Dams and Reservoirs in Evaporites [1st ed.] 978-3-030-18520-6;978-3-030-18521-3

Table of contents :
Front Matter ....Pages i-xvi
Distribution of Evaporite Karst in the World (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 1-8
Properties of Evaporites (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 9-20
Karstification of Evaporites (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 21-33
Characterization of Surface and Underground Karst Features (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 35-51
Geohazards Associated with Dams and Reservoirs (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 53-63
Prevention and Remediation at Dam Sites and Reservoirs (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 65-92
Methods of Investigation (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 93-108
Monitoring (Petar Milanović, Nikolai G. Maximovich, Olga Meshcheriakova)....Pages 109-114
Overview of Dams and Reservoirs in Evaporites (Petar Milanović, Nikolay Maksimovich, Olga Meshcheriakova)....Pages 115-157

Citation preview

Advances in Karst Science

Petar Milanović Nikolay Maksimovich Olga Meshcheriakova

Dams and Reservoirs in Evaporites

Advances in Karst Science Series Editor James LaMoreaux, Tuscaloosa, AL, USA

This book series covers advances in the field of karst from a variety of perspectives to facilitate knowledge and promote interaction thereby building stepping stones in this process. Methodologies, monitoring, data analysis and interpretation are addressed throughout the wide range of climatic, geological and hydrogeological contexts in which karst occurs. Case studies are presented to provide examples of advancement of the science. Issues to be addressed include water supply, contamination, and land use management. These issues although occurring on a local basis share many of the same features on the global stage. This book series is a critical resource to the scientific community allowing them to compare problems, results, and solutions. The presented information can be utilized by decision makers in making decisions on development in karst regions. Contributions presented may be used in the classroom and to work with stakeholders, scientists, and engineers to determine practical applications of these advances to common problems worldwide. The series aims at building a varied library of reference works, textbooks, proceedings, and monographs, by describing the current understanding of selected themes. The books in the series are prepared by leading experts actively working in the relevant field. The book series Advances in Karst Science includes single and multi-authored books as well as edited volumes. The Series Editor, Dr. James W. LaMoreaux, is currently accepting proposals and a proposal document can be obtained from the Publisher.

More information about this series at http://www.springer.com/series/15147

Petar Milanović
Nikolay Maksimovich
Olga Meshcheriakova

Dams and Reservoirs in Evaporites

123

Petar Milanović Belgrade, Serbia Olga Meshcheriakova Institute for Natural Sciences Perm State University Perm, Russia

Nikolay Maksimovich Institute for Natural Sciences Perm State University Perm, Russia

ISSN 2511-2066 ISSN 2511-2082 (electronic) Advances in Karst Science ISBN 978-3-030-18520-6 ISBN 978-3-030-18521-3 (eBook) https://doi.org/10.1007/978-3-030-18521-3 © Springer Nature Switzerland AG 2019 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Foreword

The relationship of the Milanovic and LaMoreaux families goes back many years through Petar and my father, Dr. Phil LaMoreaux, and today through Petar’s son Sasa’s family and mine. Throughout this time, we have shared many activities both professional and personal, so it was a special honor when Petar asked me to write a forward for his new book. Dams and reservoirs in evaporites are certainly an excellent culmination of Petar’s investigations into carbonate and evaporite terrains and worthy of the wide readership which we are pleased to be able to offer through Springer’s worldwide marketing and distribution network. This is one of many Springer publications that I and my father have worked with Petar on over the years. In PE LaMoreaux & Associates, Inc. (PELA) library, we have a signed copy of Petar’s book Karst Hydrogeology published in 1981. Petar gave it to Dr. LaMoreaux when he and his family came to Alabama in the 1980s and used our family’s home as the base for their trip throughout the US national parks and US geological survey and state geological survey offices around the country. Petar, Dr. LaMoreaux, Sasa, and I have served on and continue to serve as members of the International Association of Hydrogeologists (IAH) Karst Commission which has been and continues to be a catalyst for much of the research and many of the publications of these distinguished professionals and many others. It is international organizations like IAH and the International Association of Engineering Geology and the Environment that were also a factor in bringing Petar, and Perm State University colleagues Nikolay Maksimovich and Olga Meshcheriakova, together to write this book. Russian professionals have performed many investigations of dam construction in karstified rocks, and many of the publications therefrom are subsequently in Russian and English. I also had the pleasure to become familiar with Nikolai through the chapter he authored in Hypogene Karst Regions and Caves of the World (2017, Springer). It is a companion to this book in the Cave and Karst Systems of the World Book Series of which I am the editor and Springer is the publisher. At the recent Karst Symposium: Expect the Unexpected in Trebinje, Bosnia–Herzegovina, I had the opportunity to interact with Petar, Nikolai, Olga, and Derek Ford when they were working on the final draft of the book. Discussing their findings firsthand revealed the critical nature of these concerns and what has been done and is yet to be done to address them. Perhaps the final item to cover in my forward is the accolades and best wishes that were bestowed upon Petar at this symposium. Professionals from around the world came together to celebrate Petar’s 80th birthday and present him with gifts unique to his work and/or their friendships. When you experience this outpouring of admiration and acknowledgment of one man’s contribution to the field, you know that what you are about to read is a singular tome. Tuscaloosa, AL, USA

Jim Lamoreaux

v

Preface

Dam construction in karstified rocks cannot be treated as a routine undertaking. Due to the nature of karst, the differences of significant geological properties from the norm in other rocks are so extensive that they make each dam construction job exceptional. Particularly exceptional, sometimes highly problematic, is the construction of dams and reservoirs in rock formations that include even minor proportions of evaporite rock deposits. From the beginning of the twentieth century, a number of dams and reservoirs have been affected by gypsum and salt dissolution problems. Some of the dams failed to retain water up to their design levels due to extensive leakage, some of them collapsed catastrophically, others were abandoned, and some reservoirs suffered severe pollution of the stored water with solutes. Numerous dams in evaporites, mostly in gypsum, have needed costly rehabilitation. The large amount of experience accumulated during these works has been presented in many different references and languages, chiefly in English and Russian. Many of these papers and reports are not easily accessible to the wide spectrum of professionals working in dam geology today. A major purpose of this book is to summarize and present them in a concise form. Particularly important are examples of dam failures due to the conventional grouting technologies proving to be inadequate despite the high quality of remedial works (mostly grouting) undertaken. As a consequence of the high solubility of evaporites, the watertight resistance of grouting structures can quickly become reduced or eliminated by post-grouting processes. To minimize the impact of these hazardous post-grouting processes, different grout mixes and chemical components have been analyzed in laboratories and tested at sites. Because, currently, there are many dam projects under construction or in the design or investigation phases that are located in areas directly or indirectly influenced by evaporites, and the list of problem dams presented in this text cannot be final. The book is organized into nine chapters. The first four chapters cover general information, including the global distribution of evaporite formations, the properties and processes of karstification as well as an explanation of the different karst forms that may develop in evaporites. Chapter 5 presents the most frequent geohazards associated with dams and reservoirs in evaporite rocks. Chapter 6 describes the range of geotechnical methods frequently applied to improve the evaporite rock mass during the construction of dams and the remediation of their foundations while they are in operation. There is a particular focus on rock waterproofing by grouting and the development of new techniques to construct grout curtains resistant to post-grouting destructive processes. The most frequently applied methods for investigating the basic hydrogeological and geotechnical properties of new sites are described in Chap. 7. Due to the rapid increase of bedrock solution rates that is to be expected during and after the reservoir head rises to the design elevation, the importance of carefully designed monitoring is discussed in Chap. 8. The final Chap. 9 contains a list of more than 80 dams known to the authors that are located on evaporites with brief summaries of 50 of them. The presentation of case studies is not evenly balanced because the levels of information available vary a great deal; for some, there is only basic information, while others can be presented in detail. The work concludes with a list of 245 key references pertaining to the specific topic of dams and reservoirs in evaporites treated in the book. vii

viii

Preface

The authors gratefully acknowledge Dr. Derek Ford, Emeritus Professor, McMaster University, Canada, for his great efforts and help in reviewing this manuscript. We would like to express lot of thanks to him for editing the language in the manuscript in detail and for many of his valuable suggestions and instructions to improve the quality of the book. Belgrade, Serbia Perm, Russia Perm, Russia January 2019

Petar Milanović Nikolay Maksimovich Olga Meshcheriakova

Introduction

Sulfate rocks, particularly gypsiferous formations, at the surface or in the subsurface under younger deposits, cover large regions of the world. According to Maximovich (1962), the continental extent of gypsiferous rocks globally is approximately 7 million km2; Ford and Williams (1989) estimated that gypsiferous and salt deposits underlie about 25% of dryland areas. According to Klimchouk et al. (1997), gypsum karst is more common in the northern hemisphere but is to be found everywhere, from the cold Arctic to the hot arid or humid tropical areas. Sulfate rocks underlie 35–40% of the USA territory, Quinlan et al. (1986). Gorbunova (1977) estimated that 5 million km2 of the former USSR territory are underlain by sulfate rocks. The total outcrop of gypsum formations in Iran is about 80,000 km2, i.e., 5% of the total area of that country, Raeisi et al. (2013). Cooper and Calow (1998) listed 80 countries as major world producers of gypsum products. Many dams and reservoirs all over the world constructed on geological formations containing evaporites have encountered significant seepage or stability problems during their construction, first filling, or later during their operation. Numerous dams in the USA have had serious dissolution problems, sometimes ending in complete failure, as well as dams in Spain, China, Russia, Algeria, Iran, Venezuela, Argentina, Germany, Guatemala, Tajikistan, and a few other countries. All these problems are consequences of the high solubility of the common evaporite rocks, which are chiefly gypsum and salt, with carnalite, sylvite, and glauberite being less common. In many cases, failures could not be prevented or avoided despite the application of massive (and costly) protective measures to try to prevent or halt dissolution. The impoundment of a reservoir can cause rapid dissolution due to the rapid, non-natural, increase of the hydraulic head, often leading to the development of solution cavities in the dam foundations and adjoining areas, collapse sinkholes in the reservoir bottom and its near environs, while slope instability is common along the reservoir banks. These processes can provoke rapid increases in the permeability of the evaporite rock mass and reductions of its mechanical strength. Once the rapid solution has started due to increased hydraulic head at a site, additional protective measures become extremely complicated technically. During the last decade of the nineteenth century, intensive construction of dams and reservoirs began in karst regions all over the world. The result was an expensive failure in many cases, particularly where the foundation rock masses contained evaporites. In many of these cases, the planned reservoir never filled up, despite extensive investigations and remedial works. Due to the presence of evaporites in the foundations, some sites were abandoned before any building, but in many other instances the dam was constructed, with disastrous consequences. One of the earliest reported problems due to the presence of evaporites was unacceptable seepage losses from the McMillan Reservoir, New Mexico, USA, immediately after its construction in 1893, Cox (1967). A dam in Estremera (Guadalajara) failed in the 1950s due to the karstification of the gypsum bedrock overlaid by alluvial deposits, Gutierrez et al. (2003). In “Gypsum-karst problems in constructing dams in the USA,” Johnson (2008) concluded that if gypsum karst features are present at a dam site or reservoir can compromise the ability of the dam to hold water in a reservoir and can even cause its collapse. “Gypsum karst in the abutments or foundation of a dam can allow water to pass through, around, or under a dam, ix

x

and solution channels can enlarge quickly, once water starts flowing through such a karst system.” These conclusions are confirmed at a number of dams and reservoirs built in evaporites around the world. For instance, the worst American civil engineering failure of the twentieth century was that of the St. Francis Dam (California, USA) that killed 432 people along the St. Francisco Canyon and Santa Clara Valley in March 1928 (Ransome 1928). According to Cooper and Calow (1998), the strong uplift that displaced much of the dam with catastrophic rapidity was attributable partly to gypsum dissolution. The Quail Creek Dam (Utah) was constructed in 1984 and failed in 1989 due to the creation and enlargement of caverns in its foundations. More recently, the Mosul Dam, a very important water control structure in Iraq, has been declared to be the most dangerous dam in the world because of the history of nearly continuous, massive grouting works needed during the 30 years or more since its construction was completed, works that have failed to eliminate continuing dissolution beneath it (Sissakian et al. 2014). In practice, every dam and/or reservoir that is constructed in geological formations containing evaporites will face at least one of the three principal problems: seepage losses, instability of dam foundations and reservoir banks, and water pollution. Different aspects of the problems with dams and reservoirs in evaporites are discussed by: Ransome 1928; Maximovich G.A. 1948; Jiménez 1949; Maslov and Naumenko 1958; Brune 1965; Liamas 1965; Gunnar 1965; Cox 1967; Calcano and Alzura 1967; Mamenko 1967; James and Lupton 1978; Pechorkin and Pechorkin 1979; Klizas and Maksimovich 1979; James and Kirkpatric 1980; Voronkievich et al. 1983; Maximovich N.G. 1986, 2006; Anagnosti 1987; Hu 1988; Ford and Williams 1989, 2007; Gorbunova et al. 1991; Guzina et al. 1991; Araoz Sánchez-Albornoz 1992; Lykoshin et al. 1992; James 1992; Lu and Cooper 1997; Cooper and Calow 1998; Pearson 1999; Dreybrodt et al. 2001; Romanov et al. 2003; Gutierrez et al. 2003, 2015; Johnson 2003, 2004, 2008; Milanović 2004, 2011, 2018; Lu and Zhang 2006; Kiyani et al. 2008; Moradi and Abbasnejad 2011; Manchebo Piqueres et al. 2011; Barjasteh 2012; Cooper and Gutiérrez 2013; Sissakian et al. 2015; Mahjoob et al. 2014; Meshkat et al. 2018; Maximovich and Meshcheriakova 2018. In addition, the authors of many other articles concerned with the nature of evaporite deposits (at local to global scales) include remarks focused on dam construction in these hazardous environments. Dam problems in evaporites are reported in a number of different scientific journals and have presented at many international conferences and congresses in Russia, China, USA, Turkey, Iran, etc., organized at the international level by the International Association of Hydrogeologists (IAH) and the Commission Internationale des Grandes Barrages (ICOLD), or by national scientific organizations such as the Molotov (Perm) Karst Conference in Russia (1946), First world congress on Public Works Constructed Over Gypsum, Madrid, 1962, the influential series of fourteen quadrennial Multidisciplinary Conferences on Sinkholes (USA); the conference on Engineering Geological Problems of Construction on Soluble Rocks (Istanbul, 1981). Six karst conferences that include an emphasis on engineering problems have been organized in Turkey (1977, 1979, 1985, 1990, 1995 and 2000), two in Iran (1983 and 1998), and a few in China (1988, 2001, 2006, 2013). A theme section on engineering problems in Evaporite Karst was held in Denver (2002) as part of annual meeting of the Geological Society of America. Over the past two decades, similar conferences were organized in Belgrade (Serbia, 2005), Malaga (Spain, periodically from 2010), Perm (Russia, 2004, 2015), Dzershinsk (Russia, 2007, 2012), Besançon (France, 2011), Ufa (Russia, 2012), DIKTAS Conference—Karst without Borders (Trebinje, Bosnia and Herzegovina, 2016) and Neuchâtel (Suisse, Euro karst 2016). International journal Carbonates and Evaporites, established in 1979, provides a specialized forum for the exchange vide spectrum of experience including engineering issues in karstified rocks. Interest in the geological properties of evaporites has increased because of the many geotechnical problems arising from the high solubility. The most frequent and most common problems are subsidence caused by groundwater abstraction, and leakage beneath the dam due to solution channels forming in these rocks. The technical and financial losses including losses

Introduction

Introduction

xi

of human lives have caused increasing concern with building large structures in these environments. A few scientific projects have been launched to better understand evaporites and to develop effective methods of prevention and remediation to minimize the geohazards. “Avoiding gypsum geohazards: Guidance for planning and construction” was funded by the UK Department for International Development, 1998. “Research on the Development Mechanism of Sulphate Rock and Impacts of its Environmental Evolution” was supported by the National Natural Science Foundation of China, 1999–2001.

References Anagnosti, P. 1987. Prediction and control of seepage in soluble grounds. IX European Conference, International Society of Soil Mechanics and Foundation Engineering, Dublin. Araoz, A. 1992. Cimentacion de presas en terrenos terciaros con disolucion de evaporitas y erosion interna en la Cuenca del Erbo. Rev. Obras Publicas 1991–1: 265–290. Barjasteh, A. 2012. Salt tectonics impact on dam construction in Khuzestan Province of Iran. International Symposium on Dams for a changing world. Kyoto, Japan. Brune, G. 1965. Anhidrite and gypsum problems in engineering geology. Engineering Geology (Sacramento). 2: 26–33. Calcano, C.E., and P.R. Alzura. 1967. Problems of dissolution of gypsum in somedam sites. Bulletin of the Venezuelan Society of Soil Mechanocs and Foundation Engineering. 75–80. Cooper, A.H. and F. Gutiérrez. 2013. Dealing with gypsum karst problems: hazards, environmental issues and planning. In Treatise on Geomorphology. Karst Geomorphology, vol. 6, ed. Frumkin, A., 451–462. Amsterdam: Elsevier. Cooper, A.H. and R.C. Calow. 1998. Avoiding gypsum geohazards: Guidance for planning and construction. British geological survey. Technical report WC/98/5 overseas geological series. Cox, E.R. 1967. Geology and hydrology between Lake McMillan and Carlsbad Springs, Eddy County, New Mexico. Geological Survey Water-Supply Paper 1898. Washington. Dreybrodt, W., D. Romanov and F. Gabrovšek. 2001. Karstification below dam sites: A model of increasing leakage from reservoirs. 8th Multidisciplinary Conference on Sinkholes and the Engineering and Environmental Impacts of Karst. Louisville, Kentucky, USA. Ford, D. and P. Williams. 1989. Karst geomorphology and hydrogeology. London: Unwin Hyman. Ford, D. and P. Williams. 2007. Karst hydrogeology and geomorphology. Chichester, England: John Wiley & Sons. Gorbunova, K.A. 1977. Karst in gypsum of the USSR, 83. Perm, Russian: Perm University. Gorbunova, K.A., N.G. Maksimovich, V.P. Kostarev, and V.N. Andreichuk. 1991. Tectogenic impact on the karst in Perm Region, Newsletter, UNESCO, IGCP Project 299, 85. China: Guilin. Gunnar, B. 1965. Anhydrite and gypsum problems in engineering geology. American Geological Institute, vol. 7. Gutierrez, F., M. Desir and M. Gutierrez. 2003. Causes of the catastrophic failure of an earth dam built on gypsiferous alluvium and dispersive clays (Altorricon, Huesca Province, NE Spain). Environmental Geology 43: 842–851. Gutierrez, F., M. Mozafari, D. Carbonel, R. Gomez and E. Raeisi. 2015. Leakage problems built on evaporates. The case of La Loteta Dam (NE Spain), a reservoir in a large karstic depression generated by interstratal salt dissolution. Engineering Geology 185: 139–154. Guzina, B.J., M. Sarić and N. Petrović. 1991. Seepage and Dissolution at foundations of dam during the first impounding of the reservoir. In Commission Internationale Des Grandes Barages, Vienne, vol. 1459. Hu, Wuzhou. 1988. A study on the formation of Triassic “gypsum-dissolved strata” in Guizhou Province and the seepage prevention for reservoirs. In Karst Hydrogeology and Karst Environmental Protection; Proceedings 21st IAH Conference, vol. 2, ed. D. Yuan. Guilin, China: IAHS-AISH Publication. James, A.N. 1992. Soluble materials in civil engineering. Ellis Horwood Series in Civil Engineering, 434. Chichester, England: Ellis Horwood. James, A.N. and A.R.R. Lupton. 1978. Gypsum and anhydrite in foundations of hydraulic structures. Geotechnique 3: 249–272. James, A.N. and I.M. Kirkpatric. 1980. Design of foundations of dams controlling soluble rock and soil. Quaternary Journal of Engineering Geology and Hydrogeology 13: 189–198. Jiménez, B. 1949. El canal y la presa de Estremara. Revista de Obras Públicas 2812, 357–369. Johnson, K.S. 2003. Gypsum karst and abandonment of the Upper Mangum Damsite in Southwest Oklahoma. In Evaporite Karst Karst and Engineering/Environmental Problems in the United States: Oklahoma Geological Survey Circular, vol. 109, ed. Johnson, K.S. and J.T. Neal, 85–94. Johnson, K.S. 2004. Problems of dam construction in areas of gypsum karst. Karstology—XXI century: Theoretical and practical significance. In Proceedings of the International Symposium. Perm, Russia. Johnson, K.S. 2008. Gypsum-karst problems in constructing dams in the USA. Environmental Geology 53: 945–950.

