terminated, there are two compressor shutdown periods to keep the compressor safe before and after the defrosting operation. The airside surface condition shown in Fig. 9.31C was taken at the termination of the 60 min frosting operation. Second, in Case 2, the water-collecting trays are taken away in the RCD operation. In the two cases, their defrosting durations are both at 60 min. Therefore, a traditional defrosting process was conducted with this experimental ASHP unit. Also, the FEC at

290

Defrosting for Air Source Heat Pump

Fig. 9.32 Different durations at a frosting/defrosting experimental cycle.

higher than 90% was ensured. The experimental results in Case 2 could be analyzed for the most suitable DTT for RCD in this study, as listed in Table 9.7.

9.4.3 Data analysis and validation (1) Tube surface temperature analysis

Fig. 9.33 shows the measured tube surface temperature at the exit of each circuit in the three-circuit outdoor coil. It is obvious that the curve of Circuit 3 is always the lowest. This results from the negative effects of melted frost downward flowing along the surface of the outdoor coil during defrosting. In this figure, for its time (horizontal) axis, 80 s is the chosen starting time in order to clearly show the temperature rise during defrosting. Using the methodology described, the seven periods are divided and summarized in Table 9.8. From 10°C to 45°C, the eight nodes correspond to the times of 151 s for N1, 158 s for N2, 172 s for N3, 189 s for N4, 211 s for N5, 235 s for N6, 266 s for N7, and 309 s for N8, respectively. As calculated in Table 9.8, the seven time differences are kept gradually increasing from 7 s for P1 to 43 s for P7. (2) Experimental results and their special points

The other measured operating performances of the experimental ASHP unit during RCD are presented in Figs. 9.34–9.41. Fig. 9.34 presents the measured fin surface temperature at the center of the three refrigerant circuits. Figs. 9.35 and 9.36 show the measured temperatures of compressor suction and discharge, and the measured temperatures of the tube surface at the entrance and exit of the outdoor coil during defrosting, respectively. Figs. 9.27 and 9.38 show the total energy and electricity inputs coming from the system outside, respectively. The measured refrigerant volumetric flow rate and measured temperature of the air surrounding each circuit during

Defrosting control strategy

291

Fig. 9.33 Measured tube surface temperature at the exit of the three outdoor coil circuits.

Table 9.8 Experimental periods and nodes during RCD Period

Segment

P1 P2 P3 P4 P5 P6 P7

From From From From From From From

N1 N2 N3 N4 N5 N6 N7

to to to to to to to

N2 N3 N4 N5 N6 N7 N8

Tube surface temperature (°C)

Time difference (s)

10–15 15–20 20–25 25–30 30–35 35–40 40–45

158 151 ¼ 172 158 ¼ 189 172 ¼ 211 189 ¼ 235 211 ¼ 266 235 ¼ 309 266 ¼

7 14 17 22 24 31 43

defrosting are shown in Figs. 9.39 and 9.40, respectively. Finally, Fig. 9.41 presents the airside surface conditions of the outdoor coil during defrosting. The same as the tube surface temperature figure, in order to clearly show the parameters’ fluctuating states during defrosting in Figs. 9.34–9.39, for their time (horizontal) axis, 80 s is the chosen starting time. In addition, considering the DTTs used in the reported experimental studies at the range of 10–35°C, only the important periods will be analyzed. Therefore, the data analysis work would focus on the periods of P1 to P5. Fig. 9.34 shows the measured fin surface temperature at the center of the three circuits during defrosting, from 80 to 160 s. The locations of the fin surface temperature sensors are shown in Fig. 9.28. It is easy to find that the increasing trends of the three curves are similar to the tube surface temperature curves, as shown in Fig. 9.33. However, there are two differences between the tube surface temperature and the fin surface temperature trends. First, the fin curves are a little later because of the heat

292

Defrosting for Air Source Heat Pump

Fig. 9.34 Measured fin surface temperature at the center of each circuit of the outdoor coil.

Fig. 9.35 Measured temperatures of compressor suction and discharge during defrosting.

transfer process delay from the refrigerant to the tube, and then to the fins. Second, the curve of Circuit 3 is not always the lowest in Fig. 9.34, which is different from that shown in Fig. 9.33. This is because of the disturbing effects of the downward-flowing melted frost, which will cool the temperature sensors by taking latent heat from them. As listed in Table 9.9, the fin surface temperature differences at the first five divided

Defrosting control strategy

293

Fig. 9.36 Measured temperatures of the tube surface at the entrance and exit of the outdoor coil during defrosting.

Fig. 9.37 Total energy coming from outside the system during defrosting.

periods are 3.6°C for P1, 6.1°C for P2, 6.6°C for P3, 4.7°C for P4, and 4.0°C for P5, respectively. Clearly, at P3, from 172 to 189 s, the temperature difference is the biggest. In fact, this period lasts about 17 s, which is not the longest one in the five periods. Therefore, it is demonstrated that the best defrosting termination is at this period.

294

Defrosting for Air Source Heat Pump

Fig. 9.38 Total electricity inputs coming from outside the system during defrosting.

Fig. 9.39 Measured refrigerant volumetric flow rate during defrosting.

The measured temperatures of compressor suction and discharge and the temperatures of the tube surface at the entrance and exit of the outdoor coil during defrosting are separately shown in Figs. 9.35 and 9.36. As analyzed in Fig. 9.35, the compressor suction temperature will increase to its highest value, 2.6°C, at 170 s into defrosting, and then decrease to its lowest value, 1.4°C, at 190 s into defrosting. It is demonstrated that this period is special, and the most suitable DTT should be at this

Defrosting control strategy

295

Table 9.9 Fin surface temperature differences at the first five divided periods Period

Segment

P1 P2 P3 P4 P5

From From From From From

N1 to N2 to N3 to N4 to N5 to

N2 N3 N4 N5 N6

Time difference (s)

Fin surface temperature difference (°C)

158  151 172  158 189  172 211  189 235  211

9.8  6.2 ¼ 3.6 15.9  9.8 ¼ 6.1 22.5  15.9 ¼ 6.6 27.2  22.5 ¼ 4.7 31.2  27.2 ¼ 4.0

¼7 ¼ 14 ¼ 17 ¼ 22 ¼ 24

fluctuation period, 170–190 s into defrosting. Also, in Fig. 9.36, at this period, there is the highest value for the temperature of the outdoor coil entrance, 42.27°C at 175 s into defrosting. This is a special point during defrosting, which is collected for the further step of data analyzing work. During defrosting, there are three energy sources from the system outside: (1) electricity inputs of the compressor, (2) electricity inputs of the indoor air fan, and (3) thermal energy from the indoor air. The total energy and electricity inputs coming from outside the system during defrosting are shown in Figs. 9.37 and 9.38, respectively. In Fig. 9.37, the total energy curve is steeply increasing at the first half, and then slowly decreasing around 27°C. This is reasonable because most of the energy consumed in defrosting is increasing from melting the frost to vaporizing the melted frost stage. However, when it comes to the later stage, only a little of the residual water is vaporized, and most of the energy is consumed on heating the cold ambient air. This is the reason why the curve slowly decreases with time. As analyzed, it reaches its peak at a value of 29.3 kJ at 165 s into defrosting. That means that at that moment, the energy requirement for defrosting reaches its highest value. Therefore, after that moment, the energy consumption is decreasing, and the defrosting duration reaches its termination. In Fig. 9.38, the curve of electricity inputs is always increasing, which means the electricity consumption is continuing as the compressor and indoor air fan are running. However, it is clear that from point A to point B, the curve keeps increasing quickly. There is also a special time at 175 s, point C shown in Fig. 9.38, with the biggest distance from the line AB, at a value of 0.06 kJ. It is also demonstrated that this special point corresponds to the most suitable DTT. Fig. 9.39 shows the measured refrigerant volumetric flow rate during defrosting. It increased from 80 to 175 s, and then decreased suddenly. This special point is very obvious, and the peak value was about 1.466 L/min, at 175 s into defrosting. This results because a lot of energy was consumed on melting the frost and vaporizing the melted frost, which made the refrigerant’s phases change from vapor to liquid status. The pressure decreasing at the outdoor coil increased the volumetric flow rate of the refrigerant at the section from the EEV to the outdoor coil. Thereafter, the refrigerant volumetric flow rate kept fluctuating between 1.1 and 1.2 L/min. That means the energy consumption was decreased and kept at a nearly constant value at the later process of defrosting. Therefore, it is demonstrated that the most suitable DTT is at around 175 s, or at the P3.

296

Defrosting for Air Source Heat Pump

Fig. 9.40 Measured temperature of the air surrounding each circuit during defrosting.

At the later stage of the defrosting process, as the retained melted frost is vaporized, the ambient air surrounding the outdoor coil was heated by the bold fins. It is easy to find and understand that there would be a lot of white fog near the three circuits at that time. Here, Fig. 9.40 shows the measured temperature of the air surrounding each circuit during defrosting, and Fig. 9.41 presents the white fog conditions in the airside of the outdoor coil during defrosting. Clearly, the air temperature curves kept increasing slowly from 0 to 125 s into defrosting because there was not any bold fin or tube at this period. But after 125 s, their temperatures increased steeply, especially after N3 for the temperature curves of the air surrounding Circuits 1 and 2. This is because the hot air became lighter and rose at the heating stage. At the same time, as shown in Fig. 9.39, there were many white fog circles on the pictures after N3. It is not as obvious as that in a movie because it is much easier for our eyes to capture the moving white fog. But they truly exist, and the temperature curves shown in Fig. 9.40 also proved this truth. In addition, as shown in Fig. 9.40, the first decrease for the air temperature was at 185 s into defrosting. That means that at or after this moment, there would be a lot of energy consumed on heating the ambient air, which would also efficiently decrease the defrosting efficiency. Therefore, it is demonstrated that the most suitable DTT is at the P3, or a little earlier than 185 s into defrosting. Table 9.10 summarizes the DTT periods and nodes from the previously analyzed special points in different defrosting parameters for the RCD of the experimental ASHP unit. From these special values in Figs. 9.34–9.41, it seems that the most suitable DTT is at the P3, or at 175 s. When it reflects at the tube surface temperature at the exit of the lowest circuit, they are at the range of 20–25°C, or at 22°C. At the same time, the DTTs and their corresponding average fin surface temperatures are shown in Fig. 9.42. When it comes to the average fin surface temperature, the fixed DTTs are at 17.2–23.2°C, or at 18.5°C.

Defrosting control strategy

297

Fig. 9.41 White fog conditions in the airside of the outdoor coil during defrosting. (A) Node 1, (B) Node 2, (C) Node 3, (D) Node 4, (E) Node 5, and (F) Node 6.

Table 9.10 DTT phase and node from different parameters Item

Parameter

DTT period or point (s)

1 2

Total energy coming from outside the system Total electricity inputs coming from outside the system Compressor suction temperature Average of fin surface temperature Photo of white fog at the airside surface Temperature of air surrounding outdoor coil Tube surface temperature at the exit of the outdoor coil Measured refrigerant volumetric flow rate

165– 170–

3 4 5 6 7 8

170–190 172–189 172–189 172–185 175 175

298

Defrosting for Air Source Heat Pump

Fig. 9.42 DTTs and their corresponding average fin surface temperatures. (3) Validation and discussions

Defrosting efficiency can be used to evaluate the performance of a defrosting operation, which is also used as the validation index for the best DTT in this study. As defined, defrosting efficiency is the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation, as follows: ηd ¼

Em + E v  100% Qcomp + Qid, fan + Qid,air

(9.1)

where Em and Ev are the total heat used for melting frost and vaporizing the retained water, respectively, and they are evaluated by: Em ¼ Mf Lsf Lsf ¼ 334 kJ=kg Ev ¼ Mv Lv ðLv ¼ 2443kJ=kgÞ




(9.2) (9.3)

where Mf and Mv are the total mass of the frost formed on the outdoor coil and the mass of vaporized melted frost, respectively, and Lsf and Lv are the latent heat of frost melting and the latent heat of the evaporation of water, respectively. Also in Eq. (9.1), Qcom, Qfan, and Qair are the energy consumptions by the compressor and supply fan, and the thermal energy from the indoor air during defrosting, respectively. Fig. 9.43 shows the calculated defrosting efficiency for N1–N6 and the time point of 175 s into defrosting as well as the corresponding values of the energy supply for

Defrosting control strategy

299

Fig. 9.43 Energy analysis and defrosting efficiency calculated at six nodes and 175 s.

defrosting (Em + Ev) and energy consumption during defrosting (Qcom + Qfan + Qair). It is obvious that, at 175 s, the defrosting efficiency reaches its peak, at about 60.6%. Clearly, it is demonstrated that, at 175 s, the corresponding DTT is the best value. When the RCD process for the experimental ASHP unit terminates at this temperature, the defrosting efficiency will reach the highest value. And thus, the system will save the most energy during defrosting, as compared to terminating the RCD at other temperatures. Here, the standard of the DTT chosen work is the defrosting efficiency reaching the highest value. However, a frosting/defrosting cycle should be comprehensively considered. That means there is a best time for the best comprehensive energy performance for a frosting/defrosting cycle, which may not be the same as the fixed time in this study. Also, the DTT would be different from the value in this study. In fact, it is hard to evaluate the whole cycle because the frosting and defrosting processes are affected by too many factors. As shown in Fig. 9.32, the different durations contained in a frosting/defrosting experimental cycle are more problems in accurately analyzing it. Therefore, it becomes the limit for our previous work, and is a fundamental and interesting point for further study. In conclusion, an experimental study on an ASHP unit with a vertically installed three-circuit outdoor coil for its RCD termination temperature was undertaken and the study results are reported. To find the most suitable DTT for an ASHP unit during its RCD, a three-circuit outdoor coil was tailor-made and carried out. Basing on the even frosting status, the defrosting parameters were collected and effectively analyzed. The methodology for a suitable DTT or its range was first proposed and conducted by using the tube surface temperature curve as the baseline and analyzing all other defrosting parameters. The found suitable DTTs were validated by defrosting

300

Defrosting for Air Source Heat Pump

efficiency comparison. As demonstrated, the most suitable DTT for the experimental ASHP unit with the three-circuit outdoor coil is at the range of 20–25°C, or around 22°C. At this period, the average fin surface temperature is at the range of 17.2–23.2°C, or around 18.5°C. The special points in the measured tube surface temperature at the exit of the outdoor coil and the measured refrigerant volumetric flow rate are obvious. It is demonstrated that the two parameters can work as the defrosting control indexes in the future experimental study. From the view of only at the defrosting stage and in an entire frosting/defrosting cycle, the best DTTs for reverse cycle frosting were comprehensively discussed based on their energy performances.

9.5

Concluding remarks

In this chapter, experimental studies on the time-based defrosting initiation control strategies and the DTT for an ASHP unit with a multicircuit outdoor coil are presented, with the following conclusions: (1) The optimization of the time-based defrosting initiation control strategy for an ASHP unit with different frost accumulations evenly distributed on the surface of the outdoor coil was experimentally investigated, with melted frost locally drained or not. (2) Frost accumulation increases as the frosting duration is prolonged, but without a positive proportional relationship. The defrosting duration is not at a positive proportional relationship with frost accumulation. (3) In view of system stability and indoor thermal comfort, the system performance would be degraded when frost accumulation was more than 930 and 933 g for conditions of melted frost locally drained or not, respectively, in this study. The defrosting efficiency also reached its peak at this frost accumulation value. As demonstrated, frost accumulation is the most fundamental reference parameter for frosting termination in an ASHP unit. (4) A methodology for confirming a suitable DTT for RCD was proposed and experimentally examined. This method was based on the most energy saving for the ASHP unit, and realized by using the tube surface temperature curve as the baseline. Here, the suitable DTTs were also validated by a defrosting efficiency comparison. It is a fundamental study for optimizing defrosting control strategies for ASHP units.

References [1] Wang W, Xiao J, Guo QC, Lu WP, Feng YC. Field test investigation of the characteristics for the air source heat pump under two typical mal-defrost phenomena. Appl Energy 2011;88(12):4470–80. [2] Jha SK, Bilalovic J, Jha A, Patel N, Zhang H. Renewable energy: present research and future scope of Artificial Intelligence. Renew Sustain Energy Rev 2017;77:297–317. [3] Wang W, Feng YC, Zhu JH, Li LT, Guo QC, Lu WP. Performances of air source heat pump system for a kind of mal-defrost phenomenon appearing in moderate climate conditions. Appl Energy 2013;112:1138–45. [4] Iragorry J, Tao YX. Frost temperature relations for defrosting sensing system. J Heat Transf 2005;127:344–9.

Defrosting control strategy

301

[5] Lawrence JMW, Evans JA. Refrigerant flow instability as a means to predict the need for defrosting the evaporator in a retail display freezer cabinet. Int J Refrig 2008;31:107–12. [6] Suleyman Yigit K, Metin Ertunc H. Prediction of the air temperature and humidity at the outlet of a cooling coil using neural networks. Int Commun Heat Mass Transf 2006;33 (7):898–907. [7] Kim MH, Lee KS. Determination method of defrosting start-time based on temperature measurements. Appl Energy 2015;146:263–9. [8] Da Silva DL, Hermes CJL, Melo C. First-principles modeling of frost accumulation on fan-supplied tube-fin evaporators. Appl Therm Eng 2011;31(14–15):2616–21. [9] Ye HY, Lee KS. Performance prediction of a fin-and-tube heat exchanger considering airflow reduction due to the frost accumulation. Int J Heat Mass Transf 2013;67:225–33. [10] Qu ML, Xia L, Jiang YQ, Deng SM. A study of the reverse cycle defrosting performance on a multi-circuit outdoor coil in an air source heat pump-part I: experiments. Appl Energy 2012;91:122–9. [11] Song MJ, Xia L, Mao N, Deng SM. An experimental study on even frosting performance of an air source heat pump unit with a multi-circuit outdoor coil. Appl Energy 2016; 164:36–44. [12] Steiner A, Rieberer R. Parametric analysis of the defrosting process of a reversible heat pump system for electric vehicles. Appl Therm Eng 2013;61:393–400. [13] Zhang L, Jiang YQ, Dong JK, Yao Y, Deng SM. An experimental study of frost distribution and growth on finned tube heat exchangers used in air source heat pump units. Appl Therm Eng 2018;132:38–51. [14] Amer M, Wang CC. Review of defrosting methods. Renew Sustain Energy Rev 2017; 73:53–74. [15] Song MJ, Mao N, Deng SM, Xia YD, Chen Y. Y, An experimental study on defrosting performance for an air source heat pump unit at different frosting evenness values with melted frost locally drainage. Appl Therm Eng 2016;99:730–40. [16] Song MJ, Deng SM, Xia L. A semi-empirical modeling study on the defrosting performance for an air source heat pump unit with local drainage of melted frost from its three-circuit outdoor coil. Appl Energy 2014;136:537–47. [17] Song MJ, Xia L, Deng SM. A modeling study on alleviating uneven defrosting for a vertical three-circuit outdoor coil in an air source heat pump unit during reverse cycle defrosting. Appl Energy 2016;161:268–78. [18] Song MJ, Wang ZH, Mao N, Li Z, Chen Y. An experimental study on the uneven refrigerant distribution over a vertically installed multi-circuit outdoor coil in an air source heat pump unit during reverse cycle defrosting. Appl Therm Eng 2015;91:975–85. [19] Qu ML, Xia L, Deng SM, Jiang YQ. Improved indoor thermal comfort during defrost with a novel reverse-cycle defrosting method for air source heat pumps. Build Environ 2010;45 (11):2354–61. [20] Li LT, Wang W, Sun YY, Feng YC, Lu WP, Zhu JH, Ge YJ. Investigation of defrosting water retention on the surface of evaporator impacting the performance of air source heat pump during periodic frosting-defrosting cycles. Appl Energy 2014;135:98–107.

Technoeconomic performances 10.1

10

Introduction

To further improve the RCD performance for an ASHP unit, different studies were conducted globally, such as heating and/or dehumidifying the inlet air of the outdoor coil, structure and dimension adjustments for the outdoor coil, fin type and surface treatment, additional defrosting energy supply with a PCM-TES system, FEC improvement on the surface of the outdoor coil, control strategy optimization via refrigerant distribution adjustment, etc. For an outdoor coil in an ASHP unit, a multicircuit structure is usually used in order to enhance its heat transfer and minimize its refrigerant pressure loss. To save the floor space, the multicircuit outdoor coil is always vertically installed in its practical application. For an ASHP unit with a vertically installed multicircuit outdoor coil, it was easy to find the uneven defrosting phenomenon in the open literature. When the other circuit was waiting for the lowest circuit to terminate its defrosting process, the heat transfer between the hot refrigerant tube and fins and the ambient cold air would consume a lot of energy [1]. Not only would defrosting efficiency be degraded, but also the defrosting duration would be prolonged and the indoor thermal comfort level affected. As demonstrated by previous experimental and numerical studies, the downward flowing of melted frost along the surface of the outdoor coil was one of the important reasons for uneven defrosting [2, 3]. At the same time, a lower FEC would also affect RCD. It was reported that when the FEC was increased from 82.6% to 96.6%, the defrosting efficiency could increase from 42.0% to 48.7%. After the negative effects of melted frost flowing down were eliminated, the defrosting efficiency was increased by about 5.7% as the FEC was increased from 79.4% to 96.6%. Furthermore, it was proved that frosting COP was increased from 4.10 to 4.26 as the FEC was increased from 75.7% to 90.5% [4]. Among all the previous experimental studies, a series of valves was used to adjust the refrigerant distribution, according to the tube surface temperatures at the circuit exits. However, for a new technology or innovation, a technoeconomic analysis is very important and should always be carried out before its wide application [5–7]. The total cost of technoeconomic proposed new ASHP unit is increased by the additional investment of valves, and a longer payback period is expected. To solve this problem, the economy has to be improved based on the characteristics of the ASHP unit. Although many methods were used to improve the operating performance of an ASHP unit [8, 9], especially to optimize its RCD performance [10, 11], only a few studies with a technoeconomic analysis were reported [12]. Horton et al. gave an economic analysis when evaluating a high-performance cold-climate heat pump [13]. As reported, the maximum additional cost of the system changes for the Minneapolis location in the United States was $430 for a vapor-injected system and $391 for an oil-flooded system. These estimates were based on an assumed 3-year simple payback Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00010-2 © 2019 Elsevier Ltd. All rights reserved.

304

Defrosting for Air Source Heat Pump

period that was accepted by the customer. Using a lifetime of 20 years for groundcoupled and air-coupled heat pump systems, their performances were compared based on the experimental COP results by Esen et al. [14, 15]. The economic analysis clearly shows that the ground-coupled system is economically preferable to the air-coupled one. In addition, Dong et al. discussed the economy of an ASHP unit with a PCM-TES system added to improve its RCD performance. It was concluded that before the replacement of the PCM in the TES system, using the novel RCD method, the running cost could save approximately $97.47 over 7 years of service life. Recently, a technoeconomic analysis of ASHP applied to space heating in northern China was also carried out, with the pollution emission considered [16]. Although the authors neglected the unavoidable frosting/defrosting problem in northern China, the lowtemperature ASHP heating system was demonstrated to have a better economical performance than a coal-fired boiler, a gas boiler, a direct electric heating mode, or combined heat and power generation systems. Many ASHP unit studies on energy performance improvement have been reported, although only a few of them considered economic performance [17, 18]. To analyze the economic performance of the new ASHP unit, an economic analysis on its novel frosting/defrosting operations is given in this chapter. First, the two conditions, with and without the valves installed in the multicircuit outdoor coil of an ASHP unit, were analyzed, with the fundamental frosting and defrosting state assumptions clearly listed. This is followed by an investigation on the four working conditions of the ASHP unit, with the water-collecting trays placed under each circuit to take away the melted frost before flowing into the downside circuit further considered. This study is helpful for the relative technologies applied in the industry and the real market. The conclusions given in this chapter can also play a role in the pricing or governmental subsidy of ASHP units.

10.2

The influence of the refrigeration adjustment valve

To analyze the economic performance of the new ASHP unit, with valves installed in its multicircuit outdoor coil, an economic analysis on its novel frosting/defrosting operations is given in this study. First, the results of the designed frosting/defrosting experiments will be presented, as well as a series of assumptions given. To clearly show the frosting and defrosting influence on the economic performance of the ASHP unit, a typical city in severe cold regions in China will be chosen as an example. This is followed by the development of economic analysis equations, which will be used in cases with and without valves installed. After the results are calculated in this study, the proportion of costs at different stages and payback periods will be analyzed.

10.2.1 Methodology In order to carry out this economic analysis, the methodology was first illustrated in Fig. 10.1. As shown, the first step is an experimental study with an ASHP unit selected and a three-circuit outdoor coil specially made. Then, the two typical conditions were

Technoeconomic performances

Experimental study • An ASHP unit selected • A three-circuit outdoor coil made • Two typical conditions designed • Frosting/defrosting experiments undertaken

Techno-economic evaluation • • •

Running cost analysis Total cost evaluation as operating year Discussions on payback periods

305 Experimental results & assumptions • Frosting operation data collected: COP, indoor air thermal energy supplied • Defrosting experimental results obtained: Defrosting duration, power inputs, and indoor air thermal energy consumed • Four types assumptions given Economic analysis & energy equations • First cost: ASHP unit, valves and trays • Running cost: Stage 1, Heating season with frost formation Stage 2, Heating season without frost formation Stage 3, Cooling season

Fig. 10.1 Flow chart for the methodology used in this study.

designed, considering the installation of valves or not. After the experimental procedures were confirmed, a series of frosting and defrosting experiments was undertaken with the frost accumulated at different FECs. Thereafter, these experimental results and calculations, such as COP, indoor air thermal energy supplied at the frosting operation, defrosting duration, power inputs, etc., could be used as the known parameters for the later technoeconomic evaluations. In addition, many assumptions would be given at this step. The third step is the development work of equations, including the first and running costs at different operating stages. At the fourth step, evaluation results and relative discussions are given. Finally, the technoeconomic analysis on the frosting/defrosting operations for the new ASHP unit is undertaken.

10.2.1.1 Experimental study (1) Experimental case studies

An experimental ASHP unit was specifically established for carrying out the experimental work. The experimental ASHP unit was a split-type one and it consisted of a swing-type compressor, an accumulator, a four-way valve, an EEV, an indoor coil, and an outdoor coil. It was modified from a commercially variable speed Daikin ASHP unit, with its performance parameters listed in Table 10.1. In cooling mode, the rated capacity was 5.2 kW, the rated power consumption 1600 kW, and the rated COP 3.25. In heating mode, the rated capacity was 6.5 kW, the rated power consumption 1840 kW, and the rated COP 3.53. In Harbin, the price of this unit was about $1200. In the experimental ASHP unit, the indoor coil used was the prototype, with its model name of FTXD50FVM. However, to calculate the FECs in the experiments, water-collecting trays were designed and located between the circuits. Thus, considering the installation of valves and trays, a three-parallel refrigerant circuit outdoor coil was specially designed and carried out. Three solenoid valves and three manual stop valves were installed at the entrance and exit of each circuit in the outdoor coil. The solenoid valves can stop the refrigerant flowing into/out of a circuit, and the manual stop valves were used to adjust their flow rates. Using this method, the frost

306

Defrosting for Air Source Heat Pump

Table 10.1 Parameters of the ASHP unit used in this study Item

Parameters

Value

Unit

1 2 3 4 5 6 7 8 9

Rated cooling capacity Rated heating capacity Rated cooling power consumption Rated heating power consumption Rated cooling COP Rated heating COP Brand Model name of indoor unit Total price in Harbin, China

5.2 6.5 1600 1840 3.25 3.53 Daikin FTXD50FVM $1200

kW kW kW kW – – – – USD

accumulation on the surface of a circuit could be changed, and thus the FEC can be improved. In addition, under each circuit, water-collecting trays were placed by which the melted frost could be collected. After the frost that accumulated on the surface of each circuit melted, the water flowed into a measuring cylinder, which contacted the corresponding tray. And thus, these frosts could be weighed and calculated. With this method, as reported in previous energy studies [2–4], the frosting and defrosting performances of an ASHP unit could be effectively improved after the valves were installed. The experimental ASHP unit was installed in an existing environmental chamber having a simulated heated indoor space and a simulated outdoor frosting space. The sizes of both spaces were each measured at 3.8 m (L)
3.8 m (W)
2.8 m (H). The three-circuit outdoor coil was installed in the outdoor frosting space, where the frosting/defrosting cycle operations were carried out. Detailed experimental procedures and conditions were previously reported. In a frosting/defrosting cycle, the frosting operation came out first. After the valves were installed, the frosting performance would be effectively optimized. Therefore, frosting experiments should be first undertaken before the defrosting experiments in this study. As listed in Table 10.2, two frosting experimental cases were designed. Valves and trays were both installed, and thus different FECs, FEC1 and FEC2, could be reached. The frosting duration was fixed at 60 min. With this experimental study, the frosting operating performances of an ASHP unit at different FECs could be obtained. Valves in Case F1 were fully open, and the FEC1 was less than 100% due to the refrigerant and air being unevenly distributed. Differently, the valves in Case F2 were evenly adjusted, and thus the FEC2 was nearly 100%. Table 10.2 Two frosting experimental cases Case No.

Case F1

Case F2

Valves (Status) FEC Frosting duration Condition shown in Fig. 10.2

With (Fully open) FEC1 ( D4

0 0

1

2

3

4

5

6

7

8

5178.03 (6.68%)

29,530.42 20,000 10,000

3402.02 (6.26%)

31,156.43 30,000

1626.01 (5.22%)

Total cost (CNY)

77,469.29

Case D3 Case D4 Case D3 (Practical) Case D4 (Practical)

80,000

9 10 11 12 13 14 15 16

System operating duration (year)

Fig. 10.11 Total costs in two cases in the 15 operating years.

which is about 7.64% of the $3359.26 in Case D3. Also, the economic performance of the new ASHP unit with valves installed was improved in this season. It is obvious that the difference value and ratio in this season were both less than those in the heating season with frost formation. This reflects that the effects of valves on the economic performance of an ASHP unit in heating season with frost formation were higher than

Technoeconomic performances

321

380

Ration of differences (CNY/year)

360

5.00

340

15.00

320 300 120.00

280 260

First 5 Years Increase rapidly

240

Second 5 Years Increase slowly

Third 5 Years Nearly no increase

220 200 0

1

2

3

4 5 6 7 8 9 10 11 12 13 14 15 16 System operating duration (year)

Fig. 10.12 Ratio of total cost differences between the two cases in the 15 operating years.

Proportation of first cost in total cost (%)

100 Case D3 Case D4

90 80 70 60 50 40

D4 > D3

30 20 10 1.9 2.8

0 0

1

2

3

4.5 4

6.9

7.6

5 6 7 8 9 10 11 12 13 14 15 16 System operating duration (year)

Fig. 10.13 Proportion of first cost in total cost in the 15 operating years.

that in heating season without frost formation. This phenomenon results because the valves have more effects on energy performance improvement at the defrosting operation stage in the heating season with frost formation. Fig. 10.9 shows the running costs in heating season with frost formation and heating season without frost formation in the 15 operating years. As shown, the total running costs in the two heating seasons are $8084.04 in Case D3 and $7285.18 in

322

Defrosting for Air Source Heat Pump

Proportion of additional first cost in total cost (%)

2.0

1.84%

1.8 1.6 1.4 1.2 1.0 0.8 0.6

0.5%

0.4

0.3% 0.2%

0.2 0.0

0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16

System operating duration (year) Fig. 10.14 Proportion of additional first cost in total cost in the 15 operating years.

Case D4. The total difference was about $798.89, which accounts for about 9.88% of the total running costs in the heating seasons in Case D3. Obviously, after the valves were installed on the multicircuit outdoor coil, the total running costs in the two heating seasons of the new ASHP unit could save nearly 10%. In the total running cost in two cases, the running cost in heating season without frost formation accounts for about 40%, and heating season with frost formation about 60%. This results from their operating durations, at 16,200 h and 21,459.38–21,290.63 h. The electricity costs on frosting were 28,893.29 CNY in Case D3 and 26,035.71 CNY in Case D4, respectively, with a difference of 2857.58, or 9.89% of the cost in Case D3. However, the ratios of the electricity cost on frosting in the total running costs in the heating season with frost formation were nearly the same, at 53.89% in Case D3 and 53.58% in Case D4. Therefore, after the valves were installed, the running costs at all stages were decreased. The difference was not at any one stage. Fig. 10.10 shows the accumulated total costs in the two cases in the 15 operating years. The total values were $11,615.28 in Case D3, and $10,838.93 in Case D4, respectively. Their difference was $776.36, which accounts for about 6.68% of the total cost in Case D3. This directly reflects that the economic performance of the new ASHP could be effectively improved after valves were installed on the outdoor coil. With nearly the same first costs, the total costs in the two cases were mainly decided by their running costs. The running cost in Case D3 was $10,416.12, but $9617.25 in Case D4. That means the running cost could decrease by as much as $798.87, or 7.67%, after the valves were installed. In the two cases, the running costs in cooling season were the same, but the ratio increased from 20.08% to 21.51% after

Technoeconomic performances

323

the valves were installed. Therefore, the valve effect on cooling operation performance should be further studied. Fig. 10.11 shows the total costs in the two cases in the 15 operating years. As the fundamental assumptions show, the total costs kept increasing in line. As shown in Fig. 10.11, after 1 operating year, the running cost relationship of the two cases was at D3 > D4. That means the payback of the first cost is less than 1 year. When the system operations were 5, 10, and 15 years, the total costs in the two cases were 31,156.43 CNY and 29,530.42 CNY, 54,312.86 CNY and 50,910.84 CNY, and 77,469.29 CNY and 71,747.93 CNY, respectively. Their differences and ratios in the values of Case D3 were $243.85 and 5.22%, $510.19 and 6.26%, and $776.54 and 6.68%, respectively. Clearly, their differences and ratios were both increasing with the operation duration. However, in practical application, the maintenance costs increased the total running cost, which was assumed to be about 10% of the total first cost per year added. Also, the operating performances of the ASHP unit would decrease as time goes by, which was also assumed at a 10% per year range. The practical total costs’ curves in the two cases were also shown in Fig. 10.11, using dash lines. Clearly, the difference between the two cases would be increased a lot as time goes by. This figure plays an important role in the budget guiding work of equipment procurement. To clearly show the difference of total costs in the two cases and its change as time progresses, the ratio of the difference was calculated and shown in Fig. 10.12. This value was the ratio of the total cost difference in the two cases and the system operating duration. As shown, at the first 5 years, the ratio was increasing rapidly. The difference between years 1 and 5 was about 120.00. At the second five operating years, the difference became to about 15.00. The curve kept increasing slowly. However, the difference was only 5.00 at the third 5 operating years. The ratio was nearly not increased. That means that the valves effects on the total cost were mainly shown at the first 5 years. When an ASHP unit worked after 10 years, the total cost still saved a lot after the ASHP unit had installed valves. However, the effects on economic performance were not as obvious as the first years. Fig. 10.13 shows the proportion of the first cost in the total cost in the two cases in their 15 operating years. At the life span of an ASHP unit, their relationship showed D4 > D3. This also confirmed that the installation of valves could save more money in an ASHP unit’s application. As shown in this figure, when the proportions were 50%, 40%, 30%, and 20%, the operating time was about 1.9 years, 2.8 years, 4.5 years, and 7.6 years, respectively. That means that after the ASHP unit worked 8 years, the first cost only accounted for less than 20% of the total cost. In Case D3, the duration was about 6.9 years. That means that using the same running cost, the system can work 0.7 years longer in Case D4. When it works 15 years, the first cost becomes only about 10% of the total cost. It is also confirmed that the total cost is mainly decided by the running cost. To clearly show the additional first cost effect on the total cost during the operation of the new ASHP unit, Fig. 10.14 shows the proportion of additional first cost in total cost in the 15 operating years. As shown, after the new ASHP unit works 5 years, 10 years, and 15 years, the proportions of additional first cost in total cost changed from 1.84% to very small values, at only 0.5%, 0.3%, and 0.2%, respectively.