xii Kiyani, M., J. Sadrekarimi and B. Fakhri. 2008. Gypsum dissolution effects on the performance of a large dam. Lj. Trans. 21: 143–150. Klimchouk, A., D. Lowe, A.H. Cooper, and U. Sauro ed. 1997. Gypsumkarst of the world. International Journal of Speleology, 5 (1–2), 1996 Thematic Issue. Klizas, P.Y., and N.G. Maksimovich. 1979. Electrical modelling of filtration related tochemically grouted gypsiferous rocks at Kama Dam foundation (in Russian). Perm: Scientific workshop Liamas, M.R. 1965. Los terrenos yesiferos como element de function de pressa (1a parte). Servocio Geológico. Informaciones y studios, boletin No 21, Madrid, 22–55. Lu, Y., and H.A. Cooper. 1997. In Gypsum karst geohazards in China. The Engineering Geology and Hydrogeology of Karst Terraines, ed. Back and Stephenson, Rotterdam: Balkema. Lu, Y., and F. Zhang. 2006. Sulphate rock karst and sulphate-karbonate rocks’compound karst. Beijing, China: High Education Press. Lykoshin, A.G., L.A. Molokov and I.A. Parabutchev. 1992. Karst and Dam Engineering. Moskva, Russian: Gidroproekt. Mahjoob, D.F., A. Sadatifard, H. Hassani and A. Zia. 2014. Upper Gotvand Dam and Hydro Power Plant dealing with salinity in reservoir, challenges, remedies and evaluations. International Symposium on Dams in a Global Environmental Challenges. Bali, Indonesia. Maslov, N.N., and V.G. Naumenko. 1958. Stability conditions of hydro-structures foundation at rocks containing gypsum. Dissolution and disintegration of rocks. (in Russian)—M: Gosstroyizdat: 71–81. Mamenko, G.K. 1967. Kama dam on Kama River. Geology and Dams. Energy 5: 9–39. (In Russian) Manchebo Piqueras, J.A., E. Sanchez Perez, I. Menendez-Pidal. 2011. Water seepage beneath dams on soluble evaporate deposits: a laboratory and field study (Caspe dam, Spain). Bulletin of Engineering Geology and the Environment (2012). doi: 10.1007/s 10064-0379-2. (Springer-Verlag) Maximovich, G.A. 1948. Targets of Karst conference—Karstology. In Proceedings of Perm Karst Conference, 5–16. Perm, Russian: Perm University. Maximovich, G.A. 1962. Karst of gypsum and anhydrite of the globe (Geotectonic relation, distribution and major peculiarities, 108–113. Moskva, Russian: Obshchiye voprosi karstovedenyiua. Maximovich, N. and O. Meshcheriakova. 2018. The experience of the dam construction on gypsum bearing rocks (in territory of the former USSR). In Proceedings of International Symposium, KARST 2018, Expect the Unexpected, 89–96, ed. S. Milanović and Z. Stevanović. Belgrade, Serbia: Centre for Karst Hydrogeology. Maximovich, N.G. 1986. Investigation of interaction of gypsum and potassic solution as consequence of chemical grouting beneath Kama Dam foundation. In Regional Geology SSSR. MGU, No. 7, 76–83. Russia. Maximovich, N.G. 2006. Safety of dams on soluble rock (The Kama hydroelectric power station as an example), 212. Perm, Russia: Publisher “Garmonia”. Meshkat, T., D. Mahjoob Farshchi and E. Ebtekar. 2018. Evaluation of evaporate karstic challenge in Gotvand dam reservoir. In Proceedings of International Symposium, KARST 2018, Expect the Unexpected, 89–96, ed. S. Milanović and Z. Stevanović. Belgrade, Serbia: Centre for Karst Hydrogeology. Milanović, P. 2004. Water resources engineering in Karst. Boca Raton, USA: CRC Press. Milanović, P. 2011. Dams and reservoirs in karst. In Karst Management, ed. van Beynen, P.E., 47–73. Dordrecht: Springer. Milanović, P. 2018. Engineering Karstology of dams and reservoirs, 354. Boca Raton, USA: CRC Press. Moradi, G and A.R. Abbasnejad. 2011. The investigation of extra seepage of the Gheisaragh Dam and remedial method. In 6th International Congress on Civil Engineering, 1–6. Semnan, Iran. Pearson, R. 1999. Geology and Safety of Dams Case Histories in Gypsum Karst for Horsehooth Dam and Reservoir and Carter Lake Dam No. 2, Colorado Big Thomson Project, Ft. Collins and Loveland, Colorado. U.S. Department of the Interior. Bureau of Reclamation. USBR Technical Service Center Center D-8321, Denver. Pechorkin, I.A. and A.I. Pechorkin. 1979. Theoretical aspects and engineering-geological forecast on the shores of karst reservoirs. Bulletin International Association of Engineering Geologists 20. Quinlan, J.F., R.A. Smith and K.S. Johnson. 1986. Gypsum karst and salt karst of the United States of America. In Atti symposio international sul carsismo nelle evaporiti. Le Grote d-Italia 4 (13): 73–92. Raeisi, E., M. Zare and J.A. Aghdam. 2013. Hydrogeology of gypsum formations in Iran. Journal of Cave and Karst Studies. Iran 75: 68–80. Ransome, F.L. 1928. Geology of the St. Francis Dam-site. Economic Geology 23: 553–563. Romanov, D., F. Gabrovšek and W. Dreybrodt. 2003. Dam sites in soluble rocks: A model of increasing leakage by dissolution widening of fractures beneath a dam. Engineering Geology 70: 17–35. (Elsevier) Sissakian, V.K., N. Al-Ansari, and S. Knutson. 2014. Karstification problems in Mosul Dam and assessment. North Iraq, Engineering 6 (2): 84–92. http://www.scirp.org/journal/eng. Sissakian, V., N. Al-Ansari, I.E. Issa, N. Adamo and S. Knutsson. 2015. Mystery of Mosul Dam the most dangerous dam in the world (five volumes). Journal of Earth Sciences and Geotechnical Engineering 5 (3). Voronkevich, S.D., L.A. Evdokimova, S.N. Emelyanov, N.G. Maksimovich, and V.I. Sergeev. 1983. Construction of grout curtains of high permeability in gypsiferous rocks of Kama dam foundation (in Russian). Moscow: Construction in karstfied regions.

Introduction

Contents

1 Distribution of Evaporite Karst in the World . 1.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Distribution of Gypsum Karst . . . . . . . . . . . 1.3 Distribution of Saline and Gypsiferous Soils 1.4 Distribution of Karstified Salt . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

1 1 1 4 5 7

2 Properties of Evaporites . . . . . . . . . . . . . 2.1 Introduction . . . . . . . . . . . . . . . . . . . 2.2 Properties of Gypsum and Anhydrite 2.3 Properties of Gypsiferous Soils . . . . . 2.4 Properties of Halite (Salt) . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

9 9 9 13 15 19

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

3 Karstification of Evaporites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Karstification (Dissolution Processes) in Gypsum and Anhydrite . . . . . . . 3.2 Karstification (Dissolution Process) in Salt . . . . . . . . . . . . . . . . . . . . . . 3.3 Computer Modeling of Seepage Increases Due to Karstification Beneath Dams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

... ... ...

21 21 27

... ...

28 31

4 Characterization of Surface and Underground Karst Features 4.1 Discussion of Some Terms . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Surface Karst Landforms in Gypsum . . . . . . . . . . . . . . . . . . 4.3 Surface Karst Landforms in Salt . . . . . . . . . . . . . . . . . . . . . 4.4 Underground Forms in Evaporites . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

35 35 37 41 43 51

5 Geohazards Associated with Dams and Reservoirs . 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Seepage from the Reservoir . . . . . . . . . . . . . . . 5.3 Dam Foundation Instability . . . . . . . . . . . . . . . 5.4 Induced Collapses (Subsidence) . . . . . . . . . . . . 5.5 Slope Stability . . . . . . . . . . . . . . . . . . . . . . . . . 5.6 Water Pollution . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

53 53 53 54 56 58 60 62

......

65

...... ...... ......

65 65 69

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

6 Prevention and Remediation at Dam Sites and Reservoirs . . . . . . . . . . 6.1 Prevention Against Dissolution and Seepage in Evaporites—General Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Waterproofing Methods Based on Grouting Technology . . . . 6.1.2 Different Waterproofing Structures . . . . . . . . . . . . . . . . . . . .

xiii

xiv

Contents

6.2 Characteristics of Post-grouting Processes and Their Consequences . . . . . 6.2.1 Investigating the Changes of Properties in the Materials Used . . . . 6.2.2 Diffusion and Deposition of Components of the Liquid Phase of the Gel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 Physicochemical Processes Leading to the Transformation of the Gel Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4 Growth of Crystals in Gels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.5 Effects of Post-injection Processes on the Solubility of Gypsum . . 6.2.6 Mathematical Modeling of Deterioration of Grout Curtains . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.. ..

73 75

..

85

. . . . .

. . . . .

85 85 86 89 91

7 Methods of Investigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 General Setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Objectives of the Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 A General Investigation Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Common Methods of Investigating the Relevant Properties of Evaporites 7.5 Hydrogeological Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 Inventory and Investigation of the Surface and Underground Features . . . 7.7 Geophysical Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.8 Hydrochemical (Dissolution) Investigations . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. 93 . 93 . 93 . 94 . 94 . 96 . 96 . 97 . 105 . 106

8 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1 General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Groundwater Regime . . . . . . . . . . . . . . . . . . . . . . 8.3 Monitoring of Spring Discharges . . . . . . . . . . . . . 8.4 Monitoring of Surface Flows and Reservoir Floors 8.5 Hydrochemical Monitoring . . . . . . . . . . . . . . . . . . 8.6 Surface Deformation and Progressive Erosion . . . . 8.7 Monitoring/Maintenance of Galleries . . . . . . . . . . . 8.8 Data Analysis and Safeguard Plans . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

9 Overview of Dams and Reservoirs in Evaporites . . . . . . . . . . . . . . . . . . . . 9.1 Summary of Dams and Reservoirs in Evaporites That Have Construction Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Selected Case Studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 Dam Sites and Reservoirs in USA . . . . . . . . . . . . . . . . . . . . . . . 9.2.2 Dams and Reservoirs in Spain . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.3 Dams and Reservoirs in Russia . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.4 Dams and Reservoirs in Tajikistan . . . . . . . . . . . . . . . . . . . . . . . 9.2.5 Dams and Reservoirs in Iran . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.6 Dams and Reservoirs in Iraq . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.7 Some Other Dams in Evaporites . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

109 109 110 110 111 111 112 112 113 114

. . . 115 . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

115 116 116 120 124 129 132 141 146 154

About the Authors

Petar Milanović Belgrade, Serbia. Professor Emeritus. Member of Karst Commission of International Association of Hydrogeologists since 1973 and UNDP Consultant. President of the Serbian IAH Chapter and member of the Governing Board of the International Karst Institute, Guilin, China. Along the more than 55 years, he participated in hydrogeological and geotechnical investigations, design and constructions, and as a member of international boards at more than 120 different projects (dams, reservoirs, tunnels, tailings, water tapping structures, and environmental projects), in Europe, Near East, Asia, Africa, and South America. As Visiting Professor and Key Speaker, he delivered lectures at number of scientific conferences and universities across the world. He is the author of more than 100 professional and scientific papers as only author or first co-author and author of four books: “Karst Hydrogeology” (1981, WRP, Colorado); “Water Resources Engineering in Karst” (2004, CRC Press); “Karst of Eastern Herzegovina and Dubrovnik Littoral” (2006 ASOS, Belgrade) and “Engineering Karstology of Dams and Reservoir” (2018, CRC Press). He is the author of chapters in books: Hydrogeology of Dinaric Karst (1984, IAH), Karst Management (2011, Springer), Karst Aquifers-Characterization and Engineering (2015, Springer). He contributed in Encyclopedia of Caves and Karst Sciences (2003, Fitzroy Dearborn) and Encyclopedia of Sustainability Science and Technology (Springer). In 2017, he was awarded the President’s Award IAH. Nikolay Maksimovich Perm, Russia. He graduated from Moscow State University (1978), Ph.D. (1984). Deputy Director of the Institute of Natural Sciences of Perm State University. He was honored Ecologist of the Russian Federation, expert of the Russian Academy of Sciences. He is the author of more than 530 scientific papers, patents, monographs such as “Karst and Caves of the Perm Region” (1992), “Kungur Ice Cave: The Experience of Regime Observations” (2005), “Safety of Dams on Soluble Rock (The Kama hydroelectric power station as an example)” (2006), “Small Reservoirs: Ecology and Safety” (2012), “Orda Cave: Сognition” (2011, 2017). He is the author of chapters in books “Hypogene Karst Regions and Caves of the World” (2017, Springer), “Assessment, Restoration and Reclamation of Mining Influenced Soils” (2017, Elsevier). Since 2008, he has been the editor in chief of collection of scientific publications “The Caves” which was published since 1947. Participant of major projects is related to solving geological and environmental problems of various regions of Russia, Kazakhstan, Tajikistan, and Uzbekistan. He is a member of International Association of Engineering Geology and the Environment, International Geochemical Society, Institute of Karstology and Speleology, Russian Geographical Society, and Associate Member of International Show Caves Association. In the mid-1990s, he participated in the organization of the Italian-Russian Institute for Environmental Research and Education.

xv

xvi

Olga Meshcheriakova Perm, Russia. She graduated from Perm State University (2007), Ph.D. (2014). Senior Researcher, Deputy Dean. She is the author of more than 85 scientific papers. Her research interests are ecological geology, karstology, and hydrogeology. The main research topic is the oil pollution of the karst territories. She is a member of International Association of Engineering Geology and the Environment, International Geochemical Society, Institute of Karstology and Speleology, Russian Geographical Society, and Associate Member of International Show Caves Association. Participant of International symposia, conferences on geotechnical challenges, karst and hydrogeology, environmental pollution and environmental management.

About the Authors

1

Distribution of Evaporite Karst in the World

1.1

General

Evaporite karst is developed widely throughout the world, though it is more common in the northern hemisphere, reflecting the current distribution of evaporite formations. It develops in all climatic settings, from cold Arctic to hot arid or humid tropical, from the lowest land to high mountains. The global distribution of gypsum, anhydrite, and salt known to have accumulated over geological time is presented at Fig. 1.1 (Kozary et al. 1968). According to Ford and Williams (2007), most of the evaporites accumulated over geological time are now buried beneath later carbonate or clastic rocks. At many cases, evaporites are removed by dissolution or reduced by folding and thrusting as it is in the Andes. More than 90% of anhydrite/gypsum and more than 99% of the salt displayed here do not crop out. Hydrogeologically active karst within these evaporite rocks probably covers an area comparable to active carbonate karsts. The common belief that arid environments are preferred for gypsum karst development is not strictly correct. Although gypsiferous formations do suffer intense karstification when exposed at the surface in very humid regions, and thus may be quickly destroyed there, the development of inter- and intra-stratal gypsum karst can be widespread and vigorous in such conditions. Gypsum karstification is common in deep-seated geological settings also, i.e., those with little or no visible expression at the surface. When not only the geomorphological, but also the geological and hydrogeological evidences of karstification in gypsum are taken into account, appreciation of the extent of gypsum karst terrains throughout the world is considerably increased (Klimchouk et al. 1996). Evaporite rocks are widespread all over the world and underlay about 25% of the global continental surface. Klimchouk et al. (1996) listed more than 35 countries and regions with recorded outcrops of gypsum. Halide rocks are widespread also. In areas of the former USSR, salt deposits cover more than 2.3 million km2 (Korotkevich 1970). The

substantial number of dams and reservoirs that are located in evaporites has given rise to many problems. Due to the high solubility of these rocks, increasing seepage losses are a common consequence of reservoir filling. In the worst cases, due to the load-bearing capacity of the foundation rocks being weakened by solution, there can be catastrophic collapse of parts of the dam, etc. structure. Another important problem is water pollution where the addition of dissolved solids from evaporites renders the water unusable for human, agricultural or even, in some cases, industrial use, particularly where reservoir water is in direct contact with the salt rocks.

1.2

Distribution of Gypsum Karst

On the global scale, surface outcrops of gypsiferous strata appear quite limited. This apparent scarcity can be explained by the relatively low resistance of gypsum (a soft rock) to all types of denudation processes, rather than being a limited occurrence of sulfate rock deposits. The extent of sulfate rocks either at the surface or at depth beneath it is great: Ford and Williams (1989) estimated that gypsum/anhydrite and/or salt deposits underlie 25% of the continental surface (approximately 60 million km2), while Maximovich (1962) calculated that the area of the continents underlain by gypsum/anhydrite alone is about 7 million km2. Karst processes operate extensively in intra-stratal settings beneath various types of cover beds, particularly where the gypsum/anhydrite beds occur within at least the upper few hundred meters of a stratigraphic sequence. Taking this into account, gypsum karst appears to be a much more widely developed phenomenon than commonly believed. The largest areas of sulfate rocks are found in the northern hemisphere, particularly in the USA and Canada, Russia, China, Iran, Spain and the UK. In USA, gypsum deposits are present in 32 states, and they underlay about 35–40% of the regions where Precambrian through Quaternary strata are in outcrop (Dean and Johnson 1989, Fig. 1.2).

© Springer Nature Switzerland AG 2019 P. Milanović et al., Dams and Reservoirs in Evaporites, Advances in Karst Science, https://doi.org/10.1007/978-3-030-18521-3_1

1

2

1 Distribution of Evaporite Karst in the World

Fig. 1.1 Global distribution of evaporite karst, Kozary et al. (1968)

In Russia and the surrounding former USSR states, Gorbunova (1977) estimated the extent of sulfate rocks to be about 5 million km2. According to Lu and Cooper (1996), China has the largest gypsum resources in the world. They range in age from Precambrian to Quaternary, and their genesis includes

marine, lacustrine, thermal (volcanic and metasomatic), metamorphic, and secondary origin (Fig. 1.3). In Iran, there are large areas of rock formations containing evaporites, mostly gypsum and salt. Figure 1.4 shows the distribution of the Gachsaran, Upper Red, and Sachun formations, the chief gypsum/anhydrite hosts. The total

1.2 Distribution of Gypsum Karst

Fig. 1.2 Gypsum/anhydrite deposits in the coterminous USA (Dean and Johnson 1989) Fig. 1.3 Map showing the age and distribution of the main genetic types of gypsum in China. (1) Cambrian marine gypsum; (2) Ordovician marine gypsum; (3) Triassic marine gypsum; (4) Carboniferous marine gypsum; (5) Cretaceous lacustrine gypsum; (6) Tertiary lacustrine gypsum; (7) Late Tertiary– Quaternary lacustrine gypsum; (8) thermal and metamorphic gypsum (typical localities); (9) secondary deposits of gypsum produced by karstification. Abbreviations are for names of provinces (Lu and Cooper 1996)

3

4

1 Distribution of Evaporite Karst in the World

Fig. 1.4 Areas containing outcrops (black) of the principal gypsum formations in Iran. The Hith Anhydrite area is too small to show at this scale Raeisi et al. (2013)

outcrop of the gypsum formations is about 80,000 km2, 5% of Iran’s land area (Raeisi et al. 2013). According to Mancebo Piqueras et al. (2011), Spain has the highest proportion of gypsiferous rock outcrops in the world. Broadly, gypsiferous geological formations occupy 21% of the total surface area of the nation, with some 9% being massive evaporite formations. Many other countries within the American continents, Europe, and Asia host important, and commonly quite extensive, gypsum karst. For instance, karstified evaporites cover large area at cold regions of Wood Buffalo National Park in northern Alberta, Canada (Fig. 1.5). It is very like the Pinega karst near Arkhangelsk, Russia, in many respects. Detailed characteristics of many of these are provided within the national reviews comprising later chapters of this volume. The aim of this chapter is to present a brief overview of the geographical distribution of gypsum karst in the world, with particular reference to those areas that are not described separately, either due to a real scarcity of data or because the authors were unable to involve local experts. The general order of the reviews begins in the Americas and proceeds toward the east. The first brief global reviews specifically dealing with gypsum karst were provided by Maximovich (1962). Since

then knowledge of gypsum karst, in terms of its morphological and hydrogeological peculiarities, development mechanisms and geographical distribution, has increased dramatically. Recently, the global distribution of gypsum karst has been considered by Nicod (1976, 1993), Cooper and Calow (1998), Klimchouk et al. (1996), and many other authors.

1.3

Distribution of Saline and Gypsiferous Soils

The processes of evaporite dissolution that must be taken into account in hydrotechnical construction can also occur in soils containing salts and/or sulfates dispersed within them. Soils containing salts are widespread in many countries. The origin of salts in soils (unconsolidated rocks) is associated with the dissolution and chemical weathering of consolidated rocks, which cause the conversion of some of their constituent minerals to saline solutions. Saline accumulation in soils is typical in semiarid and arid regions with negative soil water balances, where the annual amount of precipitation often does not exceed 25 mm/year, Petrukhin (1993), Ashoor (2003).

1.3 Distribution of Saline and Gypsiferous Soils

5

Fig. 1.5 Wood Buffalo National Park, Alberta, Canada. The bedrock geology is 5–30 m of horizontally bedded dolomite underlain by 100 m+ of gypsum with silt, shale, dolostone interbeds, etc. Courtesy of D. Ford

According to their degree of solubility in water, the salts present in soils may be subdivided into readily, medium, and weakly soluble categories. The light- and medium-soluble salts are generally referred to as “water-soluble salts.” Saline and gypsiferous soils occupy large areas in Asia (Mongolia, China, Iran, Afghanistan, Iraq, Syria, Pakistan, India), in North Africa (Libya, Egypt, Algeria, especially in the Nile Delta), the Americas (USA, Canada, Mexico, Argentina, Chile, Peru), in Europe (France, Spain, Italy, Romania, Greece), and in Australia. Soils containing salts are widely distributed in the CIS countries—in the southern part of Russia and Ukraine, in Kazakhstan, Central Asia, and Caucasus (CIS—Commonwealth of Independent States members of former USSR). About 10% of the CIS is occupied by deserts and semideserts, in which soils with

significant salt content are predominant (Petrukhin 1993). According to Chokhonelidze (1957), the total area of saline and gypsiferous soils is about 750,000 km2. Nafie (1989) estimated that gypsiferous soils occupy 724,000 km2 of the continental surface. These soils are also widespread in Iraq where, according to various estimates, they cover 20% (Barzanji 1986) to 50% (Al-Mukhtar 1982) of the nation.

1.4

Distribution of Karstified Salt

About one half of the major sedimentary basins of the world (>110) contain salt (halide) strata. Saline deposits are widespread on the planet. They are found within all

6

Fig. 1.6 Salt and salt-dome basins around the world Belenitskaya (2013). (1) Folded cover strata areas with discrete residual and injected tectonic saline features; (2) basement terraces on ancient platforms; (3) continental rifts: a—Neogeodynamic, б—Pre-Cenozoic buried; (4) oceanic rifts; (5) limits of salt basins of different material– geochemical types; mixed types are shown by a combination of signs: a —chloride–sodium, б—chloride–potassium, в—sulfate–potassium, г— sulfate–sodium, д—carbonate–sodium; (6) areas of salt tectonic manifestations; (7) age of salts (field fill color corresponds to the stratigraphic age of salts; in the presence of thick salts of two or three ages in the sequence, striped shading is used); (8) known salt and sulfate occurrences in rocks of Precambrian age

continents and seas, and the fringes of the oceans, being absent only in the modern abyssal depths, Korotkievich (1970) and Belenitskaya (2013, 2016, 2017). Only those salt bodies (or parts of them) that, after their formation, lie at considerable depths and are insulated by

1 Distribution of Evaporite Karst in the World

overlying impermeable and insoluble strata may not be affected by solution by circulating groundwater. However, the probability that even such bodies will remain completely intact, especially in their peripheral parts, is small. In this regard, we may presume that the area displaying some salt karst features will, in general, be almost as large as the area of the original ancient salt deposit. An important feature of many saline basins is the complication introduced by salt-dome tectonics, producing saline domes. Locally intensive structural deformation is characteristic in more than half of the major saline basins of the world, including the vast majority of the largest of them (Fig. 1.6). In the territory of the former USSR, the total area of salt deposits exceeds 2.3 million km2, of which the Angara-Lena marginal trough (more than 0.6 million km2) contains the largest area of rock salt deposits, plus the Pre-Caspian syncline and the Pre-Ural marginal trough (0.3 million km2), etc (Fig. 1.7), Korotkevich (1970). It is worth noting that the influence of climatic factors on karst formation in rock salt that lies above the local base level of erosion is very large, and below it—very little or none. When a rock salt massif is located above the local base level, the rate of karstic processes is determined chiefly by the amount and nature of the precipitation available, i.e., by zonal geographical factors. This explains the almost complete absence of salt outcrops on the surface in the humid climatic zones today because they are so rapidly destroyed there and the presence of considerable amounts of surface salt in the arid zones. In a different manner, leaching of salt bodies can occur to varying extents below the surface of the earth and even beneath geologic horizons containing saltwater. Such karst development can also occur with very different intensities, regardless of the zonal geographic features of the region. The geological, tectonic, and hydrogeological features in the particular salt deposit area are the chief determinants of the possibility and intensity of any flow of aggressive groundwater coming into contact with the salt. Geological and tectonic factors determine these geographically azonal features. Groundwater is zonal (stratified). This zonality is reduced to nothing if a layer of saturated brines develops and remains practically motionless on the lowermost salt formation. In this situation, karst formation ceases regardless of the climatic zone (Korotkevich 1970).

1.4 Distribution of Karstified Salt

7

Fig. 1.7 Map of the distribution of salt deposits in the former USSR. (1) Areas with modern salt lakes; (2) areas with fossil salts Dzens-Litovsky (1962), Korotkevich (1970)

References Al-Mukhtar, A.N. 1982. Distribution of gypsiferous soils and their effects on the safety of structures. J.Al.Muhangiss. Baghdad (in Arabic). Ashoor, H.M. 2003. Accounting for the process of dissolution of gypsum contained in soils, when calculating the deformation of the foundations of building. PhD thesis. Saint-Petersburg, (In Russian). 125. Barzanji, A.F. 1986. Distribution of gypsiferous soils of Iraq. In Symposium on Gypsiferous Soils and Their Effects on the Structures and Agriculture. The institute of water and Soils researches, Ministry of Irrigation. Iraq (in Arabic). Belenitskaya, G.A. 2013. Tectonic aspects of the spatial and temporal distribution of saline basins of the world. Electronic Scientific Edition Almanac Space and Time 4(1) (in Russian). Belenitskaya, G.A. 2016. Salt tectonics on the margins of young oceans. Geotektonika 3: 26–41 (in Russian). Belenitskaya, G.A. 2017. Salt in the earth’s crust: Distribution and kinematic history. Litosfera 17(3): 5–28 (in Russian). Cooper, A.H., and R.C. Calow. 1998. Avoiding gypsum geohazards: Guidance for planning and construction. British geological survey. Technical report WC/98/5 overseas geological series. Chokhonelidze, G.I. 1957. To the question of assessing the physico-technical properties of saline soils used in hydraulic structures. Proceedings of the GruzNIIIP. Tbilisi. Issue. 18–19. (In Russian). 23–43. Dean, W.E., and K.S. Johnson. (eds). 1989. Anhydrite deposits of the United States and characteristics of anhydrite important for storage

of radioactive wastes, 132. U.S. Geological Survey Professional Paper 1794. Dzens-Litovsky, A.I. 1962. The karst of salt of USSR, hydrogeological regularities of its development and geographical areas of distribution. General questions of karst studies. M., Publishing House of the USSR Academy of Sciences (in Russian). Ford, D., and P. Williams. 1989. Karst geomorphology and hydrogeology. London: Unwin Hyman. Ford, D., and P. Williams. 2007. Karst Hydrogeology and Geomorphology. Chchester, UK: Wiley. Gorbunova, K.A. 1977. Karst in Gypsum of the USSR, 83. Perm: Perm University (in Russian). Klimchouk, A., P. Forti, and A. Cooper. 1996. Gypsum karst of the world: A brief overview. In Gypsum Karst of the World, vol. 25, ed. A. Klimchouk, D. Lowe, A. Cooper, and Sauro, U. International Journal of Speleology (Chap. II. 1), 159–181. L’Aquila: Societa Speleologica Italiana. Korotkevich, G.V. 1970. The Salt Karst, Leningrad, Nedra, 256 (in Russian). Kozary, M.T., J.C. Dunlap, and W.E. Humphrey. 1968. Incidence of saline deposits in global time. Geological Society of America Special Paper 88: 43–57. Lu, Y., and H.A. Cooper. 1996. Gypsum karst in China. International Journal of Speleology 25(34): Chapter II. 13. Mancebo Piqueras, J.A., E. Sanchez Perez, and I. Menendez-Pidal. 2011. Water seepage beneath dams on soluble evaporate deposits: a laboratory and field study (Caspe dam, Spain). Bulletin of Engineering Geology and the Environment Springer-Verlag. https://doi.org/10.1007/s10064-0379-2.

8 Maximovich, G.A. 1962. Karst of gypsum and anhydrite of the globe (Geotectonic relation, distribution and major peculiarities). In Russian—Obshchiye voprosi karstovedenyiua, Moskva, 108–113. Nicod, J. 1976. Karsts des gypses et des evaporates associees. Annales de Geographie 417: 513–554. Nicod, J. 1993. Resherches nouvells sur les karsts des gypses et des evaporates associees (Seconde partie geomorphologie, hydrologye at impact anthropique). Karstologia 21: 15–30.

1 Distribution of Evaporite Karst in the World Petrukhin, V.P. 1993. Regularities of deformation and calculation of bases composed of gypsum-clayey clayey groups, 435. Ph.D. thesis. Moscow (in Russian). Raeisi, E., M. Zare, and J.A. Aghdam. 2013. Hydrogeology of gypsum formations in Iran. Journal of Cave and Karst Studies (Iran) 75: 68–80.

2

Properties of Evaporites

2.1

Introduction

The most common evaporite rocks encountered at dam sites and reservoir areas are the calcium sulfate minerals, gypsum and anhydrite, and salt, the principal chloride (halite) mineral. In few cases, rarer sulfate minerals (thenardite, glauberite, mirabilite, epsomite) may be significant, and also sylvite and carnallite (chlorides) if rock salt is the predominant evaporite. The chemical properties of evaporites and their karstification are explained in detail in many studies so that full explanation of their chemistry and kinetics is unnecessary here. Anhydrite (CaSO4) and gypsum (CaSO4
2H2O) are the same minerals, hydrated in the latter case. Impurities are common in them, including autogenous minerals (other sulfates and chlorides), argillaceous, silty and sandy clastic grains. Based on a number of different geomorphological and hydrogeological factors, Klimchouk and Ford (2000) and Klimchouk (2004) distinguish the following common evaporite karst types: (1) Syngenetic; (2) Intra-stratal; (3) Deep-seated; (4) Subjacent; (5) Entrenched; Open and Denuded; and Mantled Evaporite Karst.

2.2

Properties of Gypsum and Anhydrite

Gypsum, CaSO4  2H2O, is a monoclinic hydrated calcium sulfate. Its specific gravity may vary between 2.3 and 2.37 g/cm3 and unconfined compressive strength between 24 and 35 MPa. Hardness on Mohs’ scale is 1.5–2. In normal meteoric water, the saturation concentration is 2.65 g/L. When heated, the water is lost but is readily regained on cooling. In hand specimens, gypsum displays a wide range of crystalline-to-granular structures. Gypsum is a bad conductor of heat. The refractive index of gypsum is weak, and the birefringence is significant, Budnikov (1933). Gypsum deposits may form in a variety of ways: hydrothermally; precipitated from oversaturated ionic

solutions in seawater, commonly by evaporation in shallow lagoonal settings; in desert regions, also by evaporation, where it may adopt the form of “desert roses”; crystallized from cold-water solutions in salt deposits; and by the hydration of anhydrite. At dam sites and in reservoir, gypsum most frequently appears in the form of: – massive gypsum beds (Fig. 2.1a), – compact layers between clay-rich or carbonate formations, – as gypsum breccias (Fig. 2.1b), – filling joints in other rocks, in laminated or crystalline forms (Fig. 2.2), – a mixture of gypsum fragments and blocks in weakly lithified rocks (Fig. 2.3), – as individual crystals or small crystal aggregates (Fig. 2.4), and – as thin-bedded sediments (Figs. 2.5 and 2.6). Sometimes, a borehole core taken from compact and dense gypsum can, by mistake, be identified as limestone. In examples from the Nargesi dam site (Iran) and Mujib dam site (Jordan), small-scale folding presumably results from the conversion of anhydrite to gypsum, as illustrated in Figs. 2.5 and 2.6. The most abundant form of gypsum encountered at dam sites can be crystalline, fibrous, or fine grained—“alabaster.” Figure 2.7 presents an image of fibrous gypsum under electronic microscope. Ten centimeters of the sample (borehole core) contain about 50 thin layers of fibrous gypsum. Gypsum is a mineral with a predominance of ionic-type bonds (simple salts) and is characterized by weak stability in water and vigorous chemical activity (Sergeev 1978). The molecular structure is layered: the individual layers of calcium and sulfate ions are separated by water molecules (Gorbunova 1977; Sergeev 1983). From the three-dimensional radiographic studies, the water molecules form double layers, one of the hydrogen atoms being on the

© Springer Nature Switzerland AG 2019 P. Milanović et al., Dams and Reservoirs in Evaporites, Advances in Karst Science, https://doi.org/10.1007/978-3-030-18521-3_2

9

10 Fig. 2.1 Some of varieties of gypsum: a Compact and dense, b Brecciated, and c In form of a desert rose. Photographs P. Milanović

Fig. 2.2 Nargesi dam site, Iran. Joints in marlstone filled with gypsum crystals. Photograph P. Milanović

Fig. 2.3 a Khersan III dam site area, Iran. Compact gypsum (upper core) and marlstone with gypsum fragments; b Nargesi dam site area, Iran. Gachsaran Formation, Red clay with gypsum laminae and scattered fragments

2 Properties of Evaporites

2.2 Properties of Gypsum and Anhydrite

11

Fig. 2.4 Shattered outcrop of regularly laminated crystalline gypsum, Kzldar site, Orenburg region, Russia. Photograph by N. G. Maksimovich

Fig. 2.5 Nargesi dam site. Folding resulting from the conversion of anhydrite. Photograph courtesy of W. Riemer

outer surface of the layer and the other on the inner surface, so the degree of their participation in the formation of hydrogen bonding differs. Gypsum crystals generally have a tabular appearance (habit) or, more rarely, columnar or prismatic and often twinned. At larger scale, deposits of gypsum range from micro- to coarse-grained. Gypsum is

characterized by large variability of structural characteristics such as microporosity, morphology, crystal orientation (Pisarchik 1958). Gypsum and anhydrite have significant viscosity. Fracturing in them is rough-to-brittle, often grainy. Deposits of gypsum and anhydrite can be blocky, bedded, irregularly

12

2 Properties of Evaporites

Fig. 2.6 Mujib Dam (Jordan). Gypsum deposits in a folded thin-bedded structure. Photograph courtesy of W. Riemer

formations; (2) sparry; (3) parallel or diverse columnar; (4) idiometric granular for two-component rocks, for example, anhydrite-dolomite and gypsum-dolomite. Secondary structures arise as a result of partial or complete recrystallization or replacement of the primary.

Fig. 2.7 Fibrous gypsum—selenite, under microscope, N. G. Maksimovich, 2006

calcareous, less often platy, nodular, and sharp-angled forms (Gorbunova 1977). The structures observed in gypsum and anhydrite are subdivided into primary and secondary categories (Pisarchik 1958). Primary structures are those created as the minerals precipitate from the initial solution: (1) micro-, fine-, small-, medium- and coarse-grained for continental gypsum

I. Crystalloblasts—due to diagenesis and epigenesist, may be subdivided: (1) by grain size—fine-, small-, medium- and coarse-grained; 2) by the shape and arrangement of grains—equal-grained (mosaic, bladed or tabular, parallel- and entangled- tegulate, fibrous), non-uniform (bladed, tegulate, fibrous, radialfibrous with sharply different sizes of individual components: porphyroblastic pseudoporphyroblastic, poikiloblastic, idiometrically granular). II. Metasomatic—due to replacement or decomposition in the presence of aqueous solutions, for example, as in hydration: (1) residual (residual-granular, tabular, pinnate, corona, porphyritic forms); (2) selfmetasomatic gypsum, formed by complete hydration of the anhydrite, creating all kinds of crystal-blast structures. III. Cataclastic—brecciated structures with traces of mechanical crushing and subsequent cementation of the grains. IV. Crystalloplastic structures with curved lamellar crystals.

2.2 Properties of Gypsum and Anhydrite

13

The textures of gypsum and anhydrite also display primary and secondary (Pisarchik 1958). Primary textures are those created by the processes of sedimentation: I. Homogeneous, characterized by the absence of aggregate groupings of components: massive, solid, etc. II. Inhomogeneous and layered, which may be classified: (1) by the ratio of the thickness of the alternating layers: uniformly laminated, non-uniformly laminated; (2) on the thickness of the layers: microlayered or paper-thin, thin- , fine- , medium- , thick-layered; (3) by the form of the surface of the layers—horizontal, wavy, lenticular, and lenticularirregular. III. Landslides (as a result of underwater landslides). Secondary textures appear during the stages of diagenesis, often displaying a sequence of changes: I. The textures that arise during the processes of diagenesis and epigenesis: (1) due to selective recrystallization in weakly circulating waters: spotty, lenticular, banded, nodular, spherulitic, coarsespherulite, porphyritic, spectacled, homogeneous, and reticulate; (2) due to bulk recrystallization: nodular, spotted, spherulitic, looped, and lenticular. II. Cataclastic, formed under the influence of tectonic factors: brecciated, plated, and microplated. III. Formed under the influences of weathering, hydration, leaching, and other processes in the weathering zone. All of the listed types of secondary textures can be found here. In the process of hydration of anhydrite, there may be textures similar to primary layered ones, e.g., in gypsum with an anhydrite cap. With weathering and leaching, there is often a loose, chalky, mossy, earthen texture. When there is preferential leaching of more soluble mineral aggregates in the rock, cavernous and cavernous textures are formed. The bulk of gypsum/anhydrite rocks are of marine lagoon origin. Initial deposition of the calcium sulfate in the saltwater basins was as gypsum, which was then dehydrated to anhydrite during diagenesis that normally results from increasing depth of burial. When erosion exposes the anhydrite to the weathering zone again, it is converted back

to gypsum by hydration, forming a sort of weathering crust (Gorbunova 1985). Gypsum karst develops more rapidly than carbonate karst. Its distribution is not limited neither by temperature (in the range from 0 °C to 40 °C) nor by the composition of natural waters. Gypsum karst is characterized by collapses, whose frequency can exceed 2 per km2 per year, and a high density of sinkholes (up to 1000 per km2) Gorbunova (1977). Anhydrite, CaSO4, is an orthorhombic sulfate that appears in granular or compact masses. Specific gravity varies between 2.63–3.0 g/cm3, solubility 2.1 g/L, and unconfined compressive strength between 40 and 123 MPa. Anhydrite is much dense and harder than gypsum (on Mohs’ scale, 3–3.5). Anhydrite is present in many salt deposits as the first evaporite mineral to be precipitated. Anhydrite is also found in marls and claystones in the form of concretions that derived by the action of sulfuric acid on local carbonate minerals. It sometimes forms hydrothermally within the alteration products associated with Cu, Zn, and Pb mineralization (Milovanović 2016).

2.3

Properties of Gypsiferous Soils

Gypsiferous soils are characterized by their high content of readily soluble salts. The dry residue of the salts is 7.8%. Soils with such a salt content are found up to depth of 7–8 m. The density of particles in a saline clay soil is reduced due to an increase in the content of readily soluble salts (Table 2.1). An important characteristic is the limit of the plasticity of these soils, because the moisture content at the yield point increases with increasing sulfate content, but decreases when the chloride content is increased (Fig. 2.8; Bakenov et al. 1988). The main cause of the formation of saline soils is the dry conditions inherent in arid regions that can retain ground waters relatively close to the surface, where they usually have high mineralization. With a small amount of precipitation (less than 100 mm/year), and due to evaporation throughout the aeration zone as a result of capillary action, there is high moisture volatility. At the maximum capillary rise height (which depends on the granulometric characteristics of the soil), the evaporating water releases its salts into the solid phase. They accumulate in the form of crystals in

Table 2.1 Dependence of the density of the skeleton of the soil on the content of salts and gypsum Salt content (%) 3

Density of soil particles (g/cm )

1–3

3–5

5–8

2.69–2.67

2.67–2.65

2.65–2.60

14

2 Properties of Evaporites

Fig. 2.8 Change in the plastic limits of clayey soils under sulfate: W—liquid limit (the moisture content at which the soil passes from plastic state to liquid state; C content of salts. (1) chloride-sulfate; (2) salinization Bankov et al. 1988

the soil pores or on the surface in the form of “gypsum cores”. One of the characteristic varieties of saline soils is saline lands, which are unstructured saline soils. Saline soils can be carbonate, sulfate, halite, nitrogenous, or mixed. Saline lands are widespread in areas with a dry climate, Ashoor (2003). Classification of gypsiferous soils. A soil containing 2% gypsum is considered to be gypsiferous, Van Alfen and Romero (1971). In 1973, Barzanji classified gypsum soils on the basis of gypsum content by mass (Table 2.2). Table 2.2 Classification of plaster soils, Barzanji (1973)

Table 2.3 Classification of gypsiferous loamy soils, Petruhin (1996)

Table 2.4 Classification of gypsiferous sand loams, Petruhin (1996)

Barzanji’s works show that when the gypsum content reaches 10%, it affects the characteristic properties of the soil, especially its permeability, plasticity, soil structure, and mechanical properties. In 1996, Petrukhin proposed a new classification, in which he subdivided gypsum soils into sandy loam and loam types, based on experimentally established features of suffosional compression in such soils (Tables 2.3 and 2.4). Frequently, gypsum inclusions in other rocks are present in forms other than the microscopic crystals characteristic of

Gypsum content (%)

Classification group

0–0.3

Not gypsiferous

0.3–3

Very slightly gypsiferous

3–10

Slightly gypsiferous

10–25

Medium gypsiferous

25–50

Strongly gypsiferous

>50

Excessively gypsiferous

Group

Gypsum content (do) (%)

Name

Suffosional deformation (έ)

1

100; 2 > 10; 3 > 1. (B) Dependance Kф = f (T) when filtering distilled water: (a) with pressure gradient is 10; 1

—mudstone—cement grout mix; 2—sandstone—cement grout mix; (b) with pressure gradient 100 for intact samples: 1—mudstone; 2— sandstone from foundation of the Rogun Dam. Voronkevich et al. (1987)

the flank of the bed, downstream water level; 3—on the flank of the bed, upstream water level. In Element 1, the vertical loses are summarized; in Elements 2 and 3—horizontal losses. The lifespan of the effective operation of the

high-density filtration screen is determined by the time during which the flow rates through the selected elements increase to values corresponding to the grouting costs. Analysis of the data obtained (Fig. 6.23a) shows that, with

References

Fig. 6.23 Results of mathematical modeling of changes in total discharges (a). Through the selected elements (b), of the protective screen around the salt reservoir at the foundation of the Rogun Dam (Q1, Q2, Q3—discharges through the corresponding screen elements at the densities achieved by injecting cement solutions), Voronkevich et al. (1987)

the steady passage of time, the losses through the selected elements change according to a more complex relationship. Losses through Elements 1 and 2 are close to each other and exceed the losses through Element 3 by about 20 times during the analyzed time period. After 10–12 years, the flow rates through Elements 1 and 2 will reach a level corresponding to the conditions before the grouting works, so determining the lifespan of the screen efficiency (Voronkevich et al. 1987). Laboratory experiments show that as a consequence of using a silica solution and its components, a thin screen of low solubility compounds (oxalates and Ca hydro silica) is created on all contacted gypsum surfaces. At the same time, in gypsum, there is increased fracture filling with these dense compounds. On the basis of the methods developed, including mathematical modeling, it is possible to estimate change of grout curtain waterproofing properties over time. A more detailed explanation of the investigation procedures is presented in Morozov (1985).