324

Defrosting for Air Source Heat Pump

This also reflects the dominant role of the running cost during economic analysis on an ASHP unit. In addition, this figure shows that the additional first cost played minute effects on the total cost. In this section, a technoeconomic analysis study on frosting/defrosting operations for an ASHP unit, with valves installed on the multicircuit outdoor coil, was conducted. After the valves were installed, the economic performance of the new ASHP unit was effectively improved. In 15 years’ service life, the total running cost decreased as much as $798.87, or 7.67%, and the total cost about $776.36, or 6.68%. The running cost of the new ASHP unit in the heating season with frost formation decreased $542.17, or 11.47%, and in heating season without frost formation $256.72, or 7.64%. The valve effects on total cost were mainly shown at the first 5 operating years. When an ASHP unit worked after 10 years, the effects of the valves on economic performance were not as obvious as the first several years. The payback period for the additional first cost is less than 1 year. When the operating durations were 5 or 10 years, the total costs saved were as much as $243.85 and $510.19, respectively. For the new ASHP unit with valves installed, after about 2 years, the proportion of first cost in total cost decreased to about 50%. For the new ASHP unit, if the proportion of first cost needs to be decreased to 20%, the operating duration is about 7.8 years.

10.3

Refrigeration adjustment valve and watercollecting tray

When an ASHP unit works at RCD mode, the frost accumulated on the surface of outdoor coil would melt, and thus flow downward. For a vertically installed multicircuit outdoor coil, the melted frost from the upside circuit would flow into the downside circuit. The melted frost would have negative effects on the system defrosting performance. To avoid this negative effect due to downward-flowing melted frost, and thus shorten the defrosting duration for a multicircuit outdoor coil in an ASHP unit, the water-collecting trays could be placed under each circuit. Then, the melted frost could be taken away before it flows into the downside circuit. In this section, based on the valves installed in each circuit to adjust the refrigerant distribution, the installation of water-collecting trays was further considered. That means that four working conditions, with and without water-collecting trays and/or valves installed in an ASHP unit, were investigated. Their economic performance on the frosting/defrosting operations are analyzed.

10.3.1 Methodology The methodology of this economic analysis is the same as the previous section mentioned. First, based on an experimental ASHP unit with a specially made three-circuit outdoor coil, four typical conditions were designed, considering the installation style of trays and valves. After the experimental procedures were confirmed, a series of experiments was undertaken with frost accumulated at different FECs. Second,

Technoeconomic performances

325

experimental results and calculations were collected and used in later technoeconomic evaluations. Third, the equations were developed, including the first and subsequent running costs at different operating stages. After a series of conditions were assumed, a technoeconomic evaluation is given. Finally, the technoeconomic analysis on frosting/defrosting operations for optimized ASHP units is undertaken. The methodology of this economic analysis is illustrated in Fig. 10.15. The energy performance analysis, working as the basis of this study, is shown in Section 10.3.1.1. The economic performance was analyzed in Section 10.3.1.2, which is the focus of this study.

10.3.1.1 Experimental study The experimental ASHP unit and the experimental procedures and conditions are the same as the previous section mentioned. Two frosting experimental cases were designed, with both valves installed. Two different FECs, FEC1 and FEC2, were reached with valves fully open and evenly adjusted, respectively. The frosting duration was fixed at 60 min. The frosting operation performance of an ASHP unit at different FECs could be obtained in the experiments. As the baseline for comparison, the valves in Case F1 were fully open, and the FEC1 was less than 100% due to the refrigerant and air being unevenly distributed. In Case F2, the valves were evenly adjusted, giving an FEC2 value of almost 100%. Thus, the effects of the valves were obvious in the different cases. Information on the two frosting experimental cases is listed in Table 10.2. Similar to the previous section, the experimental results that form the basis for the economic analysis process in this section mainly came from previous defrosting experimental studies. Considering the different installation styles of trays and valves, four typical defrosting conditions existed, as shown in Fig. 10.16. To compare, the prototype condition is shown in Fig. 10.16A, without any change for an ASHP unit. In Fig. 10.16B, trays were installed under each circuit, and thus the melted frost could Step 1: Experimental study

Step 2: Experimental results & calculation

• • • •

•

An ASHP unit selected A three-circuit outdoor coil made Four typical conditions designed Frosting/defrosting experiments undertaken at different FECs

•

Frosting operation data collected: COP, Indoor thermal energy supplied Defrosting experimental results obtained: Defrosting duration, power consumption, and indoor air thermal energy consumed

Step 3: Assumptions & equations Step 4: Technoeconomic evaluation • • • •

Running cost analysis Total cost evaluation vs operating year The best typical case chosen Discussions on payback periods

• • • •

Fundamental assumptions Frosting/defrosting/cooling assumptions First cost: ASHP unit, valves and trays Running cost: Stage 1, Heating season with frosting formation; Stage 2, Heating season without frost formation; Stage 3, Cooling season

Fig. 10.15 Flow chart for the methodology used in this study.

326

Defrosting for Air Source Heat Pump

Circuit 1

Circuit 1

Circuit 2

Circuit 2

Tray A

Tray B Circuit 3

Circuit 3

Tray C

(A)

(B)

Without trays and valves

SV1

SV1 Circuit 1

MV1 SV2

Circuit 1 MV1 SV2

Circuit 2 MV2 SV3

Tray A Circuit 2

MV2 SV3 Circuit 3

MV3

(C)

With trays installed

Tray B Circuit 3

MV3 With valves installed

(D)

Tray C With trays and valves installed

Fig. 10.16 Four typical conditions in this study.

be collected before it flowed downward into the lower circuits. In this condition, the negative effects of downward-flowing melted frost could be eliminated. In Fig. 10.16C, six valves were installed at the entrance and exit of the three circuits. In this condition, the FEC and defrosting evenness condition for the three-circuit outdoor coil could be adjusted. However, in Fig. 10.16D, both the trays and valves were installed. Four defrosting experimental cases were designed and listed in Table 10.8. In the defrosting experimental study, the frosting duration was also designed at 60 min, and 70 min for a frosting/defrosting cycle. To keep the compressor safe, two periods of 3–4 min were left for its shutdown. The durations in a cycle are illustrated in Fig. 10.17. Clearly, after installing the valves, the FEC should be higher at the Table 10.8 Four defrosting experimental cases Item

Parameters

Case D1

Case D2

Case D3

Case D4

1 2 3 4 5 6

Valves Water-collecting trays Frosting duration Cycle duration FEC of outdoor coil Shown in Fig. 10.16

Without Without 60 min 70 min FEC3 (A)

Without With 60 min 70 min FEC3 (B)

With Without 60 min 70 min FEC4 (C)

With With 60 min 70 min FEC4 (D)

Technoeconomic performances

327

1 Frosting duration 2 Shut-down period 3 Defrosting duration 4 Shut-down period

Fig. 10.17 Durations in a frosting/defrosting cycle in this experimental study.

Fig. 10.18 Airside surface conditions at the start of defrosting in the four cases.

start of defrosting operations. Therefore, the FECs were at FEC3 and FEC4 in the four defrosting experimental cases. The FEC4 was much higher than FEC3. Here, FEC3 and FEC4 are the FEC values of the three-circuit outdoor coil at the start of defrosting without and with valves installed in each circuit, respectively. Fig. 10.18 shows the airside surface conditions at the start of defrosting for the threecircuit outdoor coil in the four cases. Because no water-collecting trays were placed under the circuits in Fig. 10.18T1 and T3, the effects of melted frost and FEC were coupled. However, in Fig. 10.3T2 and T4, with trays installed to eliminate the effects of melted frost flowing downward, only the effects of FEC on the defrosting performance were evaluated. Due to the valves being installed on each circuit, in Fig. 10.18T3 and T4, their FECs were both at FEC4, 96.6%. However, compared to the outdoor coil without valves installed, their defrosting started at lower FECs, at 82.6% and 79.4 for Cases D1 and D2, respectively. Some experimental results were listed in Table 10.9. From Case D1 to Case D4, their total frost accumulations were 878 g, 1000 g, 881 g, and 969 g, respectively. The frosting duration was 60 min, but their defrosting durations were 205 s, 197 s, 185 s, and 175 s, respectively. To undertake this economic analysis, the total power inputs to the compressor and indoor air fan and the thermal energy from the indoor air were calculated and presented.

10.3.1.2 Assumptions and calculation conditions To analyze the economics of the frosting/defrosting operation performances, fundamental assumptions, frosting assumptions, defrosting assumptions, and cooling assumptions were given. The first type was assumed for all operating seasons, and the following three types for different operating seasons.

328

Defrosting for Air Source Heat Pump

Table 10.9 Experimental results of the four defrosting cases. Item

Parameters

Case D1

Case D2

Case D3

Case D4

1 2 3 4 5 6

FEC of outdoor coil Total frost accumulation Defrosting duration Total power inputs Energy from indoor air Condition shown in Fig. 10.16 Results shown in Fig. 10.18

82.6% 878 g 205 s 124.6 kJ 666.1 kJ (A)

79.4% 1000 g 197 s 128.0 kJ 673.7 kJ (B)

96.6% 881 g 185 s 107.6 kJ 541.2 kJ (C)

96.6% 969 g 175 s 117.3 kJ 561.5 kJ (D)

(T1)

(T2)

(T3)

(T4)

7

(1) Fundamental assumptions

In this work, the following nine fundamental assumptions were first given: i. Installation, running, and maintenance costs of solenoid valves and the water-collecting trays in the three-circuit outdoor coil were neglected. ii. Service life of an ASHP unit was assumed to be 15 years, and the unit price of electricity at 0.9 CNY/kWh [16]. iii. In the whole service life of an ASHP unit, three operating seasons were divided: the heating season with frost formation, the heating season without frost formation, and the cooling season. iv. Durations of three operating seasons were constant, and assumed as the climate conditions of Beijing (Heating season: Nov. 15–Mar. 15). v. In the heating season with frost formation, the duration of a frosting/defrosting cycle was assumed at 70 min. vi. In a frosting/defrosting cycle, the frosting duration was fixed at 60 min, and two periods of 3–4 min for compressor shutdown were left for its safety, as illustrated in Fig. 10.17 (To avoid an undesired shutdown for the ASHP unit, the frosting duration was not longer than 60 min in practical application). vii. Defrosting durations in the four typical cases, system COP, and defrosting efficiency were constant. viii. All ambient air parameters, such as air temperature and relative humidity, were constant at fixed operating seasons. ix. For the cases without valves installed, their FECs were fixed at 75.7%. For the cases with valves installed, their FECs were fixed at 96.6%.

(2) Frosting assumptions

To calculate the running cost at the frosting stage, the system COP and indoor heat supplied should be obtained. Therefore, the following assumptions were also given. i. Durations of the heating season with frost formation and the heating season without frost formation were both assumed to be 60 days/year.

Technoeconomic performances

329

ii. In heating season with frost formation, cycle operation was assumed to be 15 times/day. In heating season without frost formation, the duration of the frosting operation was assumed to be 12 h/day. iii. When an ASHP unit works at the heating season with frost formation, the system frosting COP was assumed at the average value of COP in 60 min, as listed in Item 3 in Table 10.2 (When it reached 60 min, the COP decreased dramatically). iv. When the ASHP unit works at the heating season without frost formation, the system COP was assumed at the average value of COP in the first 10 min, as listed in Item 4 in Table 10.2. v. The COP showed a good linear relationship with the FEC, allowing the values of the COP at different FECs to be calculated. vi. The total indoor heat supplied showed a good linear relationship with the FEC. The values of the indoor heat supplied at different FECs could be calculated.

Based on the six assumptions and the data listed in Table 10.2, a series of experimental results in Case F3 was calculated and summarized in Table 10.10, with its FEC at 96.6%. Clearly, all the data in Case F3 were much bigger than those in Cases F1 and F2. The COP and total indoor heat supplied at the different stages listed in Table 10.10 would be used in the economic analysis. (3) Defrosting assumptions

To calculate the running cost at the defrosting stage, the following five conditions were further assumed: i. Defrosting duration showed a good linear relationship with the FEC; thereby, the defrosting durations at different FECs could be calculated. ii. Total power inputs to the compressor and indoor air fan showed a good linear relationship with the FEC, allowing the values of the total power inputs to the compressor and indoor air fan at different FECs to be calculated. iii. Energy from the indoor air showed a good linear relationship with the FEC, allowing the relative values at different FECs to be calculated. iv. FEC showed a good linear relationship with the total frost accumulation. Therefore, the FECs could be calculated when the total frost accumulated changed. v. The frost accumulation difference between Case D2 and Case D4, 31 g, was considered as evaporated and neglected.

Table 10.10 Calculation data and experimental results of the two frosting cases Item

Parameters

Case F1

Case F3

1 2 3 4 5 6 7

Valves’ status FEC of outdoor coil Average value of COP (60 min) Average value of COP (first 10 min) Indoor heat supplied (60 min) Indoor heat supplied (first 10 min) Conditions shown in Fig. 10.16

Fully open 75.7% 4.10 4.23 11,116 kJ 1922.6 kJ (A) and (B)

Evenly adjusted 96.6% 4.55 4.58 11,719 kJ 1950.6 kJ (C) and (D)

330

Defrosting for Air Source Heat Pump

Table 10.11 Calculation data and experimental results of the four defrosting cases Item

Parameters

Case D5

Case D6

Case D7

Case D8

1 2 3 4 5

FEC of outdoor coil Total frost accumulation Defrosting duration Total power inputs Total energy from indoor air Condition shown in Fig. 10.16

75.7% 1000 g 245 s 151.5 kJ 828.8 kJ

75.7% 1000 g 202 s 130.3 kJ 697.8 kJ

96.6% 969 g 203 s 118.3 kJ 595.3 kJ

96.6% 969 g 175 s 117.3 kJ 561.5 kJ

(A)

(B)

(C)

(D)

6

Based on the five assumptions and the data listed in Table 10.9, a series of experimental results in the four typical cases was calculated. As listed in Table 10.11, the results in Case D5 to Case D8 were based on Case D1 to Case D4, respectively. In Cases D5 and D6, their FECs were changed from 82.6% and 79.4% to 75.7%. Clearly, the defrosting duration, the total power inputs, and the energy from the indoor air in Cases D5 and D6 were larger than those in Cases D1 and D2. In Cases D5, D7, and D8, their total frost accumulations were changed from 878 g, 881 g, and 969 g to 1000 g. At the same time, the defrosting duration, the total power inputs to the compressor and indoor air fan, and the thermal energy from the indoor air were calculated and listed in Table 10.11. All these data will be used in the economic analysis. (4) Cooling assumptions

Due to the trays and valves in cooling operation mode not being experimentally demonstrated, their running cost differences in the four typical cases were neglected. However, in the economic analysis, all the costs during its operating life should be considered. Therefore, to calculate the running cost of the ASHP unit working in the cooling season, the following conditions were assumed: i. Duration of cooling season was assumed to be 120 days/year, 12 h/day. ii. System COP was assumed at the rated value and the capacity of the ASHP unit at the rated cooling value, as listed in Table 10.1.

Based on the 22 total given assumptions in the four types, the cost calculation equations used in the technoeconomic analysis are given in the following section. Using the same method as the previous section, the total costs in the four typical cases were developed and separately expressed as: Cr,D5 ¼

Qid, air, F + Qid,air,DF Qid,air,NFDH QC + + COPF COPNFDH COPC + Cr , comp, DF + Cr, id, fan, DF + CASHP ,

(10.39)

Technoeconomic performances

331

Cr, D6 ¼

Qid, air, F + Qid,air, DF Qid,air, NFDH QC + + COPF COPNFDH COPC + Cr , comp, DF + Cr, id, fan, DF + CASHP + Cf ,T + Ci, T ,

(10.40)

Cr, D7 ¼

Qid, air, F + Qid,air, DF Qid,air, NFDH QC + + COPF COPNFDH COPC + Cr , comp, DF + Cr, id, fan, DF + CASHP + Cf ,V + Ci, V ,

(10.41)

Cr, D8 ¼

Qid, air, F + Qid,air, DF Qid,air, NFDH QC + + COPF COPNFDH COPC + Cr , comp, DF + Cr, id, fan, DF + CASHP + Cf ,T + Ci, T + Cf , V + Ci, V

(10.42)

10.3.2 Results and Discussions All the calculation results are shown in Figs. 10.19–10.29. Among them, the running costs of the four typical cases are presented in Figs. 10.19–10.23, and the total costs shown in Figs. 10.24–10.26. The proportions of the initial cost and additional initial cost in the total cost are shown in Figs. 10.27 and 10.28. Variations of the electricity unit price are discussed and presented in Fig. 10.29. Fig. 10.19 shows the running costs of four typical cases in three typical seasons over 15 operating years. In the cooling season, the running costs are always the highest. This is because the operating duration in this season is the longest, at 21,600 h. However, the operating duration in the heating season with frost formation 33,000

Running cost (CNY)

30,000

31,104.00

31,104.00

31,104.00

31,104.00

Heating season with frosting formation Heating season without frosting formation Cooling season

27,000 24,000

2540.36 21,000

21,251.46

20,952.48

18,000

18,758.50

18,711.10 1141.37

15,000

14,936.17

14,936.17 13,794.80

12,000

Case D5

Case D6

Case D7

Four typical cases

Fig. 10.19 Running costs in three seasons over 15 operating years.