References Adamo, N., and N. Al-Ansari. 2016. Mosul Dam the full story: Engineering problems. Journal of Earth Sciences and Geotechnical Engineering 6 (3): 213–244. (Scienpress Ltd.). Adamo, N., N. Al-Ansari, E.I. Issa, V.K. Sissakian, and S. Knutson. 2015. Mystery of Mosul dam the most dangerous dam in the world: Karstification and sinkholes. Journal of Earth Sciences and Geotechnical Engineering 5 (3): 33–45. (Scienpress Ltd.). Bolotina, I.N., S.D. Voronkevich, and N.G. Maksimovich. 1986. About possibility technogenetic biogeochemical processes during silication of gypsiferous rocks. In Russian. Bulletin of Moscow University 4, Geology. 4: 49–53. Bonstedt-Kupletskaya, E.M. 1962. New Minerals. XII—Sulphides, Selenides, Tellurides, 187–206. Notes of the All-Union

91 Mineralogical Society. M.-L.: Publishing House of the Academy of Sciences of the USSR. Ch. XCII. Issue 2, (in Russian). Buchatsky, G.V., E.V. Zernov, L.A. Evdokimova, V.I. Sergeev, S.D. Voronkevich. 1976. Creation of Anti-filtration Curtains with Experimental Application of a New Chemical Oil-Well Mortar, 4– 6. Hydrotechnical construction, No. 4. (in Russian). Chuvelev, V.K. 1968. On the syneresis of a carbamide resin gel in a capillary-pore space. In Materials of the Meeting on Consolidation and Compaction of Soils, 404–405. In Russian, Moscow. Dmitrievskii, G. E., L.G. Martynova, et al. 1971. Solubility of silica acid in solutions of alkalies and sodium and potassium carbonates. Zh. Prikl. Khim. 44(ii). Efimov, A.I. and I.V. Belorukova. 1983. Properties of Inorganic Compounds, 392. L.: Chemistry, (in Russian). Eitel, V. 1962. Physical Chemistry of Silicates, 1056. Moscow: Publishing House of Foreign Literature. Gardner, G.L., and G.H. Nancollas. 1975. Kinetics of dissolution of calcium oxalate monohydrate. Journal of Physical Chemistry 79 (24): 2597–2600. Gorshkov, V.S., V.S. Timashev, and V.G. Savel'ev. 1981. Methods of Physicochemical Analysis of Binders. (in Russian) Moscow: Vysshaya Shkola. Gutierrez, F., M. Mozafari, D. Carbonel, R. Gomez, and E. Raeisi. 2015. Leakage problems built on evaporates. The case of La Loteta Dam (NE Spain), a reservoir in a large karstic depression generated by interstratal salt dissolution. Engineering Geology 185: 139–154. GuzinaB, J. 1992. Power water ingress through gypsiferrous rock into foundations during excavation for a pumped storage scheme. Tunneling and Underground Space Technology 7 (2): 141–144. Henisch, H.K. 1996. Crystal Growth in Gels, 874. Inc.; New York: Dover Publications. Lafuente, R., J. Granell, I. Poyales, V. Florez. 2006. La loteta dam. A strategic location. In Dams and Reservoirs, Societies and Environment in the 21st Century, 105–112, ed. Berga, L., Buol, J. M., Bofill, E., De Cea, J.C., Garcia-Perez, J.A., Manueco, G., Polimon, J., Soriano, A. Yagüe, J., Taylor & Francis Group, London. Maximovich, N.G. 2006. Safety of Dams on Soluble Rock (The Kama hydroelectric power station as an example), 212. Perm, Russia: Publisher “Garmonia”. Maximovich, N.G., E.A. Khayrulina. 2011. Geochemical Barriers and Environmental Protection, 248. (in Russian). Perm. Meshkat, T., D. Mahjoob Farshchi, and E. Ebtekar. 2018. Evaluation of evaporate karstic challenge in Gotvand dam reservoir. In Proceedings of International Symposium, KARST 2018, Expect the Unexpected, ed. S. Milanović and Z. Stevanović, 89–96. Centre for Karst Hydrogeology, Belgrad. Milanović, P. 2010. Gotvand Dam Project, Iran. Evaporite problem. Mission Report. Teheran. Not published. Morozov, S.V. 1985. Estimation of Filtration Changes through the Chemical Grout Curtains in the Sedimentary Rocks. PhD dissertation. Mountain Encyclopedia. 1986. Geospheres—Kenai, vol. 2. Edition: Soviet Russia, Moscow, 575. Nikolaev, A.V. and E.I. Foregina. 1944. Protective effect of films on gypsum. In Protective Films on Salts (in Russian), Izd. Akad. Nauk SSSR, Moscow, Leningrad. Nobukazu, I. 1987. Growth of Gypsum Crystals in Gel. Gypsum & Lime. 1978 (153): 75–80. Payton, C.C., and M.N. Hansen. 2003. Gypsum karst in southwestern Utah: Failure and reconstruction of Quail Creek Dike. In Evaporite karst and Engineering/Environmental Problems in the United States, ed. K.S. Johnson and J.T. Neal, Oklahoma Geological Survey Circular 109. Rakin, V.I. 1997. Processes of Crystal Formation in Gels, 109. Publishing house Geoprint: Syktyvkar. (in Russian).

92 Rakin, V.I., V.I. Katkova, B.A. Makeev. 2005. Non-equilibrium Crystallization of Calcium Oxalate in Aqueous Solutions, 5–9. Vestn. Institute of Geology of the Komi Scientific Center of the Ural Branch of the Russian Academy of Sciences, no. 11. (in Russian). Ricketts, B.D. 1980. Experimental inverstigation of carbonate precipitation in hydrated silica gel. J. Sediment. Petrol. 50 (3): 963–970. Frye, Keith (ed.). 1981. The Encyclopedia of Mineralogy, 794. Stroudsburg, Pa.: Hutchinson Ross Pub. Co. Tvehofel, U.K. 1936. The Doctrine of the Formation of Precipitation, 916. M.-L.:ONTI NKTP. Trupak, N.G. 1941. Freezing of soils in hydraulic engineering construction. In Proceedings of the Conference on the Consolidation of Soils and Rocks, 81–96. Moscow: Publishing House of the USSR Academy of Sciences. (in Russian). Trupak, N.G. 1974. Freezing of Soils in Underground Construction. Moscow: Nedra. (in Russian). Voronkevich S.D., S.N. Emelyanov, S.V. Morozov, N.G. Maksimovich. 1987. Methods for assessing the temporal change in the

6

Prevention and Remediation at Dam Sites and Reservoirs

quality of impervious to air curtains in soluble rock soils. In Energy Building, 15–18, no. 7 (in Russian). Voronkevich, S.D., L.A. Evdokimova, S.N. Emelyanov, N.G. Maximovich, and V.I. Sergeev. 1983. Construction of Grout Curtains of High Permeability in Gypsiferous Rocks of Kama Dam Foundation (in Russian). Moskwa: Construction in karstfied regions. Wittke, W. and H. Hermening, 1997. Grouting of cavernose gypsum rock underneath the foundation of the weir, locks and powerhouse at Hessigheim on the River Neckar. Proceedengs of the 19th Congress of the ICOLD, Florence, Q75, R.44, 613–626. Zhdanov, S.P., S.S. Khvoshchev, N.N. Samulevich. 1981. Synthetic Zeolites, 264. M.: Chemistry. (in Russian). Zhemchuzhnikov, Y.A. 1960. Ginzburg AI Fundamentals of Petrology of Coals, 93. Moscow: Publishing House of the USSR Academy of Sciences. (in Russian). Zverev, V.P. 1967. Hydrochemical studies of the gypsum system groundwater, 100. Moscow: Publishing house Science.

7

Methods of Investigation

7.1

General Setting

Understanding the properties of natural evaporite karst is a complex task, whether the data are needed for dam and reservoir construction, for water supply or simply for advancing scientific knowledge. The uniqueness of each karst region tends to make the routine application of existing standard investigative methods less than adequate, although in earlier karst cases they may have given satisfactory results. The major hydrogeological and geotechnical properties that define karst very often change from place to place, both quantitatively and qualitatively. Therefore, for the evaluation of a karstified rock mass, systematic, complex, and usually lengthy investigations are needed. The crucial starting point for any dam design, particularly in evaporites, is a good geological conceptual understanding that is based on comprehensive investigations at the site and in its environs. Due to the complex genetic diversity found in evaporite rocks, the investigative approaches must be specific. First of all, the role of chemistry and hydrochemistry is very important because of the high and rapid solubility of these rocks. The rapid karstification that is a consequence has generated substantial seepage or leakage and, sometimes, failures in a number of dams and reservoirs where evaporites are a component of the local geology. Karstification and other destructive processes in evaporites are much faster than in carbonate rocks. In particular, the increase of solution channel apertures and the creation of sinkholes may be extremely rapid: in the halides (salts), the frequently rapid upward diapiric movement is also a very specific geodynamic process. To deal with these kinds of hazards, comprehensive investigations conducted by a wide spectrum of scientists and engineers should be involved. In general, the basic methods regularly applied in geologic and hydrogeologic practice are the first step in what will be a complex and phased approach: geological mapping, application of remote sensing methods, geologic structural analysis, geomorphological analyses, location and identification

of all karstic features, hydrological measurements and monitoring, tracer tests, drilling of investigation boreholes, monitoring of water tables with piezometers, geodetic monitoring and geophysical methods. Due to the high solubility, chemical analyses of a wide range of water samples from boreholes, surface flow and springs are particularly important. In some cases, physical models have also been applied to analyze important characteristics of the dissolution of gypsiferous rocks in dam foundations or reservoir banks.

7.2

Objectives of the Investigations

One of the principal reasons for the destruction of hydrotechnical structures is often the failure to take timely repair and restoration measures, which are not planned due to the lack of or insufficient information on the state of the structures (the presence of zones of decomposition in the rocks, the degree of karst development, etc.). The application of standard engineering-geological methods for this purpose does not fully provide a solution to these issues due to the discrete nature of the information obtained during a survey using primarily boreholes and wells, as well as the relatively high cost of the work. Salt and gypsum dissolution makes dam and reservoir construction hazardous. The existence of evaporites in the foundations can provoke extreme stability problems, sometimes ending in failures. In the case of reservoirs, the rapid dissolution of salt and gypsum in the surface and subsurface can result in pollution of the reservoir water and development of solution channels that rapidly extend. The consequence of rapid solution channel development is leakage from reservoirs and subsidence along the reservoir banks. According to some references, under natural conditions adjacent to a river or in a cave, gypsum can easily be dissolved at a rate of 1 m a year with a water flow rate of about 1 m per second across the rockface. This is about one thousand slower than halite (salt)! Dissolution is enhanced at

© Springer Nature Switzerland AG 2019 P. Milanović et al., Dams and Reservoirs in Evaporites, Advances in Karst Science, https://doi.org/10.1007/978-3-030-18521-3_7

93

94

7

the intersections of joints sets and joints with bedding planes, where large conduits and other cavities are readily formed, creating concentrated water flow underground and collapses at the surface. The key objective of investigation is to confirm the feasibility of building a dam and reservoir when evaporites are present. Due to the solubility of evaporites, investigations are much complex than in other geological formations. After detailed investigations, some dam sites have been abandoned (e.g., Cedar Ridge and Mangum dam sites, USA), while others have had to be comprehensively re-designed (e.g., the bottom outlet at Gotvand Dam, Iran). Investigation programmes must be detailed and comprehensive. Restriction of the funds during investigations is a crucial mistake that usually leads to expensive remedial works or complete failure.

7.3

A General Investigation Approach

In reservoir areas with potential problems related to salt and gypsum dissolution, a phased investigation approach leading from field investigations, laboratory analysis and modeling, to the selection of appropriate (most effective) treatments is required. The most likely and welcome results of those investigations should be methods to forecast the impacts of salty water on reservoir water quality and to minimize these negative effects. A suggested course of action is, in sequence: – Desk study, – Assessment of air photographs, LiDAR, and thermal imagery, etc. – Detailed geological mapping on site, – Hydrological measurements, – Inventory and investigation of the surface and underground karst features, – Laboratory mineralogical/petrographical and chemical analysis, – Laboratory hydrochemical investigations and dissolution experiments, – In situ dissolution experiments, – Geophysical investigations, – Boreholes (investigation/piezometric) including geophysical logging and video endoscopy. – Investigation adits in areas potentially containing caverns and other voids. – Development of a monitoring and safeguard program before construction and during operation. – Preparation of a three-dimensional point-source hydrodynamic model for surface waters.

Methods of Investigation

– Assessment and design of protective measures to mitigate risks of dam instability, reservoir water pollution and to prevent leakage from the reservoir.

7.4

Common Methods of Investigating the Relevant Properties of Evaporites

To understand the regime of a karst aquifer is not an easy task. Historical data collection and a good geological map are the unavoidable base and first steps in the evolving investigation. Geological analyses start at the regional scale map, followed by the phased approach through pre-feasibility, feasibility and, finally, detailed mapping of the dam site and reservoir area themselves. A number of technologies and identification methods are available and helpful during field investigations. Different remote sensing methods including satellite images, aerial photos and, particularly, LiDAR images and drone photography are excellent sources for very precise identification and point location of the different types of karst features. Investigation galleries and water pressure tests (Lugeon tests) are routinely applied. Geophysical methods are a source of valuable data also. For analysis of the geomorphological and structural properties, field survey is extremely important. These techniques are widely applied in geology and need not be explained in detail here. A crucial requirement in engineering karstology, particularly in dam engineering, is the detection of any karst caverns, empty or filled, in the aeration zone or the saturated zones. At the time of writing, there is still no truly reliable method for detecting the location and form of underground caverns or karst channels from the surface, particularly at depths of more than 20 m. Due to the great genetic diversity of karst from site to site, some investigation methods have their own particular uses: e.g., radar techniques (Ground Penetrating Radar and borehole radar); video endoscopy (TV logging); geothermal methods; echo-sounding; speleological exploration (including cave diving); and different tracer techniques using different dyes, radioactive isotopes, common salt, post-activated isotopes, optical brighteners, spores, smoke, and air. Due to the rapid destruction that can occur in evaporites, substantial (and often abrupt) changes in the surface topography are frequent. Some of these changes will happen within an average human lifetime. Because of that, interviews with local people can be a very important source of historical data. One of the basic diagnostic features is the sinkhole, ranging from initial, almost un-noticeable sagging up to deep sinkholes, recently created or known as historical

7.4 Common Methods of Investigating the Relevant Properties of Evaporites

events (paleosinkholes). Registration of these forms during the investigation stage, at the dam site or reservoir area, indicates possible seepage problems and the potential vulnerability of the structure. Opening of new sinkholes during the operation of a reservoir implies a very serious problem and possible failure. Reports from existing engineering practice show that new sinkholes are frequently registered in the bottoms and banks during the operation of number of dams and reservoirs in evaporites. The following investigation procedures are recommended where evaporites are present at the site: • Detail geological mapping includes the following: – Detailed geological mapping (scale 1:5000 or 1:2000). – Detailed mapping of selected geological cores and cross sections. – Detail mapping of all tectonic features. This procedure also includes selection of rock and core samples for mineralogical/petrographical analysis. • Locations of investigation boreholes have to be selected on the basis of the geological mapping and geophysical investigations. All boreholes should be drilled by applying rotary (coring) technology and equipped with PVC piezometric pipes (at least ∅ 2 inches). The drilling technique must allow for core recovery. The cores from salt and gypsum sequences have to be protected against dissolution by PVC foil before storing in the core boxes. The collar of each borehole has to be protected against solution or other damage; a concrete block and iron cap are recommended. Boreholes are necessary to define geological characteristics of the rock mass, the contact (or transition zone) between evaporite deposits and other rocks above and/or below them and to select (calibrate) the optimal geophysical methods for further investigations. • Geophysical investigations should include geoelectrical sounding and seismic refraction and reflection. The final geophysical program should be based on preliminary geophysical investigations. There are a number of new and sophisticated geophysical methods presently available. The prime purpose of these investigations is to determine the properties of the evaporite rocks and their contacts with other rocks. More detailed discussion is presented in Sect. 7.7 Geophysical Methods. • Monitoring the groundwater fluctuation regime is the best diagnostic procedure in groundwater analysis. The following monitoring regime is suggested:

95

– Not less than once per week in the dry period of the year. – Not less than twice per week in the humid period of year. – Daily measurements in flood periods (extreme river level fluctuations). However, modern electronic technology makes it possible for the user to have continuous monitoring and data transfer available in real time, and at much less cost than only a few decades ago. • Mineralogical/petrographical and chemical analysis should include the following: – Collection of samples at the site. – Microscopic study by polarizing microscope (
1200) and professional description of the rocks. – Description of the rock samples, including their petrographical and physical/mechanical properties. – Chemical analysis in the laboratory (Ca, Mg, Fe, Mn, SO3, SO4, Cl, H2S and any other prominent cations and anions. • Laboratory tests to define dissolution rates of the salt and gypsum under different flow conditions and pressures. These tests should also analyze the decrease of dissolution capacity in any waters found with differing chlorine concentrations. Such tests can be done only where a laboratory specializing in this kind of work is available. • In situ tests are important to define the relationship between dissolution rates and water flow velocity: – in stagnant water, – in slow water flow (laminar flow), – in turbulent water flow, and – under different pressures and water temperatures. These tests should include sampling from the local river upstream and downstream from the site. The locations for the rock sampling program will be defined on the basis of preliminary geological mapping and mineralogical/petrographical analysis. Hydrological testing should be maintained for at least one full hydrological year. During this time, the gauging areas at the site have to be under permanent supervision and protection. All final experimental and monitoring programs should be decided at the site itself. The above recommendations are presented only as a general framework. • Hydrological measurements must be closely coupled with the hydrogeological investigations. Due to the distinct nature of water circulation in karst terrains, the

96

7

combination of two allied but different sciences—hydrogeology and hydrology are essential to properly understand the groundwater properties of karst aquifers. Among the number of hydrological methods, the simultaneous hydrological measurements along the river bed are most useful, particularly if section of the river bed would be covered by future storage reservoir. On the base of geological mapping along the river bed, the locations of simultaneous measurements of flow discharge (gauge stations) should be established upstream and downstream of possible seepage zone. The best hydrological time for measurements is dry season when the undergroundwater levels are at minimum. The losses measured at selected section should be equal to the swallow capacity of ponors and ponor zones. Therefore, storage properties of space with water losses will be declared as risky. Hydrological relationships and application of investigation methods in karst are elaborated in more detail in the volume “Karst Hydrogeology” (Bonacci 1987, Springer-Verlag).

7.5

Hydrogeological Methods

Hydrogeological investigations should start with desk studies and geological field mapping, including inventory of the surface and underground karst features, and with preliminary tracer tests if necessary. In engineering practice, together with a good hydrogeological map, the key sources of data for analysis of the regime of an aquifer in evaporites are same as in carbonate karst aquifers: piezometric boreholes, flow recorders (particularly spring discharge recorders) and precipitation recorders. Due to the extreme importance of these data, particular care should be paid to their quality, i.e., as much precise and continuous recording as possible. As a consequence of nature of karst, the data obtained from piezometers can be misleading. The groundwater level in a given piezometer depends on its precise spatial location, the magnitude and direction of local major flows and regional base flow. To represent exact aquifer properties, the piezometric borehole should be in direct hydraulic contact with the active conduits or to be at least close to their zones of influence. Frequently, the groundwater level measured in a piezometer is not significant, being perched much higher than the true water table because, due to the erratic distribution of karst porosity, a given borehole may be terminated in a compact, nearly impermeable, block within the local rock. In that case, hydraulic connection between piezometer

Methods of Investigation

and any conduit or zonal flow is negligible or non-existent. In such piezometers, the measured water level does not represent the flow regime and hydraulic properties along the main flow zone. Particularly questionable are water levels recorded in deep multiple piezometers where two, three, or four separate pipes are inserted into one borehole in a fully karstified formation. Each upper piezometer is isolated by impervious plug from lower ones to prevent hydraulic communication between them. In many cases, these separate pipes drain only local surrounding volumes rock and establish a bogus but seemingly stable water level in the bottom, despite being much higher than the true (natural) water table.

7.6

Inventory and Investigation of the Surface and Underground Features

Inventorying the different surface karst features is an important procedure in karst terrains. In evaporites, it is particularly important due to their very dynamic rates of development. The most common features are: karst poljes, dry valleys, sinkholes, ponors (swallow holes), springs, faults, fault zones, landslides, and systems of caves. The quantity of these features should be inventoried during the geological mapping and interviewing of local inhabitants. New remote sensing technologies, for instance, LiDAR or photographs taken by drones, are sources of high-quality information for the more inaccessible areas. The properties of the inventoried features in natural conditions are particularly important. All features should be investigated, photographed, and recorded in a standard coordinate system. The detection of underground karst features is of primary importance in dam geology. Particularly important are investigations of any caverns and lesser voids detected along the routes proposed for the grout curtains. Every accessible cave and shaft needs detailed speleological investigation. The graphical presentation should include the layout plan, longitudinal sections and selected cross sections on a scale 1:100. The plan should be integrated into the local coordinate system. Detail geological mapping of every cave, particularly if is located at or close to the dam site, is mandatory. All important geological features (dip of faults, joints and bedding planes, characteristics of cavern deposits and infillings of faults and joints) have to be mapped and graphically presented. Caves, accessible from surface, and caverns detected by investigation adits can be used as investigation laboratories to understand the evolution of the local karst aquifer and the intensity of dissolution of the

7.6 Inventory and Investigation of the Surface and Underground Features

evaporites in particular places. Speleothems in evaporites can be sources of very important and detailed paleoclimatic information. Another important task related to underground features in karst is consideration of their underground fauna. Caves and shafts can be very rich in a variety of fauna, both subaqueous and in the vadose zone. Construction of dams and reservoirs in evaporite rocks causes major changes in the surface and underground water regimes. These impacts are likely to have distinct negative effects on the fragile subterranean fauna. According to the EU Habitats Directive (May 1992) almost all cave-dwelling (terrestrial and aquatic) species, particularly the endemic, need special protection. EU Member States are required to take measures to maintain or restore natural habitats of cave species to a favorable conservation status and to introduce robust protection for those habitats and species. In karst regions where every cave and saturated (phreatic) karst channel is a habitat for a number of species, (many of them endemic) strict requirements for protection can be a source of conflicts between the need for regional economic development and the preservation of nature and, in extreme cases, can be a source of transboundary conflicts. Many caves also contain important archeological artifacts: pre-historic human and animal bones, carvings, paintings, and sometimes ruins of complete settlements. In many cases, needs for conservation of these artifacts could create serious problems. Projects can be frozen or even abandoned for legal and environmental reasons.

7.7

Geophysical Methods

Most geophysical investigations usually use combinations of two or more different methods. The combined geophysical methods permit determination of the near-surface rock properties and some indication of large tectonic features and caverns. The key requirement is the detection of any caverns, solution conduits or other karst voids and studies of their properties. Due to the limitations of early conventional geophysical methods, there were many substantial improvements in the second half of the twentieth century and other new methods are being developed now. Application of Radar technique developing quickly particularly versions known as Ground Penetration Radar and Borehole radar. Different borehole logging methods have become inescapable sources of high-quality data, particularly where caverns are intercepted. Video endoscopy (down-hole TV logging) makes direct observation of the caverns possible. Echo-sounding can be successfully applied for precise measurement of contours of both vadose (water-free) cavities and submerged (phreatic) cavities. This method is based on the propagation of ultrasonic waves and recorded echo

97

travel times. Geothermal measurements are the most reliable methods for detecting underground karst water flows, by thermal logging in surrounding boreholes because the thermal field in the rock mass is disturbed by water flow or air circulation through conduit. In the case of karst channels with a flow of several cubic meters per second or more, there is great investigation potential developing with the so-called geobomb seismic method. However, geophysical methods still have their limitation, particularly when applied only from the surface. As stated above, all methods still have a hard time trying to detect the exact spatial position and geometry of caverns at depths more than *20 m. Microgravity and Ground Penetration Radar (GPR) techniques can be effective for near-surface cavities up to depths 10–15 m. In general, the microgravity technique can detect caverns down to depths equal to the diameter of the caverns if the surface terrain is comparatively flat. It is possible to detect a spherical cavern at a depth of 5 m only if the diameter of cavern is 5 m or more. In current engineering dam geology, a number of geophysical methods have been applied to analyze geological structures and rock properties in evaporites. The basis of applied geophysics consists of the solution of the so-called inverse problem, meaning that the results of geophysical measurements will always provide numerical physical values. In soluble rocks, the most important physical characteristics are electrical properties (first of all—the resistivity), dynamic elastic properties, thermal field properties, radar wave propagation, very low-frequency wave propagation and radioactivity. According to Shuvalov (2012), use of geophysical methods can detect areas with vigorous karst activity and rock destruction, contour and assess the sizes of underground cavities or workings, estimate the depths of karst caverns and voids, and provide information on the hydrogeology of the study area. Karst rocks (limestone, dolomite, gypsum, salt) have moderate to high acoustic rigidity, are dense, non-magnetic, non-radioactive (with the exception of potassium salts) and with high resistivity, Davydov et al. (2018). Geoelectrical methods have a long and successful history in investigations of karstified rocks. Most are combined sounding, profiling, and charged-body applications. Electrical profiling, vertical electrical sounding, dipole induction profiling, natural field, micromagnetic survey, gravity prospecting, refracted waves, thermometry, and emanation survey can be used during studies of fractured zones, karstified rocks, and underground cavities. The increasing number of geophysical methods leads to increasing the time and cost of the work, however. Therefore, one or two methods should be selected to reduce duplication and cover the entire area. The methods of electrical profiling and vertical electrical sounding are economical, efficient, and

98

productive and are considered basic for the study of karst, i.e., the detection of underground cavities. The choice of a rational combination depends on many factors: the economic and geological effectiveness of the methods, their productivity and ease of field work, their quality, the methods of processing and interpretation, the graphic presentation of the results and their geological interpretation (Shuvalov 2012). From the seismic perspective, the following properties of evaporite rocks are particularly important: The halite crystal dynamic elastic properties (Crain 2003) are: P-wave = 4550 m/s; S-wave = 2630 m/s, Poisson’s ratio (m) = 0.25; Density (q) = 2.16 g/cm3; Young’s modulus Ed = 36,450 MPa. The gypsum crystal dynamic elastic properties (Dortman 1992) are: P-wave = 5200 m/s; S-wave m/s; Poisson’s ratio (m) = 0.31; Density (q) = 2.32; Young’s modulus Ed = 41,000 MPa A comprehensive overview of geophysical methods applied to the investigation of karst features in salt deposits and overlying sediments is presented by Ezersky et al. (2017). The methods are based on the large contrast between the seismic velocity of salt formations and the surrounding rocks, and between the very low resistivity of local aquifers compared to surrounding sediments. To analyze properties of salt deposits in the area of Dead Sea (Israel and Jordan), a combination of the following methods was applied: Seismic Refraction (SRFR); Multichannel analysis of surface waves (MASW); Seismic Reflection (SRFL); Magnetic Resonance Sounding (MRS); Surface Nuclear Magnetic Resonance (SNMR); Electromagnetics (TEM); Ground Penetrating Radar (GPR); Microgravity and Micromagnetic study. Each of these methods has different resolutions and penetration depths. Data produced by the application of each method become very useful when they are analyzed together: TEM and 3D-SNMR; SRFR and MASW; MASW and 3D-SNMR. Magnetic Resonance Sounding (MRS) is one of the most promising, recently developed, surface geophysical methods for groundwater investigations in karst. The method was developed for the detection of water-filled karst caverns. The possible maximum depth of application is declared to be about 100 m (Legchenko et al. 2004). Two main parameters are derived from MRS measurements (Lachassagne et al. 2005): (a) the MRS water content which is closely related to the amplitude of the MRS signal; and (b) the MRS relaxing time. MRS is an efficient tool for characterizing aquifers as well as for locating water-filled voids in the subsurface (Vouillammoz et al. 2003). However, the lateral resolution of the method may be insufficient for identifying relatively small water-saturated formations for targeting a drilling program (Ezersky et al. 2017). Based on seismic refraction (SRFR), it is possible to recognize and delineate salt structures. “However, identification of the salt layer with the seismic refraction method

7

Methods of Investigation

poses two important problems. The first is the determination of the correct velocities and the second is the generation of a good geological model,” Ezersky (2006). A salt velocity criterion of Vpmin = 2900 m/s is accepted as the statistically substantially lower limit of compressional wave velocity (Vp) for the selected part of a structure. According to the measurements in 12 different areas, Vp ranges between 2920 and 4240 m/s. On the basis of field measurements, Ezersky and Sobolevsky (2002) noted that compressional seismic velocities of 3000–4000 m/s are the upper limit for carbonate rock in the shallow subsurface, which could raise the reliable salt boundary minimum Vp criterion to 4000 m/s. However, the alteration of thin salt and clay layers can give smaller compressional velocities (Fig. 7.1). Results of mapping compact salt by seismic measurements are presented in Fig. 7.2. It is based on seismic compressional wave velocities. A map based on 20 (5 + 15) refraction lines shows the velocity distribution throughout the central part of the investigated area. The compact salt units, 11-m thick, at the depth interval between 24 and 35 m with velocities greater than 2900 m/s have been identified, Ezersky et al. (2007). Application and efficiency of different geophysical methods in evaporites, particularly in the case of salt, is summarized by Ezersky et al. (2017) in Table 7.1. Based on experience with the wide spectrum of geophysical methods applied by different authors, Ezersky et al. (2017) recommend application of 3D-SNMR for locating water-filled caverns and estimating their volume; evaluation of groundwater aggressiveness using TEM; and the seismic refraction and MASW methods for salt layer mapping and evaluation of its karstification. Recently, more and more attention has been paid to developing new techniques for the operational non-destructive monitoring of sustainability and prediction of the physical condition of hydrotechnical structures, based on the use of geophysical methods that allow assessment of the nature of any karst development. Thus, Kolesnikov et al. (2012) proposed the following set of geophysical methods: engineering seismic prospecting, GPR sounding, vertical electrical sounding, and the natural field method.

Fig. 7.1 Range of compressional velocities in unconsolidated sediments and rocks in Israel, (Ezersky and Sobolevsky 2002)

7.7 Geophysical Methods

99

Fig. 7.2 Map of the salt edge based on seismic refraction lines. Areas with velocities higher than 2900 m/s are identified as compact salt, whereas areas with a velocity lower than 2750 m/s are interpreted as loose water-saturated sediments (from Ezersky et al. 2007)

A physical-geological general model based on the analysis of a set of models of the typical hydrotechnical structures in different areas of the Perm Region was developed, which served as the basis for numerical modeling to assess the physical characteristics of the major strata and their manifestation in the various geophysical fields. The results of numerical modeling of the seismic and electric fields, based on the general physical-geological model, showed that, in practical terms, each of the major strata units was reflected in the geoelectric and time sections, allowing the confident definition of a section through a hydrotechnical structure and identifying possible potentially dangerous zones (Figs. 7.3 and 7.4; Kolesnikov et al. 2012). An example of the interpretation of vertical electrical sounding data for one of the hydrotechnical structures in the area of gypsum karst development in the town of Kungur (Perm Region) is shown in Fig. 7.5.

According to the results of quantitative interpretation, a six-layer geoelectric section was derived that was generally consistent with the typical regional model. Four main geoelectric complexes are distinguished in the geoelectric section. Experimental seismic surveys were performed using refracted and reflected wave methods with appropriate observation systems. The RadExpro Plus Total 3.7 program was used to process, interpret, and visualize the seismic data. It allows researchers to carry out the processing and interpretation of the reflected and refracted wave data within one system: reading and visualizing seismograms, amplitude correction, two-dimensional and bandpass filtering, taking into account the relief, correlation, and linking of the hodographs (vectors) of the refracted waves, determining the velocities of longitudinal and transverse waves, and building appropriate models of the underground environment.

100

7

Methods of Investigation

Table 7.1 Some of the geophysical methods applied in investigation of evaporites Method’s name

Physical properties to be measured

Capability of detection

References

Seismic refraction

Seismic velocity of longitudinal waves (Vp) in solid salt layers (2900–4500 m/s) that is considerably higher than sedimentary overburden (2100–2700 m/s)

Salt elastic properties and salt delineation

Ezersky (2006), Ezersky et al. (2010)

Multichannel analysis of surface waves (MASW)

Seismic velocity of shear (S) waves (Vs)

Salt shear velocities and modulus to estimate salt porosity and karstification

Bodet et al. (2010), Ezersky and Frumkin (2013)

P-wave seismic reflection

Study of the deep (down to 200 m) subsurface structure, detection of subhorizontal interfaces and faults

Detection of salt layer within sediments, faults etc.

Keydar et al. (2010, 2012, 2013), Ezersky and Frumkin (2013)

S-wave seismic reflection

Study of the shallow (down to 100 m) subsurface structure, detection of subhorizontal interfaces and faults

Detection of salt layer within sediments, faults, etc.

Krawczyk et al. (2015), Polom et al. (2016)

Transient Electromagnetic (TEM, TDEM)

Low resistivity of aquifer formations (1– 0.5 Xm) in contrast to the surrounding sediments (more than to 10 Xm)

Accurate location of the top of the aquifer and mapping water salinity variations

Goldman et al. (1991), Kafri and Lang (1997), Ezersky (2011), Ezersky and Frumkin (2017)

Electric Resistivity Tomography (ERT

DC resistivity in subsurface

Search anomalously high resistivity

Ezersky (2008), Al-Zoubi et al. (2007)

Magnetic Resonance Sounding (MRS) or Surface Nuclear Magnetic Resonance (SNMR) + 3D-SNMR

Water in the karst voids and water in the surrounding sediments produce nuclear magnetic resonance signal with different relaxation time (less than 150 ms in sediments and more than 600 ms in karst

Detection of water-filled voids and estimate the volume of water in karstic caves

Legchenko et al. (2006), Legchenko (2013)

MRS + TEM

Bulk resistivity, water content (W), hydraulic conductivity (K) and relaxation time (T1) allows resolve equivalency

Identification of the subsurface lithology

Legchenko et al. (2008, 2009)

MRS + MASW

K and Vs

Estimating of the salt hydraulic conductivity

Ezersky and Legchenko (2014)

Ground penetrating radar (GPR

Structure of most shallow (10–15 m) sediments located over karstified unconsolidated salt

Detection of density deficit caused by cavities moving upward (stopping)

Frumkin et al. (2009, 2011)

Microgravity

Detection of gravity anomalies against strong noise background

Detection of caverns, estimation of their depth and volume

El-Isa et al. (1995), Rybakov et al. (2001), Eppelbaum et al. (2007, 2008, 2010, 2015a, b), Ezersky et al. (2013b)

Gravimetry + radar differential interferometry

Ground subsidence caused by deficit of mass

Potentially hazardous areas because of underground cavities

Closson and Karki (2009)

Photogrammetry + satellite

Ground reflectance, diachronic analysis, spatial co-occurrence

Areas prone to sinkholes

Al-Halbouni et al. (2017)

Nano seismicity and microseismic activity

Measuring of vibrations produced by falling stones in the underground cavities

Pre-collapse identification of sinkhole activity in unconsolidated media

Abelson et al. (2013)

The results of comprehensive studies of the hydrotechnical structure in Kungur display the mutual correlation of data that can be obtained by different geophysical methods (Fig. 7.6). Anomalous zones traced in the depth interval of 1–10 m, within the pickets 7–12, 19–25, are marked by

lower values of electrical resistivity, a decrease in the values of Young’s modulus (seismic prospecting), increased attenuation of the high-frequency electromagnetic field (GPR sounding) and an increase in the absolute potential of the natural field. Such a correlation is consistent with

7.7 Geophysical Methods

Fig. 7.3 Results of numerical simulations of electrical soundings for a typical five-layer physical-geological model of a hydrotechnical structure of the Perm Region in areas where sulfate and carbonate

101

karst with weakened zones were expected (a): b section of apparent resistivity; c sections of specific electrical resistivity (from Kolesnikov et al. 2012)

102

7

Methods of Investigation

Fig. 7.4 Results of numerical modeling of a seismic wave field for a typical five-layer physical-geological model of a hydrotechnical structure of the Perm Region with weakened zones (a): b cross section of a common depth platform for refracting border No. 1; c seismogram of the common point of excitation for a section without a weakened zone; d seismogram of a common excitation point for a section with a weakened zone (from Kolesnikov et al. 2012)

well-known petrophysical and physicochemical concepts about the manifestation (in physical fields) of areas of high moisture, clay content and porosity in the rocks, contributing to an increase in the reliability of the geological findings. Most confidently, the structural features of the section are displayed by the results of the seismometry and GPR observations, the elastic properties by the velocity characteristics of the section, the physico-mechanical properties of soils obtained from seismic data and their degree of wetting and permeability by the electrometric methods. This technology has been tested intensively under the conditions prevailing in the gypsum karst areas in the Perm Territory. Based on interpretations of the seismic prospecting and electrometric studies, recommendations were made for optimal repair and restoration measures, including river bank protection works and emplacement of drainage

facilities to prevent suffosion and landslides in overlying soils and colluvium (Kolesnikov et al. 2012). Geothermal methods can become important and powerful tools in karst, particularly when there is a net of piezometric boreholes in the investigation area. These methods can be very effective where dams and reservoirs are situated in evaporites. Using the network of boreholes (most of them equipped with piezometers) geothermometry can successfully determine underground water flow channels or zones with intensive filtration, e.g., Bodvarsson (1973), Borić (1980), Drogue (1985), Bonacci (1987), Chengjie (1988), Haenel and Mongelli (1988), Ravnik and Rajver (1989, 1998), Milanović (2004). When water percolates slowly through a rock mass, the thermal equilibrium between the water temperature and surrounding rock is more or less stationary. Where there is

7.7 Geophysical Methods

103

Fig. 7.5 An example of the interpretation of electrical sensing data: a a section of apparent resistivity; b geoelectric section (from Kolesnikov et al. 2012)

concentrated flow through conduits, the thermal field becomes disturbed due to the temperature differences between the rock and water. As a consequence, an inverted thermal gradient often exists in the upper portion of a karst aquifer. An inverse gradient measured in piezometric boreholes normally indicates the presence of a karst channel with active flow. The terrestrial heat flow density (q) is obtained by taking the product of the thermal conductivity (k) and the temperature gradient (dT/dz) so that
q ¼ k dT=dz mW=m2 with depth (z) increasing downward (q > 0). Generally, boreholes display a clear and continuous thermal gradient over their depth. However, the rock around a karst channel forms a (generally) low-temperature anomaly. According to Ravnik and Rajver (1998), three basic types of geotherms can be qualitatively distinguished by the sign of their temperature gradients: – Normal type, dT/dz > 0 °C/m – Inverse type, dT/dz < 0 °C/m – Adiabatic type, dT/dz = 0 °C/m

All the three types of geotherm are diagnostic of particular structures and of special hydrothermal and hydrogeological cases. A good example of the utility of thermal measurements was the temperature logging program in boreholes drilled from an investigation gallery behind the large limestone karst Ombla Spring near Dubrovnik, Croatia. They were undertaken as part of the design of the underground Ombla Dam. The aim of the investigation was to detect the spatial location of the key karst conduit zone, at a distance of approximately 200 m behind the spring outlet. The maximum discharge capacity of the spring is up to 130 m/s. Thermal measurements were done in piezometers drilled along the investigation gallery, which was approximately perpendicular to the assumed direction of the karst channel (Fig. 7.7a, c). Due to the surrounding rock mass having a low thermal diffusivity, the responses to temperature variations in the karst channels are slow. The temperature close to the underground flow is always lower than at greater distances from it (Fig. 7.7c, C1): Ravnik and Rajver (1998). All geotherms measured in 29 boreholes along the investigation gallery had the inverse form. Some of them indicated the presence of the karst channel with an active concentrated flow in their close vicinity. Two typical examples of graphs

104

7

Methods of Investigation

Fig. 7.6 Results of comprehensive geophysical studies: a apparent electrical resistivity; б graph of the natural electric field potential; B geoelectric section combined with the georadar results; r module schedule (from Kolesnikov et al. 2012)

with temperature change along the borehole are shown in Fig. 7.7b, boreholes P-5 and P-130. In both cases, the temperature decreases from about 15 °C to less than 13 °C at a depth of 70 m below zero in P-5, and close to 13 °C at 140 below zero in P-130. Interpretation of all geotherms in the form of isotherms at a cross section along the gallery indicates the zones where there are possible active karst channels (Fig. 7.7). The presence of these channels was confirmed by subsequent boreholes and investigated in detail by echo-sounding. On the basis of the research results of ground temperature fields and gradients in several different karst areas in China, Chengjie (1988) distinguished five types of temperature log (Fig. 7.8):

1. Straight line: Very common in cold water geothermal fields, conformable to a general positive geothermal gradient 2. Parabolic curve: Upper section conformable to positive geothermal gradient, encountering hot water in lower part where the temperature is high and stable 3. Logarithmic curve: Upper section conformable to positive gradient, encountering hot water in the lower part; temperature increasing rapidly 4. Positive abnormal: A conduit; the temperature of the recharge water is high 5. Negative abnormal: A conduit; the temperature of the recharge water is low.

7.8 Hydrochemical (Dissolution) Investigations

105

7.8

Fig. 7.7 Inverse geotherms in the area of karst channel flow behind Ombla Spring. a Spatial temperatures in the system of inverse geotherms; b selected thermograms with temperature gradients; c a cross section of the temperature field from Ravnik and Rajver (1998) and Milanović (2004)

Fig. 7.8 Geothermal curves. Chegjie (1988)

Hydrochemical (Dissolution) Investigations

Hydrochemical investigations are a powerful tool which enables characterization of numerous processes in areas of great hydrogeological complexity such as evaporites with karst. To estimate dissolution rates during reservoir operations a number of different tests on site, in the laboratory and with scale models were used during the investigation phases in a number of dams in evaporites: Kama Dam (Russia); Rogun Dam and the site of the Lower Kafirnigan Dam (Tajikistan); Mosul (Iraq); Gotvand Reservoir (Iran); Caspe Dam, (Spain), among the others. Comprehensive hydrochemical laboratory analyses and site investigations were done in the case of Kama Dam foundation, as well as for dam sites of Rogun and Lower-Kafernigan. Particularly important are long-term investigations of evaporite rock mass during operation of Kama Dam. Experiments and investigations are focussed to the processes important for filtration properties of rocks as: grout mix—groundwater, gel—groundwater, gypsum— post-grouting dissolution and gel—gypsum. Based on laboratory investigation of particularities of migration and sedimentation of silicates in natural environment the chemical and physico-chemical rules were derived of the genesis and complexity of ground water in the area of grout curtain. Detail explanation of laboratory tests and field work is presented at Chap. 6. To predict the solution potential of the gypsum at the Mosul Dam, there was laboratory modeling using rock samples from the site. At a scale of 1:200, centrifuge testing allowed investigations of dissolution rates on 20 cm long samples, representing seepage paths of 40 m. The rotating frequency in centrifuge experiments shall be of the order at least 3000–5000 rev/min, preferably up to 20,000 rev/min. One hour of this type of testing modeled the development of dissolution seepage routes over four years in real conditions, Anagnosti (1987).

106

7

Methods of Investigation

Fig. 7.9 Scale model of the Gotvand reservoir (Iran), section in evaporites. (Photograph by Mahab Ghodss)

A scale model of 150
50 cm was constructed to reproduce the geological conditions, including a gypsum layer, at the Caspe dam site. A model gypsum substrate, 20– 22 cm thick, was prepared by compacting successive layers (c. 5 cm) of gypsum to consolidate the substrate and so reduce voids. Capillary tubes were installed to serve as piezometers and for taking water samples. An impervious vertical screen served as a gravity dam. “The model of the dam seated on heavily karstified substratum (porosity 15%) shows that although the flow and porosity increase, in this case, the rate of dissolution decelerates with time”, Piqueras et al. (2001). A few different models were used in the case of the Gotvand Reservoir. One of them was a large physical model at a scale of 1:200, constructed to estimate dissolution rates in reservoir banks composed of salt and shaly-clayey sediments (Fig. 7.9). It was based on the real geology at the site (length 1:200 and high 1.5 m). The banks of the model (inclination of 35°) were made of two layers of salt blocks with 10 cm of clay between them, for a total thickness of 60 cm. Different 4-m-long sections of the model were protected with a geomembrane or clay of two different consistencies. At the downstream end of the model, a simulated dam with openings for intakes and overflow was built. Experiments started at a flow velocity of 4 cm/s. First indications of solution in the form of small collapses were detected after only two hours. After seven hours, a significant number of cracks had been created and there were some mass movements in the bank. After 24 h, *50% of the salt had been dissolved and, after 48 h, 100%. Dissolution was

almost twice as fast in the section protected by a geomembrane than in the sections protected by compacted clay. In general, these results did not give cause for an optimistic forecast!