13,794.80 Case D8

332

Defrosting for Air Source Heat Pump

1600 Heating season with frost formation Heating season without frost formation

1500

Running cost (CNY)

1400 1416.76 1300

169.35

1396.83

1200

1250.57

1247.41

1100 76.09 1000 995.74

995.74

900

919.65

919.65

800 Case D5

Case D6

Case D7

Case D8

Four typical cases

Fig. 10.20 Running costs in two heating seasons over an operating year.

1600 Electricity cost on frosting Total running cost

Running cost (CNY)

1500 169.35 1400

1416.76 1396.83 127.00

1300 1284.15

1284.15 1250.57

1247.41

1157.14

1157.14

1200

1100 Case D5

Case D6

Case D7

Case D8

Four typical cases

Fig. 10.21 Running costs at the frosting stage in the four defrosting cases over 1 year.

was only 13,500–15,750 h, and that in the heating season without frost formation was the shortest, at 10,800 h. In the four typical cases, the running costs in their cooling seasons are the same at 31,104 CNY. This is because the effects the valves have on the system COP in this operating season were neglected. For the running cost in the heating season with frost formation, the differences in the four typical cases were

Technoeconomic performances

333

90 Electricity cost on defrosting Indoor air thermal energy consumed

Running cost (CNY)

80

70

75.80 29.53 63.82

60

50

49.06

56.81

12.83 46.28

48.86 44.36

40

43.99

30 Case D5

Case D6

Case D7

Case D8

Four typical cases

Fig. 10.22 Running costs at the defrosting stage in the four defrosting cases over 1 year.

2.4 Indoor air thermal energy consumed Electricity cost on defrosting Electricity cost on frosting

Running cost (CNY)

2.0

1.574 1.6

1.552 1.389

1.386

1.427 91.93%

1.286 92.53%

1.286 92.76%

Case D6

Case D7

Case D8

1.2

0.8 1.427 90.64% 0.4

0.0 Case D5

Four typical cases

Fig. 10.23 Running costs in a frosting/defrosting cycle in heating season with frost formation.

obvious due to the effects of the trays and valves. From highest to lowest, their values were 21,251.46 CNY in Case D5, 20,952.48 CNY in Case D6, 18,758.5 CNY in Case D7, and 18,711.1 CNY in Case D8. The largest difference between Cases D5 and D8 demonstrated that the running cost of the ASHP unit in the heating season with frost formation could decrease by as much as 2540.36 CNY or 11.95%, after the trays and

334

Defrosting for Air Source Heat Pump

100,000 First cost

Running cost

90,000 80,000

75,291.63

75,007.65

67,291.63 89.37%

Case D5

71,807.29

71,774.88

66,992.65 89.31%

63,657.29 88.65%

63,609.88 88.62%

Case D6

Case D7

Case D8

Total cost (CNY)

70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 Four typical cases

Fig. 10.24 Total costs in the four typical cases over 15 operating years. 80,000 Case D5 Case D6 Case D7 Case D8

70,000

50,000 40,000

76,000 74,000

Case D5 Case D6 Case D7 Case D8

72,000

30,000

Total cost (CNY)

Total cost (CNY)

60,000

20,000

70,000 68,000 66,000 64,000

10,000

D5 > D6 > D7 ≈ D8

62,000 13

14

15

Operation time (year)

0 0

2

4

6

8

10

12

14

16

Operation time (year)

Fig. 10.25 Total costs in the four typical cases in the 15 operating years.

valves were installed. In addition, for the running cost in the heating season without frost formation, from Case D5 to Case D8, the values were 14,936.17 CNY, 14,936.17 CNY, 13,794.8 CNY, and 13,794.8 CNY, respectively. The effects of the watercollecting trays on defrosting performance did not exist in this operating season. But the effect of the valves was shown, with the difference of 1141.37 CNY between

Technoeconomic performances

335

4000 D6–D7

3200.36

3000

Total cost difference (CNY)

D5–D8

3516.75

3500

2500

2289.50

2000

2088.58

1500 1062.25 1000 976.78 500 0

0.65

–500 –2

0

2

4

6

8

10

12

14

16

Operating time (year)

Fig. 10.26 Total cost differences between the four cases in the 15 operating years.

100 Case D5 Case D6 Case D7 Case D8

Proportion of first cost in total cost (%)

90 80 70 60 1.9

50

2.9

40

D8 ≈ D7 > D6 ≈ D5 4.5

30

7.6

20

15.0

10 0 0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16

Operating time (year)

Fig. 10.27 Proportion of the first cost in the total cost in the 15 operating years.

Cases D5 and D8. It was demonstrated that the running cost of the new ASHP unit in the heating season without frost formation could decrease by as much as 7.64%. Therefore, after optimizing the ASHP unit with trays and/or valves, its economic performance would be better.

336

Defrosting for Air Source Heat Pump

Proportion of additional first cost in total cost (%)

2.4 Case D6 Case D7 Case D8

2.0 1.6 1.2 0.8

0.562 0.511

0.326 0.297

0.4 0.0

0.049

0.028

0.230 0.209

0.020

–0.4 0

1

2

3

4

5

6

7

8

9

10 11 12 13 14 15 16

Operating time (year) Fig. 10.28 Proportion of the additional first cost in the total cost in the 15 operating years.

Proportion of first cost in total cost (%)

15 14

Case D6 Case D7 Case D4

0.83

13 12

0.76

11

0.69 0.64

10

0.59

9 8 0.7

0.8

0.9

1.0

1.1

Electricity unit price (CNY/kWh)

Fig. 10.29 Proportions of initial cost of the total cost with electricity unit price variation in the 15 operating years.

Technoeconomic performances

337

To clearly compare the running costs in the heating season in four typical cases, Fig. 10.20 shows their running costs in a year. It is obvious that the running cost in the heating season without frost formation was much more than that in the heating season with frost formation. From highest to lowest, the running costs in the heating season with frost formation were 1416.76 CNY in Case D5, 1396.83 CNY in Case D6, 1250.57 CNY in Case D7, and 1247.41 CNY in Case D8. Their biggest difference was 169.35 CNY between Case D5 and Case D8. For the running costs in the four typical seasons in the heating season without frost formation, their values were at 995.74 CNY in Cases D5 and D6, and 919.65 CNY in Case D7 and D8. Their biggest difference was also between Case D5 and Case D8, at 76.09 CNY, which was smaller than the difference in the heating season with frost formation. This is because the watercollecting trays installed had no effect as there was no defrosting operation during the heating season without frost formation. In addition, the total difference showed that after the trays and valves were installed, the total running cost during heating seasons could decrease by 245.44 CNY, or 10.17%. Furthermore, in order to clearly analyze the running cost consumed in the heating season with frost formation, Figs. 10.21 and 10.22 show the running costs at frosting and defrosting stages, respectively, in the four typical cases over a year. As shown in Fig. 10.21, the two curves’ trends are nearly the same. This shows that the total running cost was mainly decided by the electricity cost on the frosting operation stage. It is reasonable because the operation duration frosting is 60 min during a 70 min frosting/defrosting cycle. However, the total running cost shows at D5 > D6 > D7 > D8, but the electricity cost on frosting at D5 ¼ D6 > D7 ¼ D8. This is because the trays and valves had no effect at this stage. As shown in Fig. 10.22, these effects were shown on the running costs of electricity on defrosting, and the indoor air thermal energy consumed during defrosting. Therefore, from highest to lowest, the running cost was 1416.76 CNY in Case D5, 1396.83 CNY in Case D6, 1250.57 CNY in Case D7, and 1247.41 CNY in Case D8. Their biggest difference was 169.35 CNY between Case D5 and Case D8. Here, the fact that the running cost could save a lot after trays and valves were installed was further confirmed. Also, from the big difference between Case D6 and Case D7, as shown in Fig. 10.21, we can find that the economic effects of the valves are bigger than that of the trays. Due to the installation of trays and valves, in Fig. 10.22, the trends of the two curves show that, from Case D5 to Case D8, their running costs became steadily smaller. Their biggest difference was also shown between Case D5 and Case D8, at 29.53 CNY in indoor air thermal energy consumed, and 12.83 CNY in electricity cost during defrosting. It is interesting that the value of the indoor air thermal energy consumed during the defrosting operation is always higher than the electricity cost of defrosting. This is because the cold refrigerant takes more thermal energy from the indoor air during defrosting. Clearly, the defrosting duration in Case D8 was the shortest. Therefore, the running cost in this case was always the lowest. In order to analyze the proportions of the three running costs, the indoor air thermal energy consumed, and the electricity cost on the defrosting and frosting stages, the data showing the running cost of a frosting/defrosting cycle in the heating season with frost formation is available in Fig. 10.23. In the four typical cases, their total running

338

Defrosting for Air Source Heat Pump

costs are decreasing steadily from Case D5 to Case D8. Their values were 1.574 CNY in Case D5, 1.552 CNY in Case D6, 1.389 CNY in Case D7, and 1.386 CNY in Case D8. However, for the electricity cost during frosting in the four cases, although their values show at D5 ¼ D6 > D7 ¼ D8, their proportions at D5 < D6 < D7 < D8. It is easy to conclude that the running cost saved at the defrosting stage was much more than that saved at the frosting stage. This also reflects that the economic performance of an ASHP unit was highly improved after the trays and valves were installed. Fig. 10.24 shows the total costs in the four typical cases over 15 operating years. From big to small, their values were at 64,880.46 CNY in Case D5, 64,716.08 CNY in Case D6, 63,910.52 CNY in Case D7, and 63,895.80 CNY in Case D8. This directly reflects that the economic performances of an ASHP could be effectively improved after trays or/and valves are installed. With nearly the same initial costs, the total costs in the four cases were mainly decided by their running costs. The running cost in Case D5 was 56,880.46 CNY, but 55,730.80 CNY in Case D8. That means that as much as 3681.75 CNY, or 5.47%, of the running cost, could be saved after the trays and valves are installed, from conditions (a) to (d) shown in Fig. 10.16. With their initial costs considered, the total cost decreased by as much as 3516.75 CNY, or 4.67%. Fig. 10.25 shows total costs in four typical conditions in the 15 operating years. As the fundamental assumptions are given, the total costs kept increasing in line with the operating time. In practical application, the total running cost would increase much quicker than expected because the energy performances of the ASHP unit would decrease with time. Also, the maintenance costs increases the total running cost. As shown in Fig. 10.25, the running cost relationship of the four typical cases shows at D5  D6 > D7  D8. This also reflects that the effects of the valves were more obvious than that of the trays on the running cost. After 2 years’ operation, their differences were obvious. As shown in the inset in Fig. 10.25, their relationship became D5 > D6 > D7  D8 after 12 years’ operation. The running period enlarged their differences. Therefore, the conclusions of this study have important roles in the budgeting work of equipment procurement. To clearly show the biggest difference of the total costs in the four typical cases, the difference between Cases D5 and D8 is further shown in Fig. 10.26. Also, to further confirm the effects of the valves to be larger than those of the trays, the total cost difference between Case D6 and Case D7 in the 15 operating years is shown in this figure. When their values were 0, the operating periods were both at 0.65 year. That means their additional initial costs in Cases D6, D7, and D8 could be covered by their running costs in 0.65 year, compared with the total cost in Case D5. When the operating durations were 5, 10, or 15 years, the total costs saved could be as much as 1062.25 CNY, 2289.50 CNY, and 3516.75 CNY, respectively. Meanwhile, the total cost differences between installing trays and valves were 976.78 CNY, 2088.58 CNY, and 3200.36 CNY, respectively. Fig. 10.27 shows the proportions of the initial cost in the total cost in four typical cases in the 15 operating years. At the end of the lifespan, their relationship showed at D8  D7 > D6  D5. This also confirmed that the installation of trays and valves could save more money in an ASHP unit’s application. As shown, when the recovery proportions of the initial costs were at 50%, 40%, 30%, and 20%, the operating durations

Technoeconomic performances

339

were at 1.9, 2.9, 4.5, and 7.6 years, respectively. That means that before the ASHP unit works 8 years, more than 80% of the initial cost could be recovered. After 15 years, the initial cost becomes only 10% of the total cost. It is also confirmed that the total cost is mainly decided by the running cost. After the trays and valves are installed, the total cost could significantly decrease. To clearly show the additional initial cost effect on the total cost during operation of an ASHP unit, Fig. 10.28 shows proportions of the additional initial cost of the total cost in the 15 operating years in three typical cases. As shown, the curve of Case D6 was the lowest because its additional cost was the smallest, only 15 CNY for the three water-collecting trays. Therefore, after the ASHP unit worked 5 years, 10 years, and 15 years, the proportions of the additional initial cost of the total cost remained very small, at only 0.409%, 0.028%, and 0.020%, respectively. However, the line of Case D8 was always the highest because its additional initial cost was the most, 165 CNY for trays and valves. Although the proportions in Cases D7 and D8 were much bigger, when the ASHP unit worked for 5 years, the additional initial costs became less than 0.6% of the total cost. This also reflects the dominant role of the running cost on the economic analysis of an ASHP unit. In addition, this figure shows that the additional initial cost had a minute effect on the total cost. When the ASHP unit is changed to another one with a higher-rated power, the running cost difference between the traditional and modified ones would be larger, implying that more money could be saved by installing valves and trays in the multicircuit outdoor coil due to operational performance improvements and less energy consumption after modification. Thus, the modification of the multicircuit outdoor coil should be fully considered when designing or optimizing a bigger scale residential ASHP unit or some commercial ones. For example, the rated heating capacity of the ASHP unit is much higher than 6.5 kW, as used in this study. In this study, the electricity unit price was assumed to be 0.9 CNY/kWh. Proportions of the initial cost of the total cost in the 15 operating years with an electricity unit price variation from 0.7–1.1 CNY/kWh were discussed and presented in Fig. 10.29. As seen, the trend is decreasing as the electricity unit price increases. It is easy to understand that the operation cost is affected by the electricity price. At the same time, the difference of D6 and D4 is decreasing, from 0.83% with the unit price at 0.7 CNY/ kWh to 0.59% with the unit price at 1.1 CNY/kWh. That means that as the electricity unit price increases, the additional initial cost effect on the total cost becomes smaller. If the electricity unit price decreases, we should also consider this modification for two reasons: (1) more energy could be saved with a higher-rated power ASHP unit. These running costs are much higher than that of the investment of modification; and (2) the environmental factor, as more energy saved leads to reducing environmental pollution. It is well known that environmental problems have a big negative influence on our life and society. Therefore, modifications should be fully considered no matter whether the unit price of electricity decreases or increases. This also demonstrated the fundamental meaning of this study. In this section, a technoeconomic analysis study on frosting/defrosting operations for an optimized ASHP unit, with trays and/or valves installed with its outdoor coil, was conducted and the results are reported. After the water-collecting trays and/or valves were installed, the economic performances of an ASHP unit were effectively

340

Defrosting for Air Source Heat Pump

improved. The total running cost decreased by as much as 3681.75 CNY, or 10.33%, and the total cost 3516.75 CNY, or 4.67%, over 15 years’ service life. After installing both the water-collecting trays and valves, the running cost of the new ASHP unit in the heating season with frost formation decreased by 2540.36 CNY, or 11.95%, and in the heating season without frost formation the decrease was 1141.37 CNY, or 7.64%. The effect of valves on the economic performance of the ASHP unit is obviously better than that of trays. After water-collecting trays were installed, the total cost decreased by 283.98 CNY. But the saved cost after the valves were installed enlarged 12.27 times, at 3484.34 CNY.

10.4

Concluding remarks

Two technoeconomic analyses on frosting/defrosting operations for an optimized ASHP unit are presented in this chapter. The first one considered two conditions, with trays installed in the multicircuit outdoor coil. For the second one, four conditions, with trays and/or valves installed in the outdoor coil, were considered. As analyzed, after the water-collecting trays and/or valves were installed, the economic performances of an ASHP unit were effectively improved. Over 15 years’ service life, the total running cost was decreased by as much as 10.33%, and 4.67%. After installing both the water-collecting trays and valves, the running cost of the new ASHP unit in the heating season with frost formation was decreased by 11.95%, and in the heating season without frost formation by 7.64%. The effect of installing valves on the economic performance of the ASHP unit is obviously better than that of installing trays. After water-collecting trays were installed, the total cost was decreased while the saved cost after the valves were installed was enlarged. The payback period for the additional initial cost was calculated at only 0.65 year. Electricity unit price variations were further discussed.

References [1] Liu SC, Li XQ, Song MJ, Li HL, Sun ZL. Experimental investigation on drying performance of an existed enclosed fixed frequency air source heat pump drying system. Appl Therm Eng 2018;130:735–44. [2] Song MJ, Pan DM, Li N, Deng SM. An experimental study on the negative effects of downwards flow of the melted frost over a multi-circuit outdoor coil in an air source heat pump during reverse cycle defrosting. Appl Energy 2015;138:598–604. [3] Song MJ, Deng SM, Pan DM. An experimental study on the effects of downwards flowing of melted frost over a vertical multi-circuit outdoor coil in an air source heat pump on defrosting performance during reverse cycle defrosting. Appl Therm Eng 2014;67(12):258–65. [4] Song MJ, Xia L, Mao N, Deng SM. An experimental study on even frosting performance of an air source heat pump with a multi-circuit outdoor coil. Appl Energy 2016;164:36–44. [5] Wu W, Wang BL, Shi WX, Li XT. Techno-economic analysis of air source absorption heat pump: improving economy from a design perspective. Energy Build 2014;81:200–10.

Technoeconomic performances

341

[6] Esen H, Inalli M, Esen M. Technoeconomic appraisal of a ground source heat pump system for a heating season in eastern Turkey. Energ Convers Manage 2006;47:1281–97. [7] Esen M, Yuksel T. Experimental evaluation of using various renewable energy sources for heating a greenhouse. Energy Build 2013;65:340–51. [8] Wang ZH, Wang FH, Ma ZJ, Wang XK. Dynamic character investigation and optimization of a novel air-source heat pump integrated with PCM and dehumidification. Appl Therm Eng 2017;111:122–33. [9] Song MJ, Liu SC, Deng SM, Sun ZL, Yan HX. Experimental investigation on reverse cycle defrosting performance improvement for an ASHP unit by evenly adjusting its refrigerant distribution. Appl Therm Eng 2017;114:611–20. [10] Bourke G, Bansal P. Energy consumption modeling of air source electric heat pump water heaters. Appl Therm Eng 2010;30:1769–74. [11] Wang ZH, Wang FH, Ma ZJ. Numerical study on the operating performances of a novel frost-free air-source heat pump unit using three different types of refrigerant. Appl Therm Eng 2017;112:248–58. [12] Bansal P. A review—status of CO2 as a low temperature refrigerant: fundamentals and R&D opportunities. Appl Therm Eng 2012;41:18–29. [13] Horton WT, Groll EA, Braun JE. Development of a high performance cold climate heat pump. DOE-Purdue-0003842. 2014. https://doi.org/10.2172/1133089. [14] Esen H, Inalli M, Esen M. A techno-economic comparison of ground-coupled and aircoupled heat pump system for space cooling. Build Environ 2007;42:1955–65. [15] Esen H, Inalli M, Esen M. Numerical and experimental analysis of a horizontal groundcoupled heat pump system. Build Environ 2007;42:1126–34. [16] Zhang QL, Zhang L, Nie JZ, Li YL. Techno-economic analysis of air source heat pump applied for space heating in northern China. Appl Energy 2017;207:533–42. [17] Satu P, Sakari P, Antti K. Life-cycle cost analyses of heat pump concepts for Finnish new nearly zero energy residential buildings. Energy Build 2017;150:396–402. [18] Willem H, Lin Y, Lekov A. Review of energy efficiency and system performance of residential heat pump water heaters. Energy Build 2017;143:191–201. [19] Hong Kong Observatory. http://www.hko.gov.hk/wxinfo/climat/world/chi/asia/china/ wuhan_c.htm, n.d.

Conclusions and future work 11.1

11

Conclusions of the present work

The present research work is devoted to the research, analysis, and methods of frosting and defrosting for ASHP units, with an emphasis on an ASHP unit having a multicircuit outdoor coil. The following six achievements are obtained. The first is the qualitative and quantitative evaluation of the effect of downwardflowing melted frost due to gravity along the surface of a multicircuit outdoor coil in an ASHP unit on its defrosting performance, based on experimental studies using a specially designed ASHP unit with a vertically installed multicircuit outdoor coil. Water collecting trays were designed and installed under each circuit when necessary, and thus the said effect could be investigated with comparative study cases. Based on the experimental results, a set of uneven defrosting models, with the melted frost locally drained considered or not, were developed and validated. The negative effect due to the downward-flowing melted frost on defrosting performance further demonstrated, as well as many difficult and impossible measured parameters predicted with the validated models, for example, the temperature of downward-flowing melted frost and the variation of thermal resistance of water film. These defrosting models are based on experimental conditions and thus have some limitations, but also high accuracy. The defrosting models are based on the outdoor coil, but not the system, which is also their advantage. It is also the first set of defrosting models that considered the thermal energy stored in the metal of the outdoor coil during RCD. Because the models have the aforementioned advantages, they were further used to study the performances of three given defrosting control strategies. And finally, one was suggested to alleviate the uneven defrosting for an ASHP unit having a multicircuit outdoor coil. Both the defrosting efficiency and defrosting duration were used in the evaluation. The conclusions are reported in Chapters 3 and 4. The second is the realization of even defrosting for an ASHP unit with a multicircuit outdoor coil. In this study, to alleviate uneven defrosting by eliminating the effect of downward-flowing melted frost, the vertically installed outdoor coil was horizontally placed. It was determined that the uneven defrosting could be alleviated by changing the installation style of the multicircuit outdoor coil. As expected, the defrosting performance could be improved when even defrosting was realized by defrosting efficiency increased and defrosting duration shorted. Meanwhile, retained water was found on the downside of the horizontally installed multicircuit outdoor coil, and the effect of the retained water on the downward surface of the coil due to surface tension was further experimentally evaluated. With this method, the effect of retained water due to surface tension on the downside surface of each circuit during defrosting for a vertically installed multicircuit outdoor coil could be further calculated. Additionally, the defrosting evenness status was also evaluated with the defrosting terminations of each circuit, with the named of DEC. This section is reported in Chapter 5. Defrosting for Air Source Heat Pump. https://doi.org/10.1016/B978-0-08-102517-8.00011-4 © 2019 Elsevier Ltd. All rights reserved.

344

Defrosting for Air Source Heat Pump

The third is the qualitative and quantitative confirmation of two influence parameters on uneven defrosting performance in a multicircuit outdoor coil. These parameters are the distributions of frost accumulation on the surface of each circuit and the refrigerant inside each circuit. First, the FEC was used to describe the frost distribution status on the surface of the outdoor coil, and both the frosting and defrosting performances of the experimental ASHP unit were tested when the FEC was at different values, at the range of 70%–100%. Here, the adjustment of FEC value was successfully achieved by using the valves installed in each circuit of the outdoor coil. In the defrosting process, the water-collecting trays were also used to eliminate the effects of downward-flowing melted frost. As reported in Chapter 6, the frosting and defrosting performances, with or without the melted frost locally drained, were all improved when the FEC was at a higher value. Second, the refrigerant’s even distribution status was adjusted with the valves installed on each circuit by operating the ASHP unit at defrosting mode while no frost was accumulated on the surface of its outdoor coil. Meanwhile, when the valves were all fully open, the uneven refrigerant distribution status was used as a reference case. With the comparative experimental study, the conclusion that even refrigerant distribution improves the defrosting performance was finally determined and quantitatively analyzed in Chapter 7. The two parameters’ distribution studies are valuable for the control strategy optimization of ASHP units having a multicircuit outdoor coil. The fourth is the reveal of the energy transfer mechanism in an ASHP unit during its RCD. When an ASHP unit works at RCD at a lower defrosting efficiency or/and a longer defrosting duration, the insufficient energy supply and wasted energy consumption are the main reasons. Before this work was carried out, nearly no study clearly and systematically reported the energy transfer mechanism in an ASHP unit during its RCD. Here, the energy supply was divided into four sections, and the heat consumption into five sections. The energy transferred in the metal of the two coils, the indoor and outdoor coils, due to their temperature variation when their roles changed in the condenser and evaporator, were also considered. Based on experimental studies, all nine sections were quantitatively analyzed. The effect of the thermal energy stored in the metal of the two coils on defrosting performance was also evaluated. In the aforementioned experimental studies, both the melted frost locally drained or not were considered. This section is described in Chapter 8. The fifth is the control strategy optimization for the initiation and termination of RCD in an ASHP unit with a multicircuit outdoor coil. The time-based defrosting initiation control strategy for an ASHP unit with different frost accumulations and even distribution on the surface of the outdoor coil was experimentally investigated, with melted frost locally drained or not. As found, the frost accumulation increased as the frosting duration was prolonged, but without a positive proportion relation. The defrosting duration was not at a positive proportion relationship with frost accumulation either. In view of system stability and indoor thermal comfort, the system performance would be degraded when the frost accumulation reached some fixed value, no matter whether the melted frost was locally drained. In addition, a methodology for confirming a suitable DTT for RCD in an ASHP unit having a multicircuit outdoor coil was proposed and experimentally examined. For the optimization of defrosting initiation and termination control strategies, the defrosting efficiency always worked

Conclusions and future work

345

as the most fundamental reference parameter. This fundamental study is meaningful for reaching an intelligent ASHP unit, which is reported in Chapter 9. Based on the aforementioned five achievements, the last one was reached, which is two technoeconomic analyses on frosting/defrosting operations for an optimized ASHP unit. The first one considered two conditions with trays installed in its multicircuit outdoor coil. For the second one, four conditions, with trays and/or valves installed in the outdoor coil, were considered. As analyzed, after the water-collecting trays and/or valves were installed, the economic performance of an ASHP unit was effectively improved. The initial and running costs over 1–15 years were calculated. The effect of installing valves on the economic performance of the ASHP unit was obviously much better than that of installing trays. After water-collecting trays were installed, the total cost was decreased while the saved cost after the valves were installed was increased. The payback period for the additional initial cost was also calculated. This section is reported in Chapter 10.

11.2

Proposal for future work

Based on the results presented in this book, some potential areas for future research related to defrosting for ASHP units are recommended as follows. (1) System and component optimization [1]. When optimizing the original components in an ASHP unit, more attention should be paid to the entire system’s operating performance. Experimental or/and numerical studies on the frosting/defrosting performance for an ASHP unit at different frost accumulations, circuit numbers, total heat exchanger areas, etc., have still not been reported. When an ASHP unit with a horizontally installed multicircuit outdoor coil is applied to some special places, such as on the roof of a building or vehicle, its system performance could be further studied. (2) New methods and materials. New frost suppression measures [2] and defrosting methods [3], taking their economic, energy and environmental performances into consideration, should be studied. New hydrophobic materials with good durability, high thermal conductivity, and a self-cleaning function are meaningful for industry application. (3) Model development. New defrosting models of an ASHP unit with PCM-TES, with indoor thermal comfort at both daytime and nighttime or sleeping thermal comfort considered, should be developed [4]. New frosting and defrosting models for a flat plate or a fin at micro/nanoscales should be developed, for example, using the lattice Boltzmann method. (4) Control strategy optimization. A defrosting control strategy of refrigerant distribution adjustment should be studied so that the melted frost can have a positive effect on the defrosting performance in an ASHP unit with a vertically installed multicircuit outdoor coil. More indoor and outdoor environmental parameters could be considered to avoid maldefrosting [5]. An operation control strategy coupled with big data on climates can be considered. (5) Mechanism study. For some mechanical-based frost-suppression methods, such as ultrasonic vibration and air jet techniques, their mechanisms of frosting suppression on the fin surface should be studied. In order to improve the FECs for a multicircuit outdoor coil, and thus improve the system frosting/defrosting performance, the mechanism of multiphase refrigerant distribution into each circuit during uneven heat transfer between the refrigerant and outside frost, melted frost, and ambient air, should also be studied [1].

346

Defrosting for Air Source Heat Pump

References [1] Song M, Deng S, Dang C, Mao N, Wang Z. Review on improvement for air source heat pump units during frosting and defrosting. Appl Energy 2018;211:1150–70. [2] Kreder MJ, Alvarenga J, Kim P, Aizenberg J. Design of anti-icing surfaces: smooth, textured or slippery? Nat Rev Mater 2016;1:15003. [3] Amer M, Wang CC. Review of defrosting methods. Renew Sustain Energy Rev 2017;73:53–74. [4] Wenju H, Yiqiang J, Minglu Q, Long N, Yang Y, Shiming D. An experimental study on the operating performance of a novel reverse-cycle hot gas defrosting method for air source heat pumps. Appl Therm Eng 2011;31:363–9. [5] Song MJ, Dong JK, Wu CL, Jiang YQ. Improving the frosting and defrosting performance for air source heat pump units: review and outlook. HKIE Trans 2017;24:88–98.

Appendices

Appendix A: Defrosting efficiency Defrosting efficiency can be used to evaluate the performance of a defrosting operation. It is defined as the ratio of the actual amount of energy consumption required to both melt the accumulated frost and vaporize the retained melted frost to the total amount of energy available from an outdoor coil during an entire defrosting operation, as follows: ηd ¼

Qm + Qv  100% Qcom + Qi, fan + Qi, air

(A.1)

where the energy consumed on melting frost and vaporizing retained water, Qm and Qv, were evaluated by: Qm ¼ mf  Lsf

(A.2)

Qv ¼ mv  Lv

(A.3)

where mf and mv are the total mass of the frost accumulated over the outdoor coil surface and the total mass of the vaporized water, and Lsf and Lv are the latent heat of frost melting and water vaporization, respectively. Qcom, Qi, fan, and Qi, air are the energy consumptions by the compressor and supply fan, and the thermal energy from the indoor air during defrosting, respectively.

348

Appendices

Appendix B: Calculation error of defrosting efficiency The calculation error of defrosting efficiency can be obtained by δy ¼ 

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi a21 δ21 + a22 δ22 + a23 δ23

(B.1)

in which δ1 and δ2 are the measurement errors of frost mass, mf and mv, at 104. δ3 is the measurement error of energy consumption, at 102. The other three parameters, a1 , a2, and a3, could be calculated by, a1 ¼

∂ηd Lsf ¼ ∂m1 Qe

(B.2)

a2 ¼

∂ηd Lv ¼ ∂m2 Qe

(B.3)

a3 ¼

Lsf m1 + Lv m2 ∂ηd ¼ ∂Qe Q2e

(B.4)

Finally, the calculation error of defrosting efficiency was calculated at 0.15%.

Appendices

349

Appendix C: Metal energy storage effect The metal energy storage (MES) effect was used to evaluate the influence of the thermal energy stored in the metal of the indoor and outdoor coils on the defrosting performance. It was defined as the ratio of the net amount of energy from the indoor and outdoor coils to the total amount of energy available from the system outside during an entire defrosting operation, as follows: ηm ¼

Qi, MES  Qo, MES  100% Ecomp + Ei, fan + Qi, air

(C.1)

in which, Ecomp and Ei, fan are the power inputs to the compressor and indoor air fan during defrosting, respectively. Qi, fan is the thermal energy supply from the indoor air. Qi, MES and Qo, MES are the MES values of the indoor and outdoor coils, respectively. The total mass of frost accumulation during the experiment, mf, could be calculated with the density of the air in the outdoor frosting space and the volumetric flow rate of the air passing through the outdoor coil, as shown in the following equations. ðtf mf ¼ Δmf  dt ¼

X Δmf  Δt

(C.2)

0

Δmf ¼ ρo, a  Vo, a  Δt  ðωo, o  ωo, i Þ

(C.3)

In Eqs. (C.2) and (C.3), ρo, a and Vo, a are the density and volume of outdoor air, and ωo, o and ωo, i are the moisture content of the air at the outlet and inlet of the outdoor coil. Δt and tf are the measuring time interval and frosting duration, respectively. The total mass of vaporized water, mv, was expressed by mv ¼ mf  mm ¼ mf  mtf  mrw

(C.4)

in which, mcf is the total mass of the melted frost collected in the cylinders, and mrw the total mass of the retained water. In Eqs. (A.1) and (C.1), Qi, a is the thermal energy transferred from the indoor air, which was evaluated by: ðtd Qi, a ¼ ci, a mi,a  dT ¼

X

ci, a ρi, a Vi, a Δt  ðTind, in  Tind, out Þ

(C.5)

0

where, ci, a is the specific heat of the indoor air and mi, a the mass rate of the indoor air. ρi, a is the density of the air in the indoor heated space, and Vi, a the volumetric flow rate of the air passing through the indoor coil. Tind, in and Tind, out are the measured air temperature at the inlet and outlet of the indoor coil, respectively.

350

Appendices

Appendix D: Frosting evenness coefficient

Refrigerant entrance (vapor)

Fig. D.1 Frost accumulations on the three circuits’ surface.

FA1

Circuit 1

FA2

Circuit 2

FA3

Circuit 3

Refrigerant exit (liquid)

A frosting evenness coefficient (FEC) was used to clearly describe the uneven defrosting that results when the frost accumulations on the surface of each circuit are different. It was defined as the ratio of the minimum frost accumulation among all circuits to the maximum one. In this study, the frost accumulations could be calculated by the melted frost collected from each water-collecting cylinder, with the water vaporized into the ambient air neglected. For example, as shown in Fig. D.1, the masses of frost accumulation on the three circuits’ surface are FA1, FA2, and FA3, respectively. When FA1 ¼ FA2 ¼ FA3, the FEC is 100%, and thus this is even frosting. If no FA1 ¼ FA2 ¼ FA3, the FEC is not 100%, and that is uneven frosting. When FA1 > FA2 > FA3, FEC is the ratio of FA3 to FA1, which is less than 100%.