References Abelson, M., Y. Yechieli, T. Aksinenko, V. Pinski, and B. Baer. 2013. Microseismic Monitoring of Concealed Sinkholes Activity It Mineral Beach, Ded Sea, Israel. GSI Report GSI/29. Al-Halbouni, D., E.P. Holohan, L. Saberia, H. Alrshdan, A. Sawarieh, D. Closson, T.R. Walter, and T. Dahm. 2017. Sinkholes, subsidence and subrosion on the eastern shore of the Dead Sea as revealed by a close-range photogrammetric survey. Geomorphology 285: 305– 324. Al-Zoubi, A., A-R. Abueladas, C. Camerlynck, R. Al-Ruzouq, S. Al-Rawashdeh,M. Ezersky, and W. Ali. 2007. Use of 2D multi electrodes resistivity imagining for sinkholes hazard assessment along the eastern part of the Death Sea, Jordan. American Journal of Environmental Sciences 3 (4): 229–233. Anagnosti, P. 1987. Prediction and control of seepage in soluble grounds. In IX European Conference, International Society of Soil Mechanics and Foundation Engineering, Dublin. Bodet, L., P.Y. Gilbert, A. Dhemaied, C. Camerlynck, and A. Al-Zoubi. 2010. Surface–wave profiling for sinkhole hazard assessment along the eastern Dead Sea shoreline, Ghor Al-Haditha, Jorda. In 72th EAGE Conference & Exhibition Incorporating EUROPEC 2010, 4. Barcelona, Spain. Bodvarsson, G. 1973. Temperature investigations in geothermal systems. Georxploration 11: 141–149. Bonacci, O. 1987. Karst Hydrology. With Specific Reference to the Dinaric Karst: Springed-Verlag, Berlin Heidelberg, New York, 184. Borić, M. 1980. The use of groundwater temperature changes in locating storage leakages in karst areas. In Proceedings of ^th

References Yugoslav Symposium, Hydrogeology and Engineering Geology, Portorož, Yugoslavia, 179. Chengjie, Z. 1988. A study of geothermal field and karstic leakage in karst area, In Proceedings of the IAH 21st Congress, Geological Publishing house, Beijing, China, 1127. Closson, D., and N. A. Karaki. 2009. Earthen dike leakage at the Dead Sea. In Engineering Geology for Society and Territory, eds. G. Lollino, A. Manconi, F. Guzzetti, M. Culshaw, P. Bobrowsky, F. Luino, Vol. 5. 461–464. Dordrecht, The Netherlands: Springer. Crain, E.R. 2003. Crain’s petrophysical handbook, chapter twenty. In Elastic properties of Rocks. Canada: Sponsored by Spectrum 2000 Mindware Ltd. Davydov, V.A., and G.A. Tsai. 2018. Study of dangerous natural and manmade geological processes using geophysical methods. In Russian. News of the USMU 2 (50): 65–71. Dortman, N.B. 1992. Petrophysics: A handbook. In Three Books. Book One. Rocks and Minerals. (In Russian), 391 Moscow: Nedra. p. 391. Drogue, C. 1985. Geothermal gradients and groundwater circulation infissured and karstified rocks. Journal of Geodynamics 4. El-Isa, Z., O. Rimawi, G. Jarrar, N. Abu-Karaki, S. Taqieddin, M. Atalah, N. Abderahman, and A. Al-Saed. 1995. Assesment of the Hazard of Subsidence and Sinkholes in Ghor Al-Haditha Area. Report submitted to Jordan Valley Authority, 141 Amman: Ubuversity of Jordan. Eppelbaum, L.V., M. Ezerski, A. Al-Zoubi, V. Goldshmidt, and A. Legchenko. 2008. Study of the factors affecting the karst volume assessment in the Dead Sea sinkhole problem using microgravity field analysis and 3D modelling. Advances in Geosciences 19: 97– 115. Ezersky, M., and A. Frumkin. 2013. Faults-dissolution front relations and the DS sinkholes problem. Geomorphology 201: 35–44. Ezersky, M. and L. Sobolevsky. 2002. Estimates of the Rock Mass Quality Using Geophysical Methods. Second stage. G.I.I. Report No.263/228/02, Israel. Ezersky, M. 2006. The seismic velocities of Dead Sea salt applied to the sinkhole problem. Journal of Applied Geophysics 58: 45–58. Ezersky, M., A. Legchenko, C. Camerlynck, and A. Al-Zoubi. 2007. The salt Formation Edge as a Major Indicator of the Sinkhole Hazard in the Dead Sea Western Coast. In 13th AEGE Near Surface Meeting and Exhibition, Istanbul, 67. Ezersky, M. 2008. Geological structure of the Ein Gedi Sinkhole occurrencesite at the Dead Sea shore in Israel. Journal of applied Geophysics 64: 59–69. Ezersky, M., A. Legchenko, C. Camerlynck, L. Appelbaum, S. Keidar, N. Baucher, and K. Chalikakis. 2010. The Dead Sea sinkhole hazard—new findings based on a multidisciplinary geophysical study. Zeitschrift für Geomorphologique 54 (2): 69–90. Ezersky, M. 2011. Improvement of the seismic refraction methods for mapping of the buried salt layers along the Ded Sea shoreline. GII Report No. 211/486/09, Lod, Israel, 24. Ezersky, M., L. Bodet, A. Al-Zoubi, C. Camerlynck, A. Dhemaied, and P.-Y Galibert. 2013. Seismic surface-wave prospecting methods for sinkhole hazard assessment along the Dead Sea shoreline. Journal Environmental and Engineering Geophysics 18 (4). Ezersky, M., A. Legchenko, L. Eppelbaum, and A. Al-Zoubi. 2017. Overview of the geophysical studies in the Dead Sea coastal area related to evaporite karst and recent sinkhole development. International Journal of Speleology Tampa, FL. 46 (2): 277–302. Ezersky, M., and A. Legchenko. 2014. Quantitative assessment of In-situ salt karstification using shear wave velocity, Dead Sea. Geomorphology 221221: 150–163. Frumkin, A., L. Kofman, and M. Ezerski. 2009. Improvement of the reliability of subsurface void detection, including sinkhole

107 development, at the Dead Sea shore area by means of Ground Penetration Radar (GPR). Hebrew University—TECHNION Research and Development Foundation LTD—Geophysical Institute of Israel, Report No. MNI-ES-36-2008, 145. Frumkin, A., M. Ezersky, A. Al-Zoubi, E. Akkawi, and A.-R. Abueladas. 2011. The Dead Sea hazard: geophysical assessment of salt dissolution and collapse. Geomorphology 134 (1–2): 102– 117. Goldman, M., D.D. Gilad, A. Ronen, and A. Melloul. 1991. Mapping of sea water intrusion into the coastal aquifer of Israel by the time domain electromagnetis method. Geoexploration 28: 153–174. Haenel, R., and F. Mongelli. 1988. Thermal exploration methods. In Handbook of Terrestrial Heat-Flow Density Determination, eds. R. Haenel, L. Rubach, L. Stegena, 353–389. Dordrecht: Kluwer Academic Publishers. Kafri, U., and B. Lang. 1997. Detection subsurface brines, freshwater bodiesand the interface configuration in-betweenby the time domain electromagneticmethod in the Dead Sea Rift, Israel. Environmental Geology 31: 42–49. Keydar, S., D. Pelman, and M. Ezersky. 2010. Application of seismic diffraction imaging for detecting near-surface inhomogenities in the Dead Sea area. Journal of Applied Geophysics 71 (2–3): 47/52. Keydar, S., B. Medvedev, A. Al-Zoubi, M. Ezersky, and E. Akkavwi. 2013. 3D imaging of Dead Sea area using weighted multipath summation: A case study. International Journal of Geopgysics. Keydar, S., B. Medvedev, M. Ezersky, and L. Sobolevsky. 2012. Imaging shallow subsurface of Dead Sea are by common short point stacking and diffraction method using weighted multipath summatiom. Journal of Civil Engineering and Science 1 (2): 75–79. Kolesnikov, V.P., A.V. Konoplev, A.M. Prigara, and A.V. Tatarkin. 2012. The Technology of integrated engineering and geophysical surveys for diagnosing the state of hydraulic structures. In Russian. Modern problems of Science and Education. 6: 630. Krawczyk, C., U. Polom, H. Alrshdan, and Al.-Halbouni, D., Sawarieh, A. and Dahm, T. 2015. New process model for the Dead Sea sinkholes at Ghor Al Haditha, Jordan, derived from shear-wave reflection seismic. In: EGU General Assembly. Viena, 5761. Id: Austria. Lachassagne, P., J.-M. Baltassat, A. Legchenko, and H.M. de Gramont. 2005. The links between MRS parameters and the hydrogeological parameters. Near Surface Geophysics 3 (4): 259–265. Legchenko, A., M. Ezerski, C. Kamerynchuk, A. Al-Zoubi, and K. Chalikakis. 2009. Joint use of TEM and MRS method in complex geological seting. Comptes Rendus Geosciences 341 (10–11): 908– 917. Legchenko, A. 2013. Magnetic Resonance Imaging for Groundwater,158. ISTE Ltd. Great Britain: Willey. Legchenko, A., J.-M. Baltassat, A. Bobachev, C. Martin, H. Robin, and J.-M. Vouillamoz. 2004. Magnetic resonance sounding applied to aquifer characterisation. Journal of Ground Water 42 (3): 363–373. Legchenko, A., M. Descloitres, A.A. Bost, L. Ruiz, M. Readdy, J.F. Girard, M. Sekhar, M.S.M. Kumar, and J.J. Braun. 2006. Resolution of MRS applied to the characterization of hardrock aquifers. Ground Water 44 (4): 545–554. Leghchenko, A., M. Ezersky, C. Camerlynck, A. Al-Youbi, K. Chalikakis, and J.-F. Giragrd. 2008. Locatink water-filled karst curves and estimating their volume using magnetic resonance soundings. Geophysics 73 (5): G51–G61. Milanović, P. 2004. Water Resources Engineering in Karst. Boca Raton: CRC Press. Polom, U., H. Alrshdan, D. Al-Halbouni, A. Sawarieh, T. Daham, and C.M. Krawczyk, 2016. Improved Dead Sea sinkhole site characterisation at Ghor Al Haditha, Jordan, based on repeated shear wave reflection seismic profiling. In EGU General Assembly, 6440.

108 Ravnik, D., and D. Rajver. 1989. Thermometric Investigations for Underground Dam Ombla: Dubrovnik (Phase I and Phase II), Report, Geologica Suravay, Ljubljana, Slovenia. Ravnik, D., and D. Rajver. 1998. The use of inverse geotherms for determining underground water flow at the Ombla karst spring near Dubrovnik, Croatia. Rybakov, M., V. Goldshmidt, L. Fleischer, and Y. Rotstein. 2001. Cave detection and 4-d monitoring: A microgravimetry case history near the Dead Sea. Leading Edge 20 (8): 896–900.

7

Methods of Investigation

Shuvalov, V.M. 2012. Integrated application of geophysical methods in solving problems of assessing the location, depth and form of local heterogeneity. In Russian. Bulletin of Perm University. Geology series. Issue 4 (17). 63–67. Vouillamoz, J.-M., A. Legchenko, Y.M. Albouy, M. Bakalowicz, J.-M. Baltassat, and W. Al-Fares. 2003. Localization of saturated karst aquifer with magnetic resonance sounding and resistivity imagery. Ground Water 41: 578–587.

8

Monitoring

8.1

General

Monitoring is a continuous process to measure and seek to control the qualitative and quantitative changes of different hydrogeological, hydrological, engineering geological, and hydrochemical parameters, first in natural conditions and later during the operation of the reservoir and dam. Because prevention is always better than cure, the role of monitoring is achieved precise diagnosis of any potentially destructive processes in real time. Organization of a good monitoring network, good instrumentation, and continuity of measurements are the important prerequisites for the protection of dam stability and the watertightness (impermeability) of the reservoir. A comprehensive monitoring network will be based on clearly defined targets and knowledge of the different conditions prevailing in the investigated area. In general, two different monitoring targets are: – Monitoring of any changes that may endanger the stability of the dam (e.g., permeability of the grout curtain, post-grouting processes, development of solution or suffusion caverns and channels in the bedrock foundation, collapses, and seepage losses in the reservoir), and – Monitoring of the regional impacts provoked by the modifications of water regime and water quality caused by the construction of dams and reservoirs (surface and groundwater regime changes, pollution, and other environmental impacts). One of the important and frequent concerns for all dam and reservoir projects is that the discharge of regional natural springs at lower elevations may decrease or cease to flow or become polluted. In engineering practice, these impacts are recorded in some cases at distances of a few, up to few tens of kilometers, downstream. Changes in the flow regime in downstream springs often create conflicts between the

owners of the dam and the users of the springs. Because of this, monitoring should start during the early stages of investigations and design. Data collected in the pre-dam natural conditions, so-called zero data, are the essential standards to be achieved and maintained as far as possible during the operation of the structure. Experience of building on bedrocks that include gypsum shows that the main problem endangering the safety of the structure is gypsum dissolution due to the major changes in the hydrochemical and hydrodynamic conditions created by installing the dam and reservoir. Serious problems may also be associated with the removal of unconsolidated deposits (clay, sand, pebble). It is necessary to organize continuous monitoring of changes in the parameters of the array that may indicate the beginning of undesirable processes from the security perspective. The major factor indicators of new or accelerated karst activity may be the levels, hydrochemical and thermal regimes of the groundwater and surface water, as well as any deformation (e.g. settling) of the surface of the ground. When a reservoir is in direct contact with a salt formation, the reservoir water is directly endangered by increasing amounts of dissolved salt. Monitoring the vertical distribution of the pertinent parameters in the reservoir should be established at the start of the impounding (chemistry—particular salinity, temperature, oxygen, turbidity). Increasing salinity in the reservoir was measured immediately after Gotvand and 15th Khordad, in Iran, were filled. Due to the high solubility of evaporites, with often rapid changes in topography and water pollution (sagging, collapsing, landsliding, and pollution with solutes) careful geodetic and hydrochemical monitoring plays a more essential role than it does in carbonate karst terrains. Design of the monitoring network is based on analysis of the current knowledge of the regional and local hydrogeological, engineering geological, and hydrological properties, and the characteristics of the structures (dams, etc.) to be constructed.

© Springer Nature Switzerland AG 2019 P. Milanović et al., Dams and Reservoirs in Evaporites, Advances in Karst Science, https://doi.org/10.1007/978-3-030-18521-3_8

109

110

8

At the very beginning, before the final monitoring network is installed, the monitoring regime (frequency of measurement) should be determined on the basis of a preliminary screening survey undertaken during dry and wet periods of the year. Only after at least one hydrological year of data has been obtained, should the final monitoring network set-up be established. Based on current experience in highly developed karst, during the rainy period(s) of year the aquifer regime is characterized by major and continuous fluctuations. In these circumstances, the data are only valid if the monitoring is continuous. Monitoring should begin immediately at the start of site preparation for construction (s), to prevent danger to human life and to equipment in the case of there being disastrous events (mostly floods) during the construction. The sources of crucial data should be weather forecasts and daily records of precipitation, groundwater levels (piezometers) and flow recorder data (rivers and springs). Underground works are particularly sensitive and dangerous. The burst of high-pressure water from a small karst channel (no more than 50 cm  50 cm) can provoke rapid flooding. If this happens immediately after heavy rain, the flood water usually contains clayey-sandy particles so a flood can be disastrous to any machinery that is caught in it. On the basis of continuous updating of all the monitored data, it is possible to determine the critical parameters that control the stability of the dam structure, grout curtain watertightness, and the unpredictable (in time) changes in the surrounding aquifer and surface. The general target of monitoring is to prevent the potential negative consequences immediately after a destructive process begins. The role of the monitoring data is to provide the information needed to detect the source of the problem and to suggest the most effective remedial measures.

8.2

Groundwater Regime

As with the monitoring of groundwater-level fluctuations during the investigation stage, piezometers and wells are of crucial importance during the operation of the dam and reservoir. The groundwater level regime in and around the dam/reservoir complex is the main parameter that allows quick judgment of any new karst activity that poses/may pose a threat to the dam. Intensive filtration and leakage from the reservoir along karst solutional groundwater channels leads to a change (usually an increase) in the temperature of the groundwater. Monitoring the temperatures in the piezometers, in the upper and lower pools and at the springs, will provide additional information about leakage from the reservoir and filtration zones along the karst channels.

Monitoring

The piezometric network can be used to conduct filtration/leakage studies using indicator methods (dye tracing, etc.). These studies can provide information about flow patterns, leakage rates, and their changes over time. An increase in filtration rates will indicate karst activity. The network of observation piezometers and wells should be capable of monitoring a large area because it is desirable to cover all the terrain within which the reservoir may have a significant influence on groundwater levels, water quality, etc. For example, at the Kama Dam, control over changes in the hydrogeological situation in order to assess the condition of gypsum in the base of the structure is carried out using a network of piezometers and wells to control three aquifers. Since 1955, the responsible department at the Kama HPP has been monitoring the chemical composition of water, water-level fluctuations, and flow rates. At the beginning of the operation, the hydroelectric power station was not sufficiently equipped with instrumentation. During its operation, the piezometric network was reengineered and now consists of 422 wells. The pressure piezometers are equipped with mercury manometers instead of the portable manometer panel used by the project in the beginning.

8.3

Monitoring of Spring Discharges

Water stored at higher elevations in reservoirs influences the regime in springs at lower elevations. The most frequent change is that springs at lower levels may cease to function. As a consequence of building, a reservoir both decrease and increase of the discharge of preexisting springs are possible, however. Decreasing spring discharge indicates successful preventive measures and a watertight dam site and reservoir. Change of discharge and chemistry of some preexisting springs can be a consequence of reservoir filling also. However, in many cases, the first filling of reservoirs in evaporites is followed by massive dissolution and the appearance of new springs downstream. In both cases, the natural regime of both surface and groundwater is disturbed by dam construction and can be source of socioeconomical and political problems. Because of this, all downstream springs (particularly springs with the dam site inside their likely catchment area) must be included in the monitoring program at least one hydrological year before construction starts. The appearance of new springs and an abrupt increase in their flow rate is an indicator of the threatening development of karst processes and requires urgent analysis and, if necessary, immediate remedial action. As stated before, an increase in the content of sulfates alone does not change the transparency and color of spring water; i.e., such these changes are not visually detectable.

8.3 Monitoring of Spring Discharges

All springs possibly affected by reservoir water should be carefully monitored. The monitoring program should include the following measurements: spring discharge, temperature, sulfate concentration, amounts of suspended particles, fluctuation of water level in the reservoir, and precipitation. Proposals to monitor any new springs that appear should be treated as a matter of urgency. Measurements of discharge, conductivity, and chemical analyses should be done on a daily basis.

8.4

Monitoring of Surface Flows and Reservoir Floors

An indicator of intensive karst development may be an imbalance in the flow of the river upstream and downstream of the dam, and its constant increase in downstream direction. At many cases, process is encountered instantaneously during the first filling of reservoir; however, this process can be slow and take lot of time, sometimes years. To estimate losses from the reservoir and its quantities due to seepage through the concentrated ponors or ponor zones, the precise hydrological measurements of discharge points are necessary. In evaporites, solution channels showed dramatic increasing of karst conduits and its opening. In spite of sizable magnitude of ponors (swallow holes) deep beneath the reservoir level, exact detection of sinking places and definition of its characteristics is complicated hydrogeological task. Particularly, hard and problematic is detection the seepage routes from sinking to the discharge points. Crucial problem is to find space position of channels with seepage flows. If seepage flow occurred through foundation rock mass, the grout curtain and dam structure are endangered. Different investigations and monitoring methods should be applied: mapping of reservoir bottom by application of sonar technique; mapping of reservoir bottom by divers; use of dye tracers to indicate zone with absorption points; to organize tracer test at every detected absorption point and to organize thermal measurements. It is necessary to conduct thermal observations periodically around the reservoir area, including measurements of the temperature of water in depth in order to identify areas of intense absorption by karst solution channels. Such studies make it possible to contour the seepage/leakage zones with great accuracy. It is particularly important to monitor the flows and springs in all adjacent lower valleys.

111

8.5

Hydrochemical Monitoring

In evaporites, hydrochemical monitoring of downstream springs, piezometric boreholes, drainage holes, and reservoirs is essential, particularly from the seepage and dam stability points of view, but also in cases where the water is used for water supply or irrigation. Dissolution of rocks, especially gypsum and salt, is accompanied by significant changes in the composition of surface and groundwater. Washout of unconsolidated sediments (clasts, sand, silt, etc.) from karst cavities leads to an increase in the content of suspended particles and the turbidity of the water. Water samples for complete chemical analysis and measurement of the suspended load should be taken from the observation wells at least once per month. From the results of the analysis, maps and profiles of mineralization are constructed, illustrating the pH and concentrations of SO24 , Ca2+, HCO3 and other components, which will make it possible to identify desalination (dilution) zones. According to the computer analytic method of Zverev (1967), the amount that the water is undersaturated with CaSO4 (the saturation deficit, i.e. the aggressivity) can be calculated. An increase in the saturation deficit indicates an increased risk of gypsum dissolution. It is also necessary to calculate the aggressiveness of water with respect to the concrete in order to get an idea of the effect of groundwater on the cement curtain and other concrete structures. During the first partial impoundment, the monitoring results for seepage through the Mosul dam site area clear indication of potentially enormous problems with seepage and dam stability. To analyze influence of the Gotvand Reservoir on solution and salinity in the reservoir, monitoring of the water quality and quantity in the river downstream of the dam site began 55 years before the reservoir becomes operational and has continued during its operation. Figure 8.1 presents a graph of the measurements of river flow (discharge) and electrical conductivity, 1968–2017. According to Meshkat et al. (2018), the influence of the reservoir has provoked meaningful changes of water quality downstream. This is due to the stratification of water in the reservoir; the highly saturated (and denser water) has been “trapped” at very bottom and the water above it can flow away downstream with only small concentrations of salt. For quantitative analysis, i.e., for measurement of total dissolved solids in the discharge, the installation of EC and flow recorders at monitored springs is mandatory.

112

8

Monitoring

Fig. 8.1 Gotvand reservoir. Plots of the discharge and electrical conductivity during the period, 1967–2017 (from Meshkat et al. 2018)

8.6

Surface Deformation and Progressive Erosion

The development of new karst underground may be accompanied by deformation of the land surface above. This can pose the greatest danger in the area of the dam and the main dam facilities. It is necessary to create a network of benchmarks and marks on the dam and other main structures to identify subsidence zones. LiDAR and drone photography and mapping are now becoming very useful because it is able to register decimetric scale changes at the land surface. Geodetic monitoring and installation of extensometers for precise observations of ground movement at subsidence sites (mostly vertical movement) can be very helpful in the detection of displacements at the centimetric scale. Such movements, particularly if they are progressive over time, indicate intense solution and the creation of voids in the rock beneath. If sinking occurs close to the dam, such movement is a clear indication of potential problems in the foundations. Recently, as a consequence of satellite technology, this monitoring technique has become quite simple and sufficiently precise. At the stage of filling, the reservoir and in the first years of its existence, leveling should be carried out quite often.

For example, a reference network has been created in the area of the Kama Dam that consists of 12 benchmarks for three reference points in the bush, 193 surface and deeper borehole markers, plus 47 spatial and tiltmeters installed on deforming rock beds.