Water collecting tray

Appendices

351

Appendix E: Defrosting evenness coefficient A defrosting evenness coefficient (DEC) was used to evaluate the evenness status of defrosting in an ASHP unit with a multicircuit outdoor coil, and thus to evaluate the energy consumption on heating ambient air due to waiting for the other circuit’s defrosting termination. Defrosting of a circuit was terminated when the tube surface temperature at its exit reached a preset defrosting termination temperature, and the duration is the defrosting duration for the circuit. The DEC was defined as the ratio of the minimum defrosting duration of all circuits to the maximum one. For example, as shown in Fig. E.1, the defrosting durations of three circuits are DT1, DT2, and DT3, respectively. When DT1 ¼ DT2 ¼ DT3, the DEC is 100%, and thus this is even defrosting. If no DT1 ¼ DT2 ¼ DT3, the DEC is not 100%, and that is uneven defrosting. When DT1 > DT2 > DT3, which is the common status for a vertically installed three-circuit outdoor coil, DEC is the ratio of DT3 to DT1, which is obviously less than 100%. Clearly, the higher the DEC, the more energy used for heating the ambient air could be saved due to waiting for the other circuit to terminate its defrosting.

Refrigerant entrance (vapor)

DT1

Circuit 1 T2

DT2

Circuit 2

Refrigerant exit (liquid)

T1

T3 DT3

Circuit 3

Water collecting tray

Fig. E.1 Defrosting durations for three circuits during reverse cycle defrosting.

352

Appendices

Appendix F: Program listing of Model 1 in Section 4.2 The following program written in Matlab R2012a was used for a modeling study of the negative effects of the downward flowing of melted frost on defrosting performance for an ASHP unit having a vertically installed three-circuit outdoor coil during reverse cycle defrosting without using any water-collecting trays between circuits; this is named Model 1 in Chapter 6 in this thesis. clear all; clc; hri=zeros(45,3); % the enthalpy value of input refrigerant, kJ/kg Mr=zeros(45,3); % the mass flow rate of refrigerant, kg/s Tri=zeros(45,3); % the temperature of input refrigerant, °C Rr=zeros(45,3); % thermal resistance of refrigerant during defrosting, (K˙m2)/W % input all the known parameters hri=xlsread(’song20130828’,’hri’); Mr=xlsread(’song20130828’,’Mr’); Tri=xlsread(’song20130828’,’Tri’); Rr=xlsread(’song20130828’,’Rr’); % input all the known parameters from the excel with the experimental results mf=zeros(45,3); % the mass of melted frost, when it comes to 4th stage, it is 0 kg/5s Tw=zeros(45,3); % the temperature of retained water on the coil, °C qr=zeros(45,3); % the energy used from the refrigerant during the 5 seconds, J qr2=zeros(46,3); % use refrigerant (R22) to calculate the energy used in defrosting; and another way is to measure and calculated the energy used with Power system, J Tro=zeros(45,3); % the temperature of exit refrigerant, °C qm=zeros(45,3); % the energy used in frost melting during the 5 seconds, J; melting and energy comes from the refrigerant, J sfrost=zeros(45,3); % the total mass of frost melted before the moment, kg qair=zeros(45,3); % the energy used in the ambient air, J hair=zeros(45,3); % the coefficient of natural convective heat transfer, W/(m2 °C) mvaw=zeros(45,3); % the mass of vaporized water into the ambient air, kg/5s smvaw=zeros(45,3); % the sum of vaporized water before this moment, kg mrw=zeros(45,3); % the mass of retained water on this coil, kg/5s hd=zeros(45,3); % the coefficient of convective mass transfer, W/(m2 °C) qvap=zeros(45,3); % the energy used in the water vaporized, J s_qvap=zeros(45,3); % total energy used in the water vaporized, J watertray=zeros(45,3); % the mass of water flowing away from the water collecting tray, kg/5s

Appendices

353

swatertray=zeros(45,3); % the total mass of water flowing away from the water collecting tray in the 5 seconds, kg s_qr2=zeros(45,3); % the energy taken in the refrigerant, J hro=zeros(45,3); % the enthalpy value of output refrigerant, kJ/kg % list the unknown parameters, and initialize these parameters with zeroes for i=1:3 % three circuits in this study based on the experiment results for j=1:18 % about 18*5 seconds in the first two stages, obtained from the experimental results if j==1 khri=hri(j,i); % , kJ/kg kMr=Mr(j,i); % kg / s kTri=Tri(j,1); % °C kRr=Rr(j,i); %(K˙m2)/W ksmrw=0.0001; % the total retained water at the beginning is 0 kg, choose 0.0001 as the value for debugging, kg kTw1=0.01; % the temperature of the melted frost at the beginning is 0 °C, choose 0 °C as the value for debugging % all the input parameters in the function listed here x0=[0.0001 0.0001 0.01 1200 0.001]; % mf=x(1), mrw=x(2), Tw=x(3); qr=x(4); Tro=x(5) the values of debugging options=optimset(’display’,’off’,’MaxIter’,100000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x)mystage1(x,ksmrw,kTw1,i,kRr,kTri, khri,kMr),x0,options); % x, ksmrw, kTw1, i, kTri, kRr, kqr mf(j,i)=A(1); % the mass of melted frost, kg/s mrw(j,i)=A(2); % the mass of retained water, kg/s Tw(j,i)=A(3); % the temperature of retained water, °C qr(j,i)=A(4); % the energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A; x00=real(A); fval exit qm(j,i)=334000*mf(j,i); % W sfrost(j,i)=5*sum(mf(:,i)); % kg mvaw(j,i)=0; % kg/s smvaw(j,i)=5*sum(mvaw(:,i)); % kg qvap(j,i)=mvaw(j,i)*2443*1000; % J s_qvap(j,i)=sum(qvap(:,i))*5; % J watertray(j,i)=0; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg

354

Appendices hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% else if j=8 khri=hri(j,i); % kJ/kg kMr=Mr(j,i); % kg/s kTri=Tri(j,i); % °C kRr=Rr(j,i); % (K˙m2)/W ksmrw=sfrost(j-1,i); % the total retained water on 5*j seconds, kg kTw1=Tw(j-1,i); % the temperature of the melted frost on 5*j seconds, °C % all the input parameters in the function listed here x0=[0.0034 0.0034 0.35 1200 0.001]; % mf=x(1), mrw=x(2), Tw=x(3); qr=x(4); Tro=x(5) the values for debugging; options=optimset(’display’,’off’,’MaxIter’,100000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x)mystage2(x,ksmrw,kTw1,i,kRr, kTri,khri,kMr),x0,options); % kRr1, kTr1 % uw(j,i)=A(1); mf(j,i)=A(1); % the mass of melted frost, kg/s mrw(j,i)=A(2); % the mass of retained water, kg/s Tw(j,i)=A(3); % the temperature of retained water, °C qr(j,i)=A(4); % the energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A x00=real(A); fval exit qm(j,i)=334000.*mf(j,i); % W effq(j,i)=qm(j,i)/qr(j,i); % 1 sfrost(j,i)=5.*sum(mf(:,i)); % kg mvaw(j,i)=0; % kg/s smvaw(j,i)=5.*sum(mvaw(:,i)); % kg hair(j,i)=0; % W/(m2 °C) qair(j,i)=0; % W s_qair(j,i)=sum(qair(:,i))*5; % W hd(j,i)=0; % W/(m2 °C) qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=0; % kg/s

356

Appendices swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % kJ/kg qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W % here is the end of stage 2: frost melting without water flow to down circuit %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% end end end end end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i=1 for j=18:45 % for the 18*5 seconds for the 1st Circuit kmw1=mf(j-1,i)+0.0192; % 0.0192 kg stands for the mass of flowing water is the sum of the melting and the melted on the coil, kg/s ksmrw=sfrost(17,i); % the total mass of water retained on the coil, the flowing part 0.0192 kg was neglected, kg kTw1=Tw(j-1,i); %, °C kTri=Tri(j,i); %, °C kRr=Rr(j,i); % (Km2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001]; % mf=x(1), mr=x(2), Tw=x(3); qr=x(4); Tro=x(5) the values for debugging; options=optimset(’display’,’off’,’MaxIter’,100000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x)mystage31(x,kmw1,ksmrw,kTw1,i, kTri,kRr,kMr,khri),x0,options); mf(j,i)=A(1); % melted water, kg/s; after this stage, mf is 0 kg/s mrw(j,i)=A(2); % retained water, kg/s Tw(j,i)=A(3); % retained water temperature, °C qr(j,i)=A(4); % energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A x00=real(A); fval

Appendices

357

exit qm(j,i)=334000.*mf(j,i); % W sfrost(j,i)=5.*sum(mf(:,i)); % after this stage, sfrost(j,i) =0.350 kg, obtained from the experimental study, kg qair(j,i)=1.4748.*Tw(j,i).^(4/3).*2.6852*2.5*0.55*((sfrost (j-1,i))./0.323).^1.5; % W s_qair(j,i)=sum(qair(:,i))*5; % W hair(j,i)=1.4748.*Tw(j,i).^(1/3); % W/(m2 °C) smvaw(j,i)=5.*sum(mvaw(:,i)); % kg/s hd(j,i)=0; % W/(m2 °C) qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=kmw1; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % kJ/kg qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W if sfrost(j,i)>=0.35; sfrost(j,i)=0.35; % after this stage, sfrost(j,i)=0.350 kg mf(j,i)=0; % at the fourth stage, the mf is always 0 kg/s kTw1=Tw(j-1,i); % the initial values are different for each circuit, °C mr0=0.008; % the water left on the first coil; kg/s smvaw=smvaw(j-1,i); % at the beginning of this stage, it is 0 kg % % % % % %

Coef7=-5800.2206; Coef8=1.3914993; Coef9=-0.04860239; Coef10=0.000041764768; Coef11=-0.000000014452093; Coef12=6.5459673;

T=Tri(j,i)+273.15; % K denspipe=exp(-5800.2206*T.^(-1)+1.3914993*T.^(0)0.04860239*T.^(1)+0.000041764768*T.^(2)-0.000000014452093*T.^(3) +6.5459673*log(T))/(8314./18.*T); % calculate the density of humidity air Tair=0+273.15;% K; % Tair=0 % °C PwSat_Air=exp(-5800.2206*Tair.^(-1)+1.3914993*Tair.^(0)0.04860239*Tair.^(1)+0.000041764768*Tair.^(2)0.000000014452093*Tair.^(3)+6.5459673*log(Tair)); % Pa dens_air=0.80*PwSat_Air/(8314/18*(273.15+0)); % relative_Humi_air=0.80 % 0.0039 density of component outside boundary layer, kg/m3 % PwSat_pipeAir(1,t)=Pressure_Air_Water(Tr(1,t)) % dens_pipe(c,t)=Pressure_Air_Water(Tw(c,t-1)).*10^6./(8314./ 18.*(273.15+Tw(c,t-1)));

358

Appendices % density of gas at interface (saturation density), kg/m3 kTri=Tri(j,i); % °C kRr=Rr(j,i); % (Km2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001]; options=optimset(’display’,’off’,’MaxIter’,10000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x)mystage41(x,kTw1,mr0,smvaw,i, denspipe,dens_air,kTri,kRr,kMr,khri),x0,options); mrw(j,i)=A(1); % retained water, kg/s mvaw(j,i)=A(2); % vaporized water, kg/s Tw(j,i)=A(3); % retained water temperature, °C qr(j,i)=A(4); % energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C

A x00=real(A); fval exit hair(j,i)=1.4748.*Tri(j,1).^(1/3); % W/(m2 °C) qair(j,i)=1.4748.*Tri(j,1).^(4/3).*2.6852*2.5*2; % W s_qair(j,i)=sum(qair(:,i))*5; % W hd(j,i)=hair(j,i)/1005./1.258./0.845^(2/3); % W/(m2 °C) smvaw(j,i)=5.*sum(mvaw(:,i)); % kg qm(j,i)=334000.*mf(j,i); % W qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=0; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % kJ/kg qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W end end end % here is the end of stage 4 for Circuit 1: water layer evaporating stage %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i=2 for j=18:45 % for the 18*5 seconds for the 2nd circuit;

Appendices

359

kmw1=mf(j-1,i-1)+0.0192; % the water comes from 1st circuit, kg/s kmw2=kmw1+mf(j-1,i)+0.0192; % the water left from this 2nd circuit, kg/s ksmrw=sfrost(17,i); % the sum of water retained on the coil, kg kTw1=Tw(j-1,i); % °C kTri=Tri(j,i); % °C kRr=Rr(j,i); % (Km2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001]; % mf=x(1), mr=x(2), Tw=x(3); qr=x(4); Tro=x(5) the values of debugging; options=optimset(’display’,’off’,’MaxIter’,100000, ’MaxFunEvals’,20000);% number [A,fval,exit]=fsolve(@(x)mystage32(x,kmw1,kmw2,ksmrw,kTw1, i,kTri,kRr,kMr,khri),x0,options); mf(j,i)=A(1); % melted water, kg/s; after this stage, mf is 0 kg/s mrw(j,i)=A(2); % retained water, kg/s Tw(j,i)=A(3); % retained water temperature, °C qr(j,i)=A(4); % energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A x00=real(A); fval exit qm(j,i)=334000.*mf(j,i); % W sfrost(j,i)=5.*sum(mf(:,i)); % after this stage, sfrost(j,i) =0.350, kg qair(j,i)=1.4748.*Tw(j,i).^(4/3).*2.6852*2.5*0.50*((sfrost (j-1,i))./0.323).^1.5; % W s_qair(j,i)=sum(qair(:,i))*5; % W hair(j,i)=1.4748.*Tw(j,i).^(1/3); % W/(K˙m2) smvaw(j,i)=5.*sum(mvaw(:,i)); % kg hd(j,i)=0; % W/(Km2) qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=kmw2; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % W/(m2 °C) qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W % here is the end of stage 3 for Circuit 2: frost melting with water flow to down circuit %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

360

Appendices if sfrost(j,i)>=0.35; sfrost(j,i)=0.35; % kg mf(j,i)=0; % at fourth stage, the mf is always 0 kg/s kTw1=Tw(j-1,i); % the initial values are different for each circuit, °C mr0=0.008 ; % the water left on the first coil, kg/s smvaw=smvaw(j-1,i); % at the beginning of this stage, it is 0 kg % % % % % %

Coef7=-5800.2206; Coef8=1.3914993; Coef9=-0.04860239; Coef10=0.000041764768; Coef11=-0.000000014452093; Coef12=6.5459673;

T=Tri(j,i)+273.15; % K denspipe=exp(-5800.2206*T.^(-1)+1.3914993*T.^(0)0.04860239*T.^(1)+0.000041764768*T.^(2)-0.000000014452093*T.^(3) +6.5459673*log(T))/(8314./18.*T); % calculate the density of humidity air, kg/m3 Tair=0+273.15;% K; %Tair=0; % °C; PwSat_Air=exp(-5800.2206*Tair.^(-1)+1.3914993*Tair.^(0)0.04860239*Tair.^(1)+0.000041764768*Tair.^(2)0.000000014452093*Tair.^(3)+6.5459673*log(Tair)); % Pa; dens_air=0.80*PwSat_Air/(8314/18*(273.15+0)); % relative_Humi_air=0.80; % 0.0039 density of component outside boundary layer, kg/m3 % PwSat_pipeAir(1,t)=Pressure_Air_Water(Tr(1,t)); % dens_pipe(c,t)=Pressure_Air_Water(Tw(c,t-1)).*10^6./ (8314./18.*(273.15+Tw(c,t-1))); % density of gas at interface (saturation density), kg/m3 kTri=Tri(j,i); % °C kRr=Rr(j,i); % (Km2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001]; options=optimset(’display’,’off’,’MaxIter’,10000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x)mystage42(x,kTw1,mr0,smvaw,i, denspipe,dens_air,kTri,kRr,kMr,khri),x0,options); mrw(j,i)=A(1); % retained water, kg/s; mvaw(j,i)=A(2); % vaporized water, kg/s; Tw(j,i)=A(3); % retained water temperature, °C; qr(j,i)=A(4); % energy used in defrosting from refrigerant, W; Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C

Appendices

361 A x00=real(A); fval exit

hair(j,i)=1.4748.*Tri(j,1).^(1/3); % W/(K˙m2) qair(j,i)=1.4748.*Tri(j,1).^(4/3).*2.6852*2.5*2; % W s_qair(j,i)=sum(qair(:,i))*5; % W hd(j,i)=hair(j,i)/1005./1.258./0.845^(2/3); % W/(K˙m2) qvap(j,i)=mvaw(j,i)*2443*1000; % W smvaw(j,i)=5.*sum(mvaw(:,i)); % kg qm(j,i)=334000.*mf(j,i); % W qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=0; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % W/(K˙m2) qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W end end end % here is the end of stage 4 for Circuit 2: water layer evaporating stage %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i=3 for j=18:45 % for the 18*5 seconds for the 1st circuit; kmw2=mf(j-1,i-2)+mf(j-1,i-1)+0.0192*2.0; % kg/s kmw3=kmw2+mf(j-1,i)+0.0192; % kg/s ksmrw=sfrost(17,i); % kg kTw1=Tw(j-1,i); % °C kTri=Tri(j,i); % °C kRr=Rr(j,i); % (K˙m2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001];%mf=x(1), mr=x(2), Tw=x(3); qr=x(4); Tro=x(5) the values of debugging options=optimset(’display’,’off’,’MaxIter’,100000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x) mystage33(x,kmw2,kmw3,ksmrw, kTw1,i,kTri,kRr,kMr,khri),x0,options);