8.7

Monitoring/Maintenance of Galleries

Monitoring/maintenance galleries located downstream of the watertight structures (grout curtains or cut-off walls) is one of the possible solutions to the problem of monitoring in evaporites. The gallery is multifunctional, used for investigation, monitoring, and maintenance and has to be excavated parallel with route of the watertight structure. The distance between the gallery and watertight structure, and its elevation, depend on local geological properties and the position of the main gallery. If the cut-off is located in the banks of the dam, the monitoring gallery (in general) should be 5– 10 m lower than the cut-off gallery. Piezometric boreholes should be arranged along the gallery and equipped with instruments for permanent monitoring of water table fluctuation, water temperature, chemistry, and other parameters (real-time data). In the case of seepage, local re-grouting should be done from the monitoring gallery. If necessary, an additional grout

8.7 Monitoring/Maintenance of Galleries

113

curtain row can be constructed along the gallery. In addition, the drainage boreholes in grouting galleries or in the drainage curtain (if they exist) should be incorporated into sampling program for chemical analysis.

8.8

Data Analysis and Safeguard Plans

According to the law and the proposals of ICOLD, the monitoring of dams is mandatory in all countries. In addition, a safeguard plan for the construction and operation is an important requirement in any engineering project, particularly if evaporites are present in the foundation rocks. The disaster caused by the catastrophic failure of the St. Francis Dam (USA), including hundreds of lost lives, or present situation with the Mosul Dam (Iraq), declared to be “the most dangerous dam in the world”, are examples that confirm the need for strict monitoring proposals and a clear safeguard plan. In both of these cases, the presence of gypsum played a crucial destructive role. The main aim of a safeguard scheme is to determine potential risk and to prevent potential hazard. The quality of data collected by the monitoring network has a crucial role. Rates of variation of solute sulfate concentrations obtained from historical data are important for analysis of the dissolution occurring during reservoir operation. Data acquisition from a monitoring network should provide continuous updating of the variation of parameters that can indicate changes in the bedrock foundations or rates of seepage through the grout curtain. Dissolution in evaporites

is extremely fast (particularly in halides); thus, the transmission of monitoring data in real time is the most important requirement. For instance, rapid change in the water chemistry at downstream springs, or any new sagging at the surface or collapses around the site area are “red flags” and need immediate rehabilitative action. An important requirement is also to establish the technical and operational rules for prevention or mitigation of destructive processes at the dam site and to minimize dam stability hazards and pollution hazards in the intake structures. These rules should contain alternative solutions, temporary or permanent, for the evacuation of people from endangered areas, the provision of alternative sources of water supply, etc. As part of a safeguard plan, the general approach to remedial works should be designed in advance. The best approach is seen when the machinery needed for remediation (drill rigs and grouting equipment) is available immediately at the site. Because dissolution in evaporites is extremely fast, the most effective approach is a prompt start with remedial works, immediately after rock deterioration has been detected. Any hesitancy allows the rapid and progressive increase of the initial voids. As a consequence, the volume and velocity of flow increases and may become turbulent. If remediation is not activated immediately after a problem is detected, truly efficient remediation becomes extremely complicated technically and its results questionable. Some indicators of hazardous initial karst degradation at a site are presented in Table 8.1. Based on a comprehensive analysis of such indicator factors, decisions can be made to provide the rehabilitation that is needed.

Table 8.1 Factors—indicators of the activation of destructive processes Monitoring network

Measured parameters

Indications of deterioration

The groundwater

The groundwater level regime

Reduction of head in the curtain zone Decrease of groundwater levels upstream of the curtain Increase of groundwater downstream from the curtain

a

The spring discharge regime

The appearance of springs downstream in the river valley; increase in their flow rates; the presence of suspended particles in spring water

The chemical composition

Increase of halides in the groundwater; presence of suspended particles Increase saturation with CaSOa4

The temperature

Groundwater temperature rise in wells and springs

The surface water

The flow

Imbalance of flow of the river upstream and downstream from reservoir; increase the downstream flow

The land surface

The geodetical mark

Appearance of the sagging zones and collapses

Alarming is also the reduction of levels of saturation in discharge waters if this is accompanied by an increase in the volume of discharge

114

References Meshkat, T., Mahjoob Farshchi, D., and Ebtekar, E. 2018. Evaluation of evaporate karstic challenge in Gotvand dam reservoir. In Proceedings of International Symposium, KARST 2018, Expect

8

Monitoring

the Unexpected, ed. S. Milanović, and Z. Stevanović, 89–96. Belgrad: Centre for Karst Hydrogeology. Zverev, V.P. 1967. Hydrochemical Studies of the Gypsum System— Groundwater, 100. Moscow: Publishing house “Science”.

9

Overview of Dams and Reservoirs in Evaporites

9.1

Summary of Dams and Reservoirs in Evaporites That Have Construction Problems

A number of dams and reservoirs constructed in soluble evaporite rocks all over the world have been affected by dissolution. From the beginning of the twentieth century, more than sixty dams have been affected by gypsum and salt dissolution problems. Some of the dams failed to retain water up to their design levels due to extensive leakage, some of them collapsed catastrophically, some others were abandoned, and some reservoirs suffer due to severe solute pollution of the stored water. Numerous dams in evaporites, mostly in gypsum, need rehabilitation: e.g., in Algeria, Argentina, Armenia, Azerbaijan, China, Dagestan, Georgia, Germany, Guatemala, Iran, Iraq, Jordan, Peru, Russia, Switzerland, Spain, Tajikistan, Tunisia, USA, Uzbekistan, Venezuela. A major role of this chapter is to summarize the experience reported in a series of scientific papers on dam and reservoir construction in evaporites. Because there are many dam projects under construction or in the design or investigation phases today that are in areas directly or indirectly influenced by evaporites, the list of dams presented below is not final. A number of the dams constructed in evaporites in the USA have encountered different problems, mainly leakage: the McMillan, Avalon, and Rio Hondo dams in New Mexico; San Fernando, Dry Canyon, Buena Vista, Olive Hills, Rattlesnake, Big Sage, Little Panoche, and Castaic dams in California; Stanford Dam (Texas); Fontenelle Dam (Oklahoma); and Moses-Saunders Power Dam (USA/Canada). Large caverns are developed in gypsum in the foundation of the Red Rock Dam (Iowa). The Clark Canyon Dam (River Overhead, Montana), constructed between 1961 and 1964, is founded on limestone, gypsum, anhydrite, dolomite. Due to the presence of gypsum in the foundation rock, catastrophic failures occurred in the St. Francis Dam (California, 1928) and Quail Creek Dike (Utah, 1989).

The proposed Upper Mangum Dam (Oklahoma) was abandoned before any construction, following investigations over the period, 1937–1999. Due to the presence of gypsum at the original Cedar Ridge dam site, a new site location was selected about 25 km upstream. Dams and reservoir in Spain reported to have problems or concerns because of foundation problems, seepage and collapse are Beninar Reservoir, San Juan Reservoir, La Loteta Dam (salt), Caspe Dam, San Loran Dam, Estremera Dam, and Alloz Dam. In Iran, a number of reservoirs suffer due to high concentrations of sulfides or the presence of evaporites in the bedrock foundations and reservoirs: Tang-E-Shemiran (Kangir) Dam and Reservoir, Khordad 15th Reservoir, Kowsar Reservoir, Seymareh Reservoir, Marun Reservoir, Gotvand Dam and Reservoir (salt), Jarreh, Gheisaragh, including dams currently under construction at Nargesi, Karachilar, and the Khersan III Dam. Some of the dams and reservoirs located in evaporites in Iraq have foundation and seepage problems: Mosul, Darbandikhan, Tharthar, and Hatra Dam. Dams and reservoirs known to be located in evaporites in China are: Huoshipo, Yangmazhain, Langdai, the Baiyanjiao, and Mahuangtian reservoirs. In Algeria, dam sites on rocks that contain evaporites are: Djedra Dam, Foum El Gherza Dam, and dam sites under investigations and design or dams under construction are at M’djedel, Fadi Naceur, Bouzina, Oued Mellagh. Dams that suffer due to the presence of evaporites in Tajikistan are: Lower Kafirnigan (not constructed), Baipazinsk, and Sangtuda Dam. The, Nurek and Rogun dams have salt in the foundation rocks. Dams exposed to risk from rapid solution of evaporites in Russia are: Kama Dam, Perm Region; Irigan Dam (not completed), Dagestan; Bratsk Reservoir, Siberia; Osa River Dam; Cheboksary Dam (not built); and a dam on the Iren River (not constructed). Dams at other sites containing evaporites include the Tbilisi Dam (Georgia); Erevan Dam (Armenia); Toktogul

© Springer Nature Switzerland AG 2019 P. Milanović et al., Dams and Reservoirs in Evaporites, Advances in Karst Science, https://doi.org/10.1007/978-3-030-18521-3_9

115

116

9

Dam (Kyrgyzstan); Mingachevir Dam (Azerbaijan); Khoda Afarin Dam (border of Iran/Azerbaijan); Meghri Dam (border of Armenia/Iran); Chardarinska Dam on the Syr Darya River (Kazakhstan); Farkhad Dam (Uzbekistan); Tannur Dam, King Talal, Mujib, and Lisan Peninsula Dike (Jordan); Joumine Dam (Tunisia); Moncenisio Reservoir (border of Italy/France); Hessigheim Dam, (Germany); Birsfelden (Switzerland); El Isiro Dam (Venezuela); Poechos Dam (Peru); the Casa de Piedra Dam (Argentina); Puebla de Pava (head race tunnel, Guatemala); and a few others known to us elsewhere in the world.

9.2

Selected Case Studies

9.2.1 Dam Sites and Reservoirs in USA Gypsum is present in thirty-two states in the USA, with salt also underlying portions of twenty-five of them. The most pronounced gypsum karst is in strata of Permian age. According to Johnson (2004), recent projects related to gypsum karst and dam construction problems in the USA include: (1) abandonment of the Upper Mangum dam site in southwest Oklahoma; (2) failure and collapse of the Quail Creek Dike in southwest Utah; (3) leakage from Horsetooth Reservoir and Carter Lake Reservoir in north-central Colorado; and (4) leakage from Anchor Dam in northwest Wyoming. The presence of evaporites has created serious problems at the Moses-Saunders Dam (Canada/USA) on St. Laurence River, the McMillan Dam in New Mexico, Cedar Ridge Dam in Texas, and the Lower St. Fernando Dam in California. St. Francis Dam (California) was a 60-m-high concrete gravity arch structure. Failure occurred in its second year of operation. It was the worst American civil engineering failure of the twentieth century, killing at least 432 people along the St. Francisco Canyon and Santa Clara valley downstream. The dam was designed and built in 1924–1926 for the water supply of Los Angeles. The foundation was in schist, with gypsum intercalations in the east abutment and beneath the dam. The rock foundation in the upper section of the west abutment was conglomerate (Fig. 9.1). According to Rogers (2007) and Rogers and Hasselmann (2013), there was no geological input on dam design and construction here in the 1920s. Several tension cracks were formed in the concrete and were recorded during the second year of operation. On March 12/13, 1928, the dam collapsed due to rejuvenation of a paleolandslide in the left abutment. About 1.52 million tons of schist slid down, provoking dam collapse, and a flood wave about 45 m deep swept down the canyon. Post-failure survey established that eleven of the twenty concrete sections forming the dam were displaced by

Overview of Dams and Reservoirs in Evaporites

the outbreak flood. Only one, the middle segment, was remained on its original foundations. According to Cooper and Calow (1998) and Rogers (2007), strong uplift was partly attributed to gypsum dissolution; “the forensic work on the St. Francis Dam failure illustrates the complex and interdisciplinary nature of working with earth, water, and structural systems, and conveys the frailties engineers and geologists possess, based on the limitations of their training and professional experience” Rogers (2007). The Upper Mangum dam site (Oklahoma). Proposals for construction of 33-m-high earth-fill dam at the Upper Mangum site began in 1937. The decision to consider the site was based only on its favorable topography, without proper respect for the geology despite a predominance of gypsum in the abutments, which were in the Blaine Formation (Permian) that is characterized by typical karst features: caves, sinkholes, and sinking streams. The bedrock foundation consists of 60 m of dolomite and shale with gypsum interbeds. West of the dam site, the formation underlying the Blaine Formation contains salt beds. Investigation and testing boreholes at the dam site showed that rock mass is highly permeable. Extensive karst features are frequent in both the surface and the subsurface: sinkholes, voids, open cavities, and clay-filled cavities (Fig. 9.2). Based on new data and field studies in 1999, it was recommended that the Upper Mangum site be abandoned: “Engineering measures needed to remediate karstic foundation construction here would add greatly to the cost of construction and still would not assure tightness of the reservoir or integrity of the dam” Johnson (2004). An alternative dam site, situated seven km downstream where the geological and hydrogeological properties are more favorable, is proposed. Quail Creek Reservoir (Utah) is formed by construction of a main dam and a dike. The dike is a 25-m-high earth-fill embankment structure constructed in 1985 over well-jointed and karstified strata consisting of dolomite, gypsiferous siltstone, gypsum, and sandstone. Gypsum is common in all the rock units in the form of laminae parallel to the bedding and as a cementation agent in many parts. Karst features were detected during the site investigations. When filling operations began, seepage appeared along the downstream toe of the dike and further downstream. Additional grouting programs were undertaken in 1986 and 1988, but despite these efforts to prevent dissolution of gypsum and leakage the flow increased, causing dam failure in 1989. Due to a huge collapse beneath the dam, about 70 m of the dike was completely demolished. The flood wave, 3–4 m high, provoked considerable damage and the evacuation of people downstream. As a remedial measure, the earth-fill embankment has been replaced with a roller-compacted concrete structure.

9.2 Selected Case Studies

117

Fig. 9.1 St. Francis Dam, showing the concrete monoliths displaced by the outbreak flood. Rogers and Hasselmann 2013 (Information by Horace Wildy)

Along the foundations, a concrete-filled trench 25 m deep and cut-off wall 600 m long have been emplaced through the gypsum karst zone (Payton 1992; Fig. 9.3). The new RCC Quail Creek Dike installations were completed in 1990. However, leakage beneath the dam foundations gradually increased during the next eleven years. Analysis of the water from a spring 130 m downstream showed that the water was transporting up to 1.4 tons of sediments per day out of the rock beneath the reservoir. This indicates that together with dissolution, massive [mechanical?] erosion has become an important process. After emptying the reservoir, more than 200 collapses were discovered in the bottom. In 2002, seepage losses had reached about 340 L/s. Dissolution of gypsum caused subsidence gap of about 12 cm to open between dam and bedrock. To intersect open fractures beneath the cut-off trench, an extensive grouting program was undertaken. However, it is expected that dissolution and erosion of the gypsum will continue. To extend the time before the next grouting program is needed a partial basin lining (*120,000 m2) was installed. This program includes compacted clay lining overlain by a geomembrane liner and topped with compacted fill (Payton and Hansen 2003). Anchor Dam (Wyoming, USA) is a 66-m-high concrete arch structure constructed in the early 1960s. The dam site is in a narrow and deep canyon in resistant sandstone and carbonate formations. The reservoir area upstream (Owl Creek) is underlain by a dolomite and red beds, with gypsum. Solution cavities in the dolomite beds were encountered during excavation at the dam site, and more than 50 sinkholes were documented in the reservoir area.

Meter-sized solution cavities were developed along subvertical discontinuities. Two of them had volumes of about 2000 m3. A first sinkhole was discovered in the reservoir area in 1955, 12 m deep and 10 m in diameter. The pre-dam sinkholes were filled with a selected material. A year after the reservoir started filling, a sinkhole nearly 100 m in diameter and 20 m deep developed approximately 350 m upstream of the dam. To prevent water loss, an earthen dike was constructed around it “but the height of the dike was below the crest of the dam”!! Since the 1960s (first filling), the reservoir has never filled to be more than just a small pond. A few tens of fresh collapses opened year by year. Failure of the Anchor Dam “is a textbook example of where not to build the dam” (Jarvis 2003). The McMillan Dam on the Pecos River in New Mexico (USA) was constructed in 1893. The rock-fill dam is 18 m high. The bedrock consists of about 100 m of dolomite and evaporites—anhydrite and gypsum with red silt and clay. Gypsum was known only in the abutments. During the site investigation and construction, no caverns and channels were observed, but during the first twelve years of operation they were quickly created in the left abutment, and the reservoir dried up. Sinkholes are exposed along the eastern shore of the reservoir (Gunnar 1965). Water leaks from the reservoir under pressure and appears as concentrated discharge at springs (the Major Johnson Springs) in the banks of the Pecos River downstream of the dam (Cox 1967). The relationship between reservoir level and volume of leakage increased tremendously (Fig. 9.4). Leakage of 0.28 m3/s at a reservoir level of 5 m increased to 2.8 m3/s for a level of 8.3 m. According to James and

118

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.2 Schematic cross section along the proposed Upper Mangum Dam, showing bedrock geology and karst features (caves, cavities, fissures, and sinkholes) in gypsum, dolomite, and salt. The section is

based on borehole data and field studies. View looking downstream to the southeast. Proposed conservation pool level (497 m) is shown (from Johnson 2003)

Lupton (1978) during the period from 1893 to 1942, dissolution of the gypsum created approximately 50 million meter cubic of karst channels and caverns! Horsetooth Reservoir (Colorado, USA, 1940) was created by four earth-fill dams with heights of 51.5, 73, 75, and 80 m. The reservoir is situated in the Permian/Triassic Lukins Formation, composed of sedimentary claystone and siltstone, with several limestone units that include brecciated and karstified gypsum and anhydrite. The gypsum and anhydrite contain local cavities and vertical collapses that cross all units. During the site investigations, evaporites were not observed in boreholes at the dam foundation. Over the period, 1980–1990, sinkholes formed in the reservoir and seepage losses increased dramatically. In 2000, a sinkhole

formed in the upstream toe of the dam (Pearson 1999; Johnson 2004). For repair work at the Horsetooth Dam, the reservoir level was lowered in 1999–2001, exposing a few collapses in the banks. Detailed water chemistry was undertaken to determine the changes in seepage chemistry, with emphases on mass loading (calcite + gypsum) and void volume formation. A summary of high and low seepage void formation rates (observed since 1986) that are assumed to be constant over the 50-year life of Horsetooth Dam along an assumed flow path length of 900–1000 m is (i) for a low scenario, daily void formation rate—0.0495 m3/day; for 50 years—905 m3; and simulated conduit diameter— 1.12 m and (ii) for a high scenario (maximum dissolution), daily void formation rate—0.157 m3/day; for 50 year—

9.2 Selected Case Studies

119

Fig. 9.4 Relationship between stage and leakage from Lake McMillan, based on 1954–1959 data (from Cox 1967)

Fig. 9.3 Configuration of the cut-off trench in the main part of the new Quail Creek Dike (from Payton 1992)

2860 m3; and simulated conduit diameter—1.91 m (Craft and Pearson 2002). Carter Lake Dam (Colorado, 1950) is an earth-fill structure 71 m in height and with a crest length of 410 m. The dam foundation consists of the same geological formations as at Horsetooth Dam, limestone and gypsum with cavities, interstratal breccias, paleocollapse chimneys, and other karstic features (Pearson 1999). Seepage downstream of the dam is more than 100 L/s.

The Cedar Ridge Dam (Clear Fork Valley, Texas) is an example where there was a major relocation following adequate field investigations at the originally planned site. By exploratory drilling along the 1.6 km site, beds of Permian gypsum were encountered, although at the surface gypsum was exposed in outcrop at only one place a little upstream of the proposed dam location. To determine properties of the subsurface gypsum formations, the geoelectrical logs of almost 100 oil wells around the broad area were examined and correlated with cores from nearby oil wells. Gypsiferous sequence, 30–40 m thick, discovered by drilling, consisting of eight gypsum beds (1–3 m thick) is interbedded with shale layers (Fig. 9.5). “If a dam is constructed upon gypsum, or if lake water is impounded too closely above gypsum in Clear Fork Valley, it could be detrimental to dam integrity. Potential karst development in gypsum could provide pathways for impounded water to escape from the reservoir and be discharged downstream of the dam” (Johnson and Wilkerson 2013). This led to a new dam site (A) being adopted upstream from borehole SB-4 and about eight km upstream of the original site (CR). At the new location, the uppermost gypsum is at least 23 m beneath the dam site. It is assumed that the upper shale and other strata are sufficient to separate the reservoir water from the gypsum sequences.

120

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.5 Cedar Ridge dam site. Schematic cross section showing westward dip of gypsum beds beneath Clear Fork and dam sites, CR and A. The top of the gypsum sequence is above stream level at CR and is about 23 m below stream level at dam site A (from Johnson and Wilkerson 2013)

The Moses-Sounders Hydropower Dam on the St Lawrence River (USA/Canada border, 1958/59) consists of several concrete structures: a main dam, 60 m in height and 979 m long; the Long Sault Spillway Dam (33 m high and 902 m long); an intake structure (33 m high and 220 m long); and 18 km of dikes. One single bed of gypsum 0.6 m thick within the 50-m-deep dolomite foundation has been a recurring source of problems. Lower St. Fernando Dam (California, 1912) is a hydraulic-filled structure 44 m high. After impounding, there was seepage through gypsum intercalated in shales, siltstones, and sandstones. Remedial grouting was applied. A force 6.6 earthquake in February 1971 caused the upstream face of the dam to fail due to liquefaction. To avoid the possible danger, about 80,000 people were temporarily evacuated. Clark Canyon Dam on the River Overhead (Montana) was constructed in 1961–1964. At the base of the dam are readily soluble rocks: calcite, gypsum, anhydrite, dolomite (Craft 2006). The presence of leaks from the reservoir caused repairs to work in 1984–1985, after which the filtration losses were judged acceptable. However, the design maximum water level in the reservoir was not reached. Vigorous karst solutional activity was noted here. On the left bank and in the central section of the dam, a significant increase in dissolved sulfates has been measured. It is predicted that further dissolution of gypsum and anhydrite and increase of leakage may occur, especially in the years of high water.

Dam №6 was built in 1957 on the Vashita River, Oklahoma (Gunnar 1965). Gypsum is widespread in the area. After the reservoir filled in the spring period, karst solution funnels began to form on its right side, leaking large amounts of water. The cost of the compensatory cementing works was several times higher than the entire cost of the dam construction itself. A special method was developed to eliminate the funnels.

9.2.2 Dams and Reservoirs in Spain The San Juan Reservoir (Huesca Province), with a capacity of 850,000 m3, was built in 1999 for irrigation purposes. This earth dam with a clay core forms a closed basin roughly trapezoidal in shape (Fig. 9.6). The length of the reservoir is 400 m and width 230 m. Depth of the basin is 19 m and depth of water 17.5 m. A drainage system was constructed along the top of the dam (Gutierrez et al. 2003). The reservoir is situated on a Quaternary age footslope pediment overlying tertiary clays. The pediment consists of silty sand with up to 40% gypsum, plus paleochannels filled with limestone, quartzite, gypsum, and conglomerate pebbles. These deposits cover the Serinena Formation of shales and marls with sandstone intercalations. Shortly after filling, a small (6 L/s) seepage zone was detected. The chemical composition of seepage water indicated intensive dissolution with an SO4 concentration of 1482 ppm. A failure occurred during the first test filling in 2001. There was a huge breach in the earth dam due to the

9.2 Selected Case Studies

Fig. 9.6 San Juan Reservoir, Spain. a Cross section in the failure zone and b plan view of the reservoir (from Gutierrez et al. 2003)

use of old alluvial deposits containing considerable amounts of gypsum in the dam foundations. The breach size was 10 m deep and 16  10 m in width. A flood wave of about 300,000 m3 discharged over a short period. The early analyses had not noticed the presence of gypsum dispersed in the alluvial deposits. However, according to studies of Gutierrez et al. (2003), most probably, the failure of San Juan earth dam was the consequence of a combination of processes: dissolution (karstification) of the detrital and secondary gypsum in the pediment deposits, piping that affected the embankments and core of the dam, and biogenic burrows in the pediment deposits. According to these authors “since the San Juan reservoir will operate in the future with the same design, it would be desirable to impermeabilize the reservoir basin with a geotextile.” The La Loteta Reservoir, with a storage capacity of 105 hm3, was created by construction of a 34-m-high and 1.5-km-long embankment dam in the Zaragoza region of the Ebro River Valley (Gutierrez et al. 2015). The bedrock is an 850-m-thick Oligocene–Miocene formation including anhydrite, halite, and glauberite. The reservoir is located “in a large subsidence depression around 6 km long resulting from interstratal karstification of halite- and glauberite-bearing evaporites” (Gutierrez et al. 2015). The foundation and abutments of the dam and entire reservoir area include a horizontally bedded, laterally extensive, gypsum unit eleven

121

meters in thickness. Due to the structure being horizontal, the potential for leakage was high. Other evaporites (halite and glauberite) are present 45 and 70 m below the dam body. The water-retaining structures include a vertical clay core, a horizontal clay blanket, and a cut-off wall (Fig. 9.7). Grout curtains 675 and 255 m long are constructed in the left and right abutments. A cement–fly ash–bentonite cut-off wall was built along a continuous single-slot cutter trench using the panel wall technique: It is 1587 m long, 0.8 m wide, with an average depth of 23 m (Lafuente et al. 2006). The lower portions of the grout curtains penetrate into the clayey sediments; however, it was not possible to extend them down to deeper impervious formations. Figure 9.8 presents the general layout of the curtains in both abutments and partial sketches showing the distribution of grout borehole rows in different sectors. Sections with higher grout consumption are indicated. Locally, in these sections, grout mix consumption up to 1496 kg/m was recorded. Such high take indicates the existence of a cavernous zone in the lower part of the gypsum unit, directly beneath the body of the dam. Groundwater levels recorded in piezometric boreholes show that water flows (seepage) occur through the grout curtain in both abutments (Fig. 9.9). As a consequence of this destructive infiltration, several collapse sinkholes have occurred upstream of the left end of dam, and slow subsidence (14 mm/year) has been measured on the left edge of the dam crest (Gutierrez et al. 2015). Sinkholes detected in the reservoir area identify possible underground directions of the main leakage routes. The distribution of the discharges (springs) indicates that there is flow through the number of different branches. When the reservoir reached its maximum level (2011), total leakage was only 46 l/s, i.e., negligible. In spite of additional grouting in the left abutment after impoundment, in 2013 leakage had risen to 55.5 L/s. This, still small but increasing, volume of leakage suggests the possibility that there is progressive dissolution along the flow route(s), the size of conduits is increasing, and seepage may become unacceptably high. Dam integrity becomes questionable, also. According to Gutierrez et al. (2015), the possible additional remedial measures suggested include a cut-off wall linked to the existing clay core in the left abutment, from the dam crest for around 100 m of length, 31 m deep, and 0.6 m thick. Caspe Dam (Guadalupe River) is an earth-fill structure with a crest length of 400 m and 51 m in height. The dam and reservoir are located on subhorizontal beds of tertiary marls and gypsum (in joints) which also contain smaller amounts of anhydrite, halite, glauberite, mirabilite, and epsomite. A comprehensive analysis of the Caspe Dam seepage problem was presented by Mancebo Piqueras et al. (2011), based on

122

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.7 La Loteta Dam. Simplified cross section through the dam (from Gutierrez et al. 2015)

Fig. 9.8 La Loteta Dam, Spain. General layout of the grout curtain (from Gutierrez et al. 2015)

detailed hydrogeological field investigations and experimentation in a scale physical model of the dam site, including a dam structure model (Fig. 9.10). The near upstream bottom of the reservoir is protected by a clay blanket (Fig. 9.10a). A grout curtain, 25 m deep, was

constructed from an inspection gallery and connected to the clay core. Distance between grouting holes was 3 m. Dam construction was completed at the end of 1987. During the first filling of the reservoir seepage was recorded at three places. Three months after filling started seepage was found in an old tunnel plugged with concrete, where discharge rapidly increased to 200–250 L/s. A year after the impounding started, new seepage of about 30 L/s began on the left margin of the dam: After additional grouting, this was reduced to 0.3 L/s. Seepage in the storm overflow channel (Gauge No. 6) was also plugged by grouting (Fig. 9.11). Forty-nine piezometric boreholes were drilled to analyze parameters of the seepage. Investigations included geophysical methods (borehole logging and seismic tomography) and tracer tests. Groundwater flow velocities were found to range between 20 m/day beneath the dam and 40 m/day in the abutments. When a dam is constructed on evaporites, the groundwater regime is drastically changed. As consequences of a number of processes and reactions between the highly soluble minerals, the groundwater flow may increase or diminish the porosity. Processes that can increase porosity are mineral dissolution, increase of voids by transformation of glauberite to gypsum, and increase of voids by physical erosion. Porosity is reduced by hydration (anhydrite can increase its volume between 30 and 67% as it hydrates to gypsum) (Mancebo Piqueras et al. 2011). Besides dissolution, the transformation of glauberite to gypsum with formation of mirabilite may lead to a volume reduction of 28%, creating cavities. With additional remedial works, the Caspe leakage was brought under control, “although injections are required to maintain the grout curtain.”

9.2 Selected Case Studies

123

Fig. 9.9 La Loteta Dam, Spain. The equipotential map at the time of maximum groundwater level indicates seepage through the grout curtain in both dam banks (from Gutierrez et al. 2015)

Fig. 9.10 Caspe Dam, Spain. a Schematic cross section perpendicular to dam and b geological cross section along the dam axis and the location of seepage (from Morlans et al. 2005)

124

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.11 Caspe Dam, Spain. Effects of grouting on seepage into the overflow channel (Sanchez-Albornoz 1992, from Mancebo Piqueras et al. 2011)

9.2.3 Dams and Reservoirs in Russia Kama Dam, on the Kama River, is a concrete gravity structure 25 m high and 2.5 km long (Fig. 9.12) that was constructed in 1954. At the dam site and reservoir area, there

Fig. 9.12 Kama Dam, Perm Region, Russia. Photograph E. Tyurin

is a complex lithological sequence: argillites, sandstone, gypsum, dolomites, and anhydrite (Gorbunova et al. 1991; Lykoshin et al. 1992; Maksimovich N.G. 1983–2009). Upper sandstones, alleurolites, argillites, and limestone are 5–10 m in thickness, underlain by a more continuous carbonate

9.2 Selected Case Studies Table 9.1 Content of gypsum in the Kama Dam bedrock foundations (Maksimovich 2006)

125 Layer

Thickness (m)

Type of gypsum

Average percentage of gypsum %

Argillite and aleurolites

5–7

Dispersed, separate crystals up to 5 cm

1.3

Compact limestone

4–5

Dispersed

2.0

Dolomite

2.5–3

Thin interbeds up to 0. 5 m

15.9

Shaly limestone

6–8

Dispersed; joint fillings

3.1

Gypsiferous shale and dolomite

11–19

Interbedded lenses, total thickness about 9.5 m

35.2

Clayey dolomite

9.0

Dispersed

4.9

Dolomites, porous

3.5

Dispersed

Gypsum

5.0

100

Anhydrite

115

–

formation with gypsum beds totaling 40–45 m thick: This formation has six stratigraphic sequences, with gypsum present in five of them (Turcev 1938; Kuznecov 1953). Maksimovich (2006) recognized that the percentage of gypsum varies in the different beds, etc (Table 9.1). In the rocks in Table 9.1, gypsum is mostly present in a dispersed form. Crystals of gypsum in argillites and aleurollites are usually damaged (Fig. 9.13). When separated, the contact between gypsum and other rock surfaces is smooth (Fig. 9.13a). Dispersed gypsum displays different forms: as microscopic crystals (Fig. b and c) and nests and lenses that consists of chaotically oriented crystals within an admixture of other minerals (Fig. d). The rocks beneath the dam body are characterized by heterogeneous fracturing and have uneven vertical distribution. In the sulfate-carbonate strata as a whole, there is an increase in the content of calcium sulfate with depth. At the design stage, a few different waterproofing concepts and techniques were considered. One was emplacement of a permanently frozen curtain, 10 m thick, extending deep into the gypsum and anhydrites. Also discussed were the possibilities of applying different chemicals to build an insoluble layer over the gypsum. Finally, to prevent intensive dissolution, an upstream impervious blanket was constructed extending 110 m from the dam on the upstream side, with a double row grout curtain along the upstream perimeter of the blanket and a vertical drainage curtain in the middle (Fig. 9.14). The right bank grout curtain is 200 m long and at left bank 70 m. These structures should prevent pressure in formations with gypsum and decrease dissolution capability in lower formations. However, during operation stage of dam these structures were not able to achieve the designed decrease of pressure (27%). Karstification processes became active immediately after reservoir impounding (Maksimovich 2006, 2009). Over time, the grout curtain lost its efficiency. During impoundment up to a height of 22 m, seepage began to be detected below the

12.6

dam foundations. An abrupt increase of sulfates in the seepage water indicated that gypsum dissolution was increasing significantly. As a consequence of fluctuations in reservoir water levels, new collapse sinkholes were created in its vicinity. Before dam construction, two (natural) collapses had been recorded during a period of 50 years. After the reservoir began to operate, there were eleven new collapses within six years. According to Pechorkin and Pechorkin (1979) due to the intensive solution, 40–50 m3 of gypsum has been removed from each linear meter of the reservoir banks. The problems encountered during the operations at Kama Dam confirm the inefficiency of using only classic cement-based grout curtains. The data collected indicate that there was serious deterioration of the waterproofing properties over time. To improve grout curtain efficiency where gypsiferous rocks are present, application of a silica solution was analyzed by Buchatsky et al. (1976) and Voronkevich et al. (1986). This kind of solution achieves effective waterproofing by becoming a gel (from colloidal solution) after its injection into the rock mass. Detailed information on this process is presented in Chap. 6. To improve the resistance and density of the grout curtain in the gypsiferous rock under Kama Dam, a chemical gel-forming grout mix with penetration capacity similar to that of water was introduced. This grout mix (an oxaloaluminosilicate) consists of two components—sodium silicate with a density of 1.19 g/cm3 and a hardener (Maksimovich 2006). In 1974, an additional grout curtain in the reservoir bottom area (465 m long) and two lateral curtains (each 100 m long) were constructed. After application of this chemical mixture, seepage through the Kama curtain was reduced. This led to decrease of the filtration pressure in the foundation beneath the dam structure and, over time, has increased the stability of the dam (Maksimovich 1983). Thus, the example of the Kama Dam shows that careful geological studies, groundwater regime observations, and a combination of engineering and geological measures to

126

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.13 a and b, gypsum crystals in argillites and dispersion gypsum, c in argillites, and d in aleurollites, Maksimovich 2006

increase rock stability have allowed the successful operation of hydraulic structures in this area of sulfate rocks for a long time (Maximovich and Meshcheryakova 2009). Irigan Dam in Dagestan is an earth-fill structure 111 m in height and 317 m long. Dam impounding was completed in 1996. The dam site area consists of dolomite above sandstone formations. The upper part of the dolomite, 20 m in thickness, now contains between 3 and 5% gypsum, which increases to 18% below. Carbonate and sulfate karst are well developed in the area. Waterproofing structures include a cement-based grout curtain beneath the dam foundation and laterally into the banks, a concrete plate beneath the dam core, and a drainage curtain. To prevent increasing karstification, an impervious blanket has been constructed for 100 m upstream from the dam toy. Along the upstream perimeter of this blanket, a second grout curtain was built that included a drainage structure extending downstream under the curtain. These measures have made successful operation of the dam possible.

Chirkeiskaya Dam on the Sulak River (Dagestan) is a concrete arch structure (Figs. 9.15 and 9.16). The dam was fully commissioned in 1981 and is the second highest dam in Russia and the highest arch in the nation (230 m). Chirkeiskaya Dam is located in a narrow gorge with a depth of more than 200 m. The width of the gorge is 300 m in the upper parts, narrowing to only 12–15 m in the bottom. The walls of the gorge are very steep, with a volume of about 300 thousand m3 of potentially unstable blocks. The bedrocks are of Upper Cretaceous age, mainly platy limestones with intercalated marls and clays that display considerable fracturing. Fractures have opened along the gorge to depths of 150 m and widths up to 0.5 m. The rock is weakened by the presence of soluble gypsum and karst features. Miatlinskaya Dam is the counter-regulator of the largest in the cascade of the Chirkeiskaya HPP. The Miatlinskaya Dam is a medium-pressure, dam-derived hydroelectric power station. Most of the hydraulic pressure for the turbines is created by the dam itself, plus a smaller proportion of the

9.2 Selected Case Studies

127

Fig. 9.14 Kama Dam, Russia. a Cross section perpendicular to dam. b Cross section along the dam axis. Maksimovich 2006

water directed through a derivational diversion tunnel. The base of the dam is also composed of highly fractured rocks, including gypsum (Fig. 9.17). In the initial design, the dam was supposed to be located further downstream. But in the process of construction there was a huge landslide on the right bank, displacing a volume of more than 17 million cubic meters (Fig. 9.18)! The access road was shifted downward for several hundred meters. The lower part of collapsed mass entered the river. After that, the sliding became very slow (one to two centimeters per day), moving toward the foot of the slide. It was necessary to change the project drastically, halting building of the hydropower plant and moving the dam upstream. Water is now routed through two-kilometer tunnels in the bedrock on the opposite shore. To assess the safety of the hydraulic structures on a fractured gypsum foundation, a procedure was developed that included calculating the main groundwater seepage characteristics prior to and after leaching. For laboratory analysis were used three rock samples. To verify resistance of deposits in fractures of the foundation massif against local filtration, three rock samples were tested in laboratory. Based on the methods developed, the safety of the fractured gypsum base of the Miatlinskaya arch dam was

calculated, with allowance for residual strength after seepage solution at the base, and found to guarantee reliable operation of the dam for a long period, significantly exceeding the estimated service life. Comparison of the calculated, experimental, and field data on seepage characteristics in the rock base of the dam showed satisfactory agreement (Balamirzoev 2005). Bratsk Reservoir in the Angara River valley is retained by a concrete gravity dam 124.5 m high and 4417 m long. A part of the reservoir is situated on a soluble Lower/Middle Cambrian formation (dolomite, limestone, anhydrite–dolomite, and gypsum). Under natural conditions, karst forms at the surface included large collapse sinkholes and shafts (up to 60 m in diameter and 25 m in depth), blind creeks, and caves and karst channels underground (Kozyreva and Trzhtsinsky 2004). During reservoir filling (1963–1966), about 200 new sinkholes per km2 were induced every year, ranging 2–30 m in diameter and 5–28 m in depth. Genesis of these sinkholes was a consequence of reactivation of paleokarst features and the fluctuations in water levels in the reservoir. Some of them occurred at distances as great as six km from the river channel. Close to the reservoir, fresh caverns in gypsum are now up to 30 m deep and about 2 m in diameter. There have been a number of landslides due to

128

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.15 Chirkeiskaya Dam and Reservoir https://commons.wikimedia.org/wiki/User:Andshel`` \o ' User:Andshel

Fig. 9.16 Panoramic view of the Chirkey Gorge, reservoir and hydroelectric power station (https://commons.wikimedia.org/wiki/User: Caйгa20К)

9.2 Selected Case Studies

129

Fig. 9.17 View from downstream of the Miatlinskaya Dam (https://russos.livejournal.com/1004470.html)

Fig. 9.18 Landslide on the bank of the Sulak River (https://russos.livejournal.com/1004470.html)

solution. Based on detailed investigations and cartographic volumetric modeling, it is concluded that the reservoir shoreline will retreat as much as 10 m per year in some sectors. In many areas of gypsum/anhydrite outcrop, new collapses of 20–38 m-depth have been created. As a consequence, a number of buildings have been damaged and much land has become unsuitable for use.

9.2.4 Dams and Reservoirs in Tajikistan Lower Kafirnigan Dam is situated on the Kafirnigan River, Tajikistan. An earth-fill dam structure about 70 m high is proposed (Fig. 9.19). The foundation consists of argillaceous rocks with frequent layers of gypsum up to 35 cm in thickness. Gypsum layers are visible in outcrop along the

130

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.19 Lower Kafirgan Dam (Tajikistan). Measures suggested to be applied to prevent intensive solution and seepage losses through the rock that contains gypsum (from Nadriga and Demyanova 1986)

valley banks. Estimated permeability of the rock mass with gypsum layers down to a depth of 20 m is K3 = 0.25 m/24 h and deeper below K4 = 0.037 m/24 h (Nadriga and Demyanova 1986; Lykoshin et al. 1992). The suggested waterproofing structure consists of: 1. a grout curtain through the higher permeability zone; 2. horizontal impervious apron (asphalt concrete) extended 210 m upstream of the top of the dam; and 3. a dike built of gypsum along the upstream border of apron. The concept here is that when water percolates through the thick gypsum dike it will become saturated with respect to gypsum and thus have low dissolution potential. Intensity of dissolution (karstification) of gypsum beneath the dam foundation will be negligible or completely prohibited. Creation and enlargement of karst solution channels will be very limited. According to detailed analyses, including modelling by Lebedev and Lehov (2011), , a gypsum dike along the edge of the impervious blanket will be efficient about 100 years. Rogun Dam on the Vakhsh River is a rock-fill structure 335 m in height and 1300 m long. The foundation rocks are a terrigenous Cretaceous formation over Jurassic gypsiferous argillites that include a layer of salt (Fig. 9.20). At depths more than 200 m, the thickness of salt deposits is more than 10 m. The key problem is the outcrop of the salt at dam/rock foundation contact. Two crucial remediation approaches are required: to decrease salt dissolution to a minimum and to prevent transport of dissolved products away from the salt layer. To achieve these, the structure consists of a hydraulic grout curtain coupled with a very salty liquid screen (Fig. 9.21). The grout curtain consists of one row of grouting boreholes situated downstream (down-gradient) of the salt deposits. The salt screen comprises a row of boreholes for the injection of highly concentrated salt water. This protective system was installed from two galleries constructed for the purpose.

Fig. 9.20 Rogun Dam, Tajikistan. Geological properties of the dam foundation and schematic presentation of protective measures against salt dissolution (from Osadchi and Bahtiyarov 1975)

Nurek Dam is a 300-m-high rock-fill structure with a cement–clay core on the Vakhsh River, constructed between 1961 and 1980 (Fig. 9.22). The dam site consists of the following formations: Jurassic—clay, gypsum, and salt formations that are extremely tectonized and karstified (Demyanova 1986); Lower Cretaceous—sandstones and aleurolites, locally containing beds of limestone and gypsum; Upper Cretaceous—argillites, aleurolites, sandstone, and claystone, with rare layers of limestone and gypsum; Paleogene—limestones, carbonate clays, and gypsum; and Neogene—sandstone and sandy clay. The monitoring network of hydrogeological wells showed that there was salt saturation (brine strength), creating significant seepage losses after construction of an upstream cofferdam, and then became most intense after raising the reservoir to the Normal Water Level (NWL). In order to reduce the intensity of seepage, it was recommended to create a blanket over the upstream slope and the adjacent section of the reservoir bottom by soil (fill) with a high gypsum content: Fill thickness was recommended to be not less than 0.5 m, i.e., to be in accordance with the

9.2 Selected Case Studies

131

Fig. 9.21 Rogun Dam, Tajikistan. Protective structures to prevent salt dissolution in the dam foundations (from Lykoshin et al. 1992)

Fig. 9.22 Nurek Dam, Tajikistan. Geological cross section perpendicular to the dam structure (from Lykoshin 1992)

length of the section necessary to achieve complete saturation of the water with gypsum. In the bottom of the valley, a concrete plate 25 m was constructed under the dam foundation. At its base, consolidation grouting was carried out to a depth of 20 m, and a cement grout curtain 80 m deep was constructed beneath the core of the dam. No other special salt protection measures were made. During the construction of the dam and filling of the reservoir, there was intensive sedimentation on the bottom, but no indications of dissolution of bedrocks in the foundation have been detected during its period of operation. Baipazinsk Dam on the Vakhsh River was constructed in 1968 by controlled blasting of local rock walls. The river bed and adjoining parts of the valley were covered by blasted rock debris. The dam height is 55 m. The site has a complex geological structure including Cretaceous,

Paleogene, and Quaternary formations. A major property of these formations is intensive karstification of carbonate rocks and the presence of gypsum in the reservoir bank (Fig. 9.23). Particularly important is a *20 m gypsum/ clayey formation. Thickness of gypsum layers in it ranges from a few cm up to 2–3 m. The frequency of gypsum veins, lenses, and layers increases with depth. The upstream dam face is protected by an impervious screen and the bottom with an impervious blanket. There is similar protection on the left and right banks close to the dam itself. There is an anti-filtration screen and a drain from the upper slope of the dam to the reservoir bottom, on the slopes of the left and right banks, on the section between the dam and the hydrotunnel in order to reduce seepage through the dam or bypassing the dam, as well as to achieve the proper tight seal between dam and banks. In the middle of the spillway heel, part of it with a width of 23 m is in an Altaic gypsum clay