362

Appendices mf(j,i)=A(1); % melted water, kg/s; after this stage, mf is 0 kg/s mrw(j,i)=A(2); % retained water, kg/s Tw(j,i)=A(3); % retained water temperature, °C qr(j,i)=A(4); % energy used in defrosting from refrigerant. W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A x00=real(A); fval exit qm(j,i)=334000.*mf(j,i); % W sfrost(j,i)=5.*sum(mf(:,i)); % after this stage, sfrost (j,i)=0.350, kg qair(j,i)=1.4748.*Tw(j,i).^(4/3).*2.6852*2.5*0.45* ((sfrost(j-1,i))./0.323).^1.5; % W s_qair(j,i)=sum(qair(:,i))*5; % W hair(j,i)=1.4748.*Tw(j,i).^(1/3); % W/(K m2) smvaw(j,i)=5.*sum(mvaw(:,i)); % kg hd(j,i)=0; % W/(K m2) qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=kmw3; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % kJ/kg qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W % here is the end of stage 3 for Circuit 3: frost melting with water flow to down circuit %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

if sfrost(j,i)>=0.35; sfrost(j,i)=0.35; % kg mf(j,i)=0; % 4th stage the mf is always 0, kg/s kTw1=Tw(j-1,i); % the initial values are different for each circuit, °C mr0=0.008 ; % the water left on the first coil, kg/s smvaw=smvaw(j-1,i); % at the beginning of this stage, it is 0, kg % % % % % %

Coef7=-5800.2206; Coef8=1.3914993; Coef9=-0.04860239; Coef10=0.000041764768; Coef11=-0.000000014452093; Coef12=6.5459673;

Tair=0+273.15; % K; % Tair=0; % °C

Appendices

363

PwSat_Air=exp(-5800.2206*Tair.^(-1)+1.3914993*Tair.^(0)0.04860239*Tair.^(1)+0.000041764768*Tair.^(2)-0.000000014452093 *Tair.^(3)+6.5459673*log(Tair));% Pa dens_air=0.80*PwSat_Air/(8314/18*(273.15+0));% relative_Humi_air =0.80; % 0.0039 density of component outside boundary layer, kg/m3 % PwSat_pipeAir(1,t)=Pressure_Air_Water(Tr(1,t)) % dens_pipe(c,t)=Pressure_Air_Water(Tw(c,t-1)).*10^6./ (8314./18.*(273.15+Tw(c,t-1))) % density of gas at interface (saturation density), kg/m3 T=Tri(j,i)+273.15; % K denspipe=exp(-5800.2206*T.^(-1)+1.3914993*T.^(0)0.04860239*T.^(1)+0.000041764768*T.^(2)-0.000000014452093*T.^(3) +6.5459673*log(T))/(8314./18.*T); % calculate the density of humidity air, kg/m3 kTri=Tri(j,i); % °C kRr=Rr(j,i); (K˙m2)/W kMr=Mr(j,i); % kg/s khri=hri(j,i); % kJ/kg % all the input parameters in the function listed here x0=[0.0042 0.0042 0.335 1200 0.001]; options=optimset(’display’,’off’,’MaxIter’,10000, ’MaxFunEvals’,20000); % number [A,fval,exit]=fsolve(@(x) mystage43(x,kTw1,mr0,smvaw,i, denspipe,dens_air,kTri,kRr,kMr,khri),x0,options); mrw(j,i)=A(1); % retained water, kg/s mvaw(j,i)=A(2); % vaporized water, kg/s Tw(j,i)=A(3); % retained water temperature, °C qr(j,i)=A(4); % energy used in defrosting from refrigerant, W Tro(j,i)=A(5); % the temperature of tube surface at exit of each circuit, °C A x00=real(A); fval exit hair(j,i)=1.4748.*Tri(j,1).^(1/3); % W/(K m2) qair(j,i)=1.4748.*Tri(j,1).^(4/3).*2.6852*2.5*2; % W s_qair(j,i)=sum(qair(:,i))*5; % W hd(j,i)=hair(j,i)/1005./1.258./0.845^(2/3); % W/(K m2) smvaw(j,i)=5.*sum(mvaw(:,i)); % kg qm(j,i)=334000.*mf(j,i); % W qvap(j,i)=mvaw(j,i)*2443*1000; % W s_qvap(j,i)=sum(qvap(:,i))*5; % W watertray(j,i)=0; % kg/s swatertray(j,i)=sum(watertray(:,i)); % kg

364

Appendices hro(j,i)=44518+1170.36*Tro(j,i)+1.68674*Tro(j,i)^2+5.2703/ 1000*Tro(j,i)^3; % W/(K m2) qr2(j,i)=kMr*(khri-hro(j,i)); % W s_qr2(j,i)=sum(qr2(:,i))*5; % W end end end % here is the end of stage 4 for Circuit 3: water layer evaporating stage %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% for i=1:3 s_qm=0.350*3*334000; % this is fixed value for different cases, W s_qvap1=sum(s_qvap(37,:)); % W s_qr1=sum(s_qr2(37,:))*0.65; % W s_qair1=sum(s_qair(37,:)); % W s_q_heatingmeltedfrost=sum(swatertray(45,:)*Tw(32,3) *4.2*1000); % W Defrostingefficiency=(s_qm+s_qvap1)/s_qr1; % 1 end %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % The following programs are function programs used in the previous main program. function F=mystage1(x,ksmrw,kTw1,i,kRr,kTri,khri,kMr) % solve mf=x(1), mw=x(2), Tw=x(3), qr=x(4), kRr=x(5) F=[x(1)-x(2); % kg/s x(4)-334000.*x(1)-4195.2.*x(2).*x(3)-4195.2.*ksmrw.*(x(3)-kTw1); % J/s 334000.*x(1)-1480.54.*((i+1)^0.5-i^0.5).*2.6852*2.5*x(3); % 0.85 stand for the water side modification x(4)-0.0679*(kTri-2.*x(3))./(kRr+3.6316e-06); % 0.55 is the modification value of refrigerant side x(4)-0.32*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] %0.40 make the s-frost suitable. % -5.2703/ 1000*kTro^3, W end function F=mystage2(x,ksmrw,kTw1,i,kRr,kTri,khri,kMr) % solve mf=x(1), mw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)-x(2); % kg/s x(4)-334000.*x(1)-4195.2.*x(3).*x(2)-4195.2.*ksmrw.*(x(3)-kTw1)1.4748.*x(3).^(4/3).*2.6852*2.5*0.15; % W 0.4 1.4748.*x(3).^(4/3).*2.6852*2.5*0.15+334000.*x(1)-1480.54.*((i +1)^0.5-i^0.5).*2.6852*2.5*0.85*x(3); % 0.85 stand for the area, which is not 2.6852*2.5*2.0

Appendices

365

x(4)-0.0679*(kTri-2.*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-0.32*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.40 is to modify the Mr. -5.2703/1000*kTro^3, W end function F=mystage31(x,kmw1,ksmrw,kTw1,i,kTri,kRr,kMr,khri) % solve mf=x(1), mrw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)-x(2)-kmw1; % mass conservation, frost changes to water, kg/s x(4)-334000.*x(1)-4195.2.*ksmrw*(x(3)-kTw1)-1.4748.*x(3).^(4/3). *2.6852*2.5*0.55; % the energy in refrigerant used in frost melting, retained water temperature improve, heating the air, W % 2.6852*2.5*4.0*0.75 is the actual area of contact fins surface and the ambient air 1.4748.*x(3).^(4/3).*2.6852*2.5*0.55+334000.*x(1)-1480.54.*((i +1)^0.5-i^0.5).*2.6852*2.5*x(3)*0.45; % the heat transfer from water layer; 0.85 stand the area is not 2.6852*2.5*4.0; x(4)-0.0679*1.6*(kTri-2.*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-0.42*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.40 is to modify the Mr.-5.2703/1000*kTro^3, W end function F=mystage32(x,kmw1,kmw2,ksmrw,kTw1,i,kTri,kRr,kMr,khri) % solve mf=x(1), mrw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)+kmw1-x(2)-kmw2; % mass conservation, frost changes to water, kg/s x(4)-334000.*x(1)-4195.2.*(ksmrw+kmw1-kmw2)*(x(3)-kTw1)-1.4748. *x(3).^(4/3).*2.6852*2.5*0.50; % the energy in refrigerant used in frost melting, retained water temperature improve, heating the air % 2.6852*2.5*4.0*0.75 is the actual area of contact fins surface and the ambient air 1.4748.*x(3).^(4/3)*2.6852*2.5*0.50+334000.*x(1)-1480.54.*((i+1) ^0.5-i^0.5).*2.6852*2.5*x(3)*0.50; % the heat transfer from water layer; 0.85 stand for the area, which is not 2.6852*2.5*4.0; x(4)-0.0679*1.3*(kTri-2*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-0.375*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.30 is to modify the Mr.-5.2703/1000*kTro^3, W end function F=mystage33(x,kmw2,kmw3,ksmrw,kTw1,i,kTri,kRr,kMr,khri) % solve mf=x(1), mrw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)+kmw2-x(2)-kmw3; % mass conservation, frost changes to water, kg/s x(4)-334000.*x(1)-4195.2.*(ksmrw+kmw2-kmw3).*(x(3)-kTw1)-1.4748. *x(3).^(4/3).*2.6852*2.5*0.45; % the energy in refrigerant used in frost melting, retained water temperature improve, heating the air

366

Appendices 1.4748.*x(3).^(4/3).*2.6852*2.5*0.45+334000.*x(1)-1480.54.*((i +1)^0.5-i^0.5).*2.6852*2.5*0.55*x(3); % the heat transfer from water layer % 2.6852*2.5*4.0*0.75 is the actual area of contact fins surface and the ambient air; x(4)-0.0679*1.0*(kTri-2*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-0.33*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.40 is to modify the Mr. -5.2703/1000*kTro^3, W end function F=mystage41(x,kTw1,mr0,smvaw,i,denspipe,dens_air,kTri, kRr,kMr,khri) % solve mrw=x(1), mvaw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)-x(2); % the water left on the coil is the water vaporized from this coil; mass conservation law x(4)-4195.2*((mr0-smvaw-x(2))*x(3)-(mr0-smvaw)*kTw1)-1480.54.* ((i+1)^0.5-i^0.5).*2.6852*2.5*0.1*x(3)-1.4748*x(3)^(4/3) *2.6852*2.5*0.9-2443*1000*x(2); % the energy comes from refrigerant was used in improve the temperature % of retained water, and heat the ambient air, and retained water % vaporized; energy conservation, 2.0 x(2)-1.4748.*x(3).^(1/3)./1005./1.258./0.845^(2/3).*(denspipedens_air)*2.6852*2.5*0.9*((mr0-smvaw)./mr0).^1.5*7.8; % the mass of water vaporized into the ambient air is equal with the x(4)-0.0679*7.0*(kTri-2.0*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-1.35*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.30 is to modify the Mr. -5.2703/1000*kTro^3, W end function F=mystage42(x,kTw1,mr0,smvaw,i,denspipe,dens_air,kTri, kRr,kMr,khri) % solve mrw=x(1), mvaw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)-x(2); % the water left on the coil is the water vaporized from this coil; mass conservation law x(4)-4195.2*((mr0-smvaw-x(2))*x(3)-(mr0-smvaw)*kTw1)-1480.54.* ((i+1)^0.5-i^0.5).*2.6852*2.5*0.2*x(3)-1.4748*x(3)^(4/3) *2.6852*2.5*0.8-2443*1000*x(2); % the energy comes from refrigerant was used in improve the temperature % of retained water, and heat the ambient air, and retained water % vaporized; energy conservation, 2.0

Appendices

367

x(2)-1.4748.*x(3).^(1/3)./1005./1.258./0.845^(2/3).*(denspipedens_air)*2.6852*0.8*2.5*((mr0-smvaw)./mr0).^1.5*7.8; % the mass of water vaporized into the ambient air is equal with the x(4)-0.0679*5.5*(kTri-2.0*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-1.25*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.30 is to modify the Mr. -5.2703/1000*kTro^3, W end function F=mystage43(x,kTw1,mr0,smvaw,i,denspipe,dens_air,kTri, kRr,kMr,khri) % solve mrw=x(1), mvaw=x(2), Tw=x(3), qr=x(4), Rr=x(5) F=[x(1)-x(2); % the water left on the coil is the water vaporized from this coil; mass conservation law x(4)-4195.2*((mr0-smvaw-x(2))*x(3)-(mr0-smvaw)*kTw1)-1480.54.* ((i+1)^0.5-i^0.5).*2.6852*2.5*0.3*x(3)-1.4748*x(3)^(4/3) *2.6852*2.5*0.7-2443*1000*x(2); % the energy comes from refrigerant was used in improve the temperature % of retained water, and heat the ambient air, and retained water % vaporized; energy conservation, 2.0 x(2)-1.4748.*x(3).^(1/3)./1005./1.258./0.845^(2/3).*(denspipedens_air)*2.6852*0.7*2.5*((mr0-smvaw)./mr0).^1.5*7.8; % the mass of water vaporized into the ambient air is equal with the x(4)-0.0679*4.0*(kTri-2.0*x(3))./(kRr+3.6316e-06); % 0.755 is the modification value of refrigerant side x(4)-1.15*kMr*(khri-44518-1170.36*x(5)-1.68674*x(5)^2-5.2703/ 1000*x(5)^3)] % 0.30 is to modify the Mr. -5.2703/1000*kTro^3, W end

Index Note: Page numbers followed by f indicate figures and t indicate tables. refrigerant mass flow rate, 101–102, 103f, 105–106f surface tension on, 138 tube surface temperature, 101, 102f, 104–106, 107–108f with two refrigerant circuits, 48–51, 48f, 49t, 50–51f

A Absorption heat pumps, 1 Air jet technique, 17, 345 Air source heat pump (ASHP) unit control strategy optimization, 345 defrosting methods compressor shutdown, 23–25, 27t electric heating, 25–26, 27t frost suppression vs., 25t hot gas bypass, 26, 27t hot water spraying, 26, 27t reverse cycle, 26–28, 27t energy transfer mechanism, 224 local drainage of melted frost, 242 mechanism study, 345 melted frost elimination, on uneven defrosting, 118–121, 120t model development, 345 multicircuit outdoor coil, 155–156 new methods and materials, 345 outdoor coil, 5–7 refrigerant distribution, melted frost flowing, 204 running cost (see Running cost, ASHP unit) surface tension, on uneven defrosting, 138 system and component optimization, 345 uneven defrosting, 95–110 airside of three-circuit outdoor coil, 98, 98f assumptions, 102–103 defrosting durations, 104, 104t experimental cases, 99–100, 100t, 100f experimental study, 96–98 heat supply and energy consumption, 108–110, 109f melted frost elimination on, 118–121, 120t modeling study, 98–99 outdoor coil, 96, 97f refrigerant distribution, 100f, 101

C Carnot cycle, 11 Clean development mechanism (CDM), 115 Coefficient of performance (COP), 1, 163f, 164 Compression heat pumps, 1 Compressor shutdown defrosting (CSDD) method, 23–25, 27t Convective heat transfer, 77 Convective heat transfer coefficient, 78 CSDD. See Compressor shutdown defrosting (CSDD) method D DAS. See Data-acquisition system (DAS) Data-acquisition system (DAS), 96, 119–121 Data analysis and validation, defrosting control strategy DTT periods and nodes, 296–298 experimental results, 290 RCD termination temperature, 299–300 tube surface temperature analysis, 290 DEC. See Defrosting evenness coefficient (DEC) Defrosting, 3–5 demand-based, 35 efficiency, 67 methods compressor shutdown, 23–25, 27t electric heating, 25–26, 27t frost suppression vs., 25t hot gas bypass, 26, 27t

370

Defrosting (Continued) hot water spraying, 26, 27t reverse cycle, 26–28, 27t Defrosting control strategy artificial intelligence, 257 frost accumulation, 257–258 holographic interferometry technique, 257–258 infrared thermometer, 257–258 initiation, 35–36 Internet of Things, 257 melted frost, local drainage of compressor suction and discharge, measured pressures of, 279–280, 281f durations of melted frost, 281t energy supply and effective energy consumption, 281–283, 283f flow chart, 274f frost accumulation, 274, 284f frosting durations, p0110 indoor air thermal and electricity, 280–281, 282f mal-defrosting problems, 273 measured temperatures, 278–279, 280f outdoor coil airside surface conditions, 275–276, 276f preheating stage of, 275–277 pressure difference between suction and discharge, 279–280, 282f refrigerant distribution resistance, 278–279 system stability and indoor thermal comfort, 284–285 trial-and-error manual adjustments, 274 termination of, 36–38 data analysis and validation, 290–300 enhanced heat transfer efficiency, 285–286 experimental cases, 288–290 methodology, 286–287 minimized refrigerant pressure loss, 285–286 surface temperature of outdoor coil, 285–286 time-based defrosting initiation control strategy, 258 DX A/C system, 260–261 energy supply and, 271f

Index

flow chart of procedure, 261f fluctuation of tube surface temperature, 264–265f frost accumulation, 260, 272f indoor coil air temperature difference, fluctuation of, 269f melted frost, fluctuation of, 266f outdoor coil airside surface conditions, 262–263, 262–263f refrigerant pressure difference, fluctuation of, 267f refrigerant volumetric flow rate, fluctuation of, 266f trial-and-error manual adjustments, 260 unequal frost accumulation, 258 Yao’s distributed mathematical frosting model, 257–258 Defrosting duration, 343 Defrosting efficiency, 225, 343 calculation error, 348 definition, 347 energy transfer mechanism, 223, 225 frosting evenness coefficient, 174–175, 174f, 189, 190f metal energy storage effect, 349 technoeconomic performances, 303 uneven defrosting, on outdoor coil, 67 Defrosting evenness coefficient (DEC), 132, 133t, 351 Defrosting evenness status, 343 Defrosting performance, 344 air wet-bulb temperatures, 229–230 defrosting efficiency, 225 energy input, to compressor, 236f energy supplies and consumptions, 225 experimental cases, 230 experimental results, 239t experimental setup, 225 heat consumptions, 238f, 240 heat supplies, 237f, 250f local drainage of melted frost airside surface conditions of outdoor coil, 245 air temperature differences, 246–249 ASHP unit, 242 experimental cases, 242–243 heat consumption, 252 mean measured tube surface temperatures, 247–248f

Index

MES effects, 241–242, 252 temperature variations, 249–251, 249f MES calculations, 231 MES effect, 241t precalibrated K-type thermocouples, 229–230 prototypical three-circuit indoor coil, 227t refrigerant distribution energy analysis, 202–204 experimental cases, 194–197 experimental results, 197–202 tailor-made three-circuit outdoor coil, 227t temperature differences, indoor and outdoor coils, 234–235 temperature values, 239–240 three-circuit outdoor coil, 225 three water-collecting cylinders, 228 three-working-circuit outdoor coils, 241 total mass of frost accumulation, 228 two-working-circuit, 225, 241 Defrosting termination temperature (DTT), 36–37, 37t Demand-based defrosting, 35 Direct expansion air-conditioning (DX A/C) system, 50 Dittus-Boelter correlation, 78 Downward-flowing melted frost, 343 Dry-bulb temperature sensors, 119

371

heat supplies, 237f local drainage of melted frost, 241–253 MES calculations, 231 and MES effect, 241t precalibrated K-type thermocouples, 229–230 prototypical three-circuit indoor coil, 227t tailor-made three-circuit outdoor coil, 227t temperature differences, indoor and outdoor coils, 234–235 temperature values, 239–240 three-circuit outdoor coil, 225 three water-collecting cylinders, 228 three-working-circuit outdoor coils, 241 total mass of frost accumulation, 228 two-working-circuit, 225, 241 dynamic energy transfer, 223–224 measured tube surface temperature of outdoor coil circuit, 234f melted frost, effects of, 223, 253 periodic defrosting, 223 system performance improvement, 223 thermal comfort, 254–255 vaporize melted frost, 223 Environmental chamber, 50, 50f, 118, 155–156 Even frosting, 204–205

E Electric heating defrosting (EHD), 25–26, 27t Electronic expansion valve (EEV), 214–215 Energy performance improvement, 304 Energy transfer mechanism, 344 airside surface conditions of outdoor coil, 232–233 ASHP unit, 224 defrosting duration and defrosting efficiency, 223 defrosting performance air wet-bulb temperatures, 229–230 defrosting efficiency, 225 energy input, to compressor, 236f energy supplies and consumptions, 225 experimental cases, 230 experimental results, 239t experimental setup, 225 heat consumptions, 238f, 240

F FEC. See Frosting evenness coefficient (FEC) Fin surface, coating treatment on, 18–19, 20–21t Fin-type adjustment, 18 Fraction force (Ff ), 123 Frost accumulation, 153–154, 154f Frost deposition, 3–4 Frost formation, on cold flat plate surface, 3–4, 3f Frosting evenness coefficient (FEC), 350 defrosting performances, 164–175 airside surface conditions of outdoor coil, 166–167, 166f defrosting durations, 171–172, 174t, 174f defrosting efficiency, 174–175, 174f differential analysis, 174t

372

Frosting evenness coefficient (FEC) (Continued) energy analysis, 173–175 energy supplies, 173, 173f experimental cases, 165, 166t fin surface temperature, 167–172, 169–170f with local drainage of melted frost, 175–191 melted frost temperature, 167–173, 171f refrigerant volumetric flow rate, 167–172, 170f tube surface temperatures, 167–172, 167–168f multicircuit outdoor coil airside surface conditions, 154, 155f, 158, 159f air temperature, 162f, 164 ASHP unit, 155–156 COP variation, 163f, 164 even frosting, control method of, 156–157 experimental cases, 155–157, 159t fin surface temperatures, 163f, 164 refrigerant pressure drop, 159–160, 161f, 162–163 refrigerant volume flow rate, 156–157, 159–160, 161f, 162–163 tube surface temperatures, 156–157, 157–158f, 159–162, 160f, 162f Frosting mechanism, 3–5 Frost-suppression measures, 12–23, 12f air jet technique, 17 external heating source, 22–23 external type, 12 internal type, 12 outdoor coil coating treatment on fin surface, 18–19, 20–21t fin and tube geometry adjustment, 17–18 fin-type adjustment, 18 increasing inlet airflow rate, 16 preheating inlet air, 13–16 reducing inlet air humidity, 13 solid and liquid desiccants for, 13, 14–15t two-stage technique, 22 ultrasonic vibration technique, 16–17 vapor-injection technique, 19–22

Index

G Gravity force (G), 123 Ground-source heat pumps (GSHPs), 1–2 H Heat exchangers, 116–117 Heat pumps, 1 absorption, 1 compression, 1 heat sources, 1–2 Holographic interferometry technique, 36 Horizontal heat exchangers, 116–117 Horizontal installation, 116, 117f, 118, 119f, 123–124 Hot gas bypass defrosting (HGBD), 26 Hot water spraying defrosting (HWSD) method, 26, 27t HWSD. See Hot water spraying defrosting (HWSD) method Hybrid heat pump systems, 2–3 Hygrosensor, 50 I Infrared thermometer, 36 Inlet air humidity, 13 preheating, 13–16 Inlet airflow rate, 16 K K-type thermocouples, 50–51, 119–121, 229–230 L LGUs. See Load-generating units (LGUs) Liquid refrigerant heat transfer coefficient, 78 Load-generating units (LGUs), 50, 96, 118 Local drainage, of melted frost frosting evenness coefficient with airside surface conditions of outdoor coil, 178–179, 179f defrosting durations, 188–189, 188t, 190f defrosting efficiency, 189, 190f downward-flowing melted frost, 176, 177t energy analysis, 188–191, 189t, 190f experimental cases, 177–178, 178t

Index

373

air temperature, 162f, 164 ASHP unit, 155–156 COP variation, 163f, 164 even frosting, control method of, 156–157 experimental cases, 155–157, 159t fin surface temperatures, 163f, 164 refrigerant pressure drop, 159–160, 161f, 162–163 refrigerant volume flow rate, 156–157, 159–160, 161f, 162–163 tube surface temperatures, 156–157, 157–158f, 159–162, 160f, 162f

fin surface temperatures, 179, 181–182f, 186 frost accumulations, 176, 177f melted frost temperatures, 179, 183–185f, 187–188 refrigerant volumetric flow rate, 185f, 188 tube surface temperatures, 179–183, 180–181f M Mal-defrosting, 36, 273 Manual stop valve (MV), 48 Melted frost, 52–53, 57–58 Melted frost downward flowing (MFDF), 204 Melted frost elimination, on uneven defrosting, 116–135 airside surface conditions of outdoor coil, 119, 120f, 124–126, 125f, 127f ASHP unit, 118–121, 120t defrosting duration, 132–134, 133–134t defrosting evenness coefficient, 132, 133t energy performance analysis, 134, 135t experimental cases, 122–124 experimental conditions, 121–122, 121t, 124t fin surface temperatures, 126, 129–130f, 131–132 force analysis of retained water droplets, 123, 123f horizontally installed three-circuit outdoor coil, 119f mass transfer of retained water, 123–124, 124f measurement/calculation errors, 121–122, 122t three-circuit outdoor coil, 118, 119f, 120t tube surface temperatures, 126–131, 128–129f vertically installed outdoor coil, 117f Metal energy storage (MES) effect, 349 Moving boundary technique, 33 Multicircuit outdoor coil, 193–194, 200–201, 204–205, 303–304, 321–322, 324, 338–339 frosting evenness coefficient airside surface conditions, 154, 155f, 158, 159f

O Outdoor coil. See also Uneven defrosting in ASHP unit, 5–7 defrosting process, 33, 33f frost-suppression measures coating treatment on fin surface, 18–19, 20–21t fin and tube geometry adjustment, 17–18 fin-type adjustment, 18 increasing inlet airflow rate, 16 preheating inlet air, 13–16 reducing inlet air humidity, 13 multicircuit outdoor coil, 193–194, 200–201, 204–205, 303–304, 321–322, 324, 338–339 P Payback period, 303–304, 324, 340 Phase change materials (PCM), 29 Photocoupler, 36 Polyvinyl chloride (PVC), 48 R RCD. See Reverse cycle defrosting (RCD) Refrigerant charge compensator, 28 Refrigerant distribution, 344 defrosting performance energy analysis, 202–204 experimental cases, 194–197 experimental results, 197–202 downward-flowing melted frost, 193 gravity and tube internal resistance, 193–194 gravity force, 193–194

374

Refrigerant distribution (Continued) melted frost flowing airside surface conditions, of outdoor coil, 208–209, 208f ASHP unit, 204 compressor suction and discharge pressures, 214–215 conditions of, 206f effects of, 219–220 electronic expansion valve, 214–215 energy analysis, 216–219 experimental work, 206–207 gravity and tube internal resistance, 207–208 measured tube surface temperatures, 209–212, 209–210f, 212f practical applications, 205 refrigerant volumetric flow rate, 214–215, 214f temperature difference of the outdoor coil entrance and exit, 212–214, 213–214f trial-and-error manual adjustments, 205–207 tube internal resistance and gravity, 205–206 semiempirical mathematical models, 193 tube surface temperature, 193 two-circuit and three-circuit outdoor coils, 193 Refrigerant distribution evenness values (RDEVs), 195 Relative humidity (RH), 11 Residual water, 57–58 Retained melted frost, 139, 140f Retained water, 123–124, 124f Reverse cycle defrosting (RCD), 26–28, 27t, 115 defrosting model, 33–34, 33f, 34t energy, 26–27 improvements for, 28–34 airflow and refrigerant distribution adjustment, 29, 32t basic component optimization, 28, 32t PCM-TES-based, 29, 30–32t sensible heat defrosting method, 29–32, 32t Reynolds numbers, 17–18 Running cost, ASHP unit, 303–305, 310–313

Index

refrigeration adjustment valve, influence of cooling season, 316–318 economic analysis, 323–324 heating season, 318, 319f with frost formation, 313–315, 321–322 without frost formation, 316, 321–322 maintenance costs, 323 and operating durations, 318, 318f proportion of additional first cost, 318, 322f ratio of total cost differences, 318, 321f system operating durations, 318, 319f total cost of, 318, 320f water-collecting tray additional initial cost effect on total cost, 335f, 338–339 cooling season, 331–335 defrosting stage, 331, 332–333f frosting/defrosting cycle, 331, 333f heating seasons with frost formation, 337 heating seasons without frost formation, 337 indoor air thermal energy, 337–338 proportion of additional first cost, 331, 336f proportion of first cost in total cost, 331, 335f running cost difference, 338–339 total cost differences, 331, 335f S Sensible heat defrosting method, 29–32 Solar-assisted heat pump, 2–3 Solenoid valve (SV), 48 Surface tension, on uneven defrosting, 136–149 airside surface conditions of outdoor coil, 140–141, 141–142f ASHP unit, 138 defrosting duration, 145–146, 146t energy analysis, 147–149, 147f, 148t experimental cases, 138–139 experimental conditions, 140t face velocity of outdoor coil, 138, 139f fin surface temperatures, 143–145, 144f mass transfer of retained melted frost, 139, 140f tube surface temperatures, 143–145, 143f

Index

T Technoeconomic performances defrosting duration, 303 defrosting efficiency, 303 energy performance improvement, 304 frosting and defrosting state assumptions, 304 hot refrigerant tube and fins, 303 multicircuit outdoor coil, 303 novel RCD method, 303–304 refrigeration adjustment valve, influence of cooling assumptions, 312 defrosting assumptions, 311 defrosting operations, airside surface conditions of, 308–309, 308f economic analysis model, 317 economic analysis process, 307 first costs, 312 flow chart of methodology, 305f frosting assumptions, 310 frosting/defrosting cycle, 306 frosting operations, airside surface conditions of, 307–308, 307f fundamental assumptions, 309 installation of valves and trays, 305–306 performance parameters, 305, 306t refrigerant distribution, 307 running cost, 313 swing-type compressor, 305 three-circuit outdoor coil, 305–306 and water-collecting tray, 324–340 water-collecting trays, 305–306 Thermal conductivity, 78–79 Thermal energy storage (TES) system, 28 Thermal expansion valve (TEV), 28 Three-circuit outdoor coil, 116, 119f Time-based defrosting initiation control strategy, 258, 344–345 DX A/C system, 260–261 energy supply and, 271f flow chart of procedure, 261f fluctuation of tube surface temperature, 264–265f frost accumulation, 260, 272f indoor coil air temperature difference, fluctuation of, 269f melted frost, fluctuation of, 266f

375

outdoor coil airside surface conditions, 262–263, 262–263f refrigerant pressure difference, fluctuation of, 267f refrigerant volumetric flow rate, fluctuation of, 266f trial-and-error manual adjustments, 260 Two semiempirical models, 73–85, 74t airside of three-circuit outdoor coil, 73–74, 73f assumptions and calculation conditions, 75–76 computational algorithm, 84, 85f defrosting process, 74 development of, 75–85 energy used from refrigerant, 91–92, 93f experimental validation, 86–90 frost melting without water flowing away from circuit, 79–81, 80f frost melting with water flowing away from circuit, 81–82 limitations, 94 mass and energy flows in defrosting stages, 80f, 84f mass of melted frost, 91–92, 93f melted frost temperatures, 87–90, 87–89f melted water temperature, 91–92, 92f model extrapolation, 90–91 preheating, 76–79 refrigerant mass flow rate, 90–91, 91f refrigerant temperature, 90–91, 90f thermal resistance of refrigerant, 91–92, 92f tube surface temperatures, 86–87, 86f, 88f uses, 94 water-collecting cylinder, 83–84, 85f water-collecting trays, 73–74, 83–84, 85f water layer vaporizing, 82–83 Two-stage technique, 22 U Ultrasonic vibration technique, 16–17 Uneven defrosting, 303, 343 for ASHP unit, 95–110 airside of three-circuit outdoor coil, 98, 98f assumptions, 102–103 defrosting durations, 104, 104t

376

Uneven defrosting (Continued) experimental cases, 99–100, 100t, 100f experimental study, 96–98 heat supply and energy consumption, 108–110, 109f modeling study, 98–99 outdoor coil, 96, 97f refrigerant distribution, 100f, 101 refrigerant mass flow rate, 101–102, 103f, 105–106f tube surface temperature, 101, 102f, 104–106, 107–108f five-parallel refrigerant circuit outdoor coil, 48, 49t, 49f frosting evenness coefficient, 350 melted frost elimination on, 116–135 airside surface conditions of outdoor coil, 119, 120f, 124–126, 125f, 127f ASHP unit, 118–121, 120t defrosting duration, 132–134, 133–134t defrosting evenness coefficient, 132, 133t energy performance analysis, 134, 135t experimental cases, 122–124 experimental conditions, 121–122, 121t, 124t fin surface temperatures, 126, 129–130f, 131–132 force analysis of retained water droplets, 123, 123f horizontally installed three-circuit outdoor coil, 119f mass transfer of retained water, 123–124, 124f measurement/calculation errors, 121–122, 122t three-circuit outdoor coil, 118, 119f, 120t tube surface temperatures, 126–131, 128–129f vertically installed outdoor coil, 117f performance of, 344 semiempirical mathematical models, 73–85 assumptions, 75–76 calculation conditions, 75–76 experimental validation, 86–90 limitations, 94 model development, 76–85, 352–367 model extrapolation, 90–91

Index

uses, 94 surface tension on, 136–149 airside surface conditions of outdoor coil, 140–141, 141–142f ASHP unit, 138 defrosting duration, 145–146, 146t energy analysis, 147–149, 147f, 148t experimental cases, 138–139 experimental conditions, 140t face velocity of outdoor coil, 138, 139f fin surface temperatures, 143–145, 144f mass transfer of retained melted frost, 139, 140f tube surface temperatures, 143–145, 143f three-circuit experimental study, 57–67 airside surface conditions, 60, 61f defrosting efficiency, 67 downward flowing of melted frost, 60, 60t experimental cases, 60 experimental setup, 58–59 fin surface temperatures, 62, 64–65f, 65 measurement/calculation errors, 59–60t melted frost temperatures, 65–66, 66f tailor-made three-circuit outdoor coil, 58, 59t tube surface temperatures, 62, 62–63f valve location, 58f with two refrigerant circuits, 47–57 airside surface conditions, 50, 51f, 52, 53f ASHP unit, 48–51, 48f, 49t, 50–51f experimental cases, 51–52, 52t experimental conditions, 51, 52t experimental setup, 48–51 fin surface temperatures, 53, 54–55f, 55–56 melted frost temperature, 53, 56, 56f RCD energy sources, 57 vs. three-circuit outdoor coil, 67, 68t tube surface temperature, 53–55, 54–55f

V Valve adjustment, 200–201 Vapor-injection technique, 19–22

Index

W Water-collecting tray, in ASHP unit, 48, 94, 96–97, 304, 343 airside surface conditions, 326t, 327 cooling assumptions, 330 defrosting assumptions, 329 defrosting experimental cases, 325–327, 326t durations in frosting/defrosting cycle, 325–327, 327f economic analysis, methodology of, 325f frosting assumptions, 328 frosting operation performance, 325 fundamental assumptions, 328 prototype condition, 325–327, 326f running costs

377

additional initial cost effect on total cost, 335f, 338–339 cooling season, 331–335 defrosting stage, 331, 332–333f frosting/defrosting cycle, 331, 333f heating seasons with frost formation, 337 heating seasons without frost formation, 337 indoor air thermal energy, 337–338 proportion of additional first cost, 331, 336f proportion of first cost in total cost, 331, 335f running cost difference, 338–339 total cost differences, 331, 335f Water-source heat pumps (WSHPs), 1–2 Wet-bulb temperature sensors, 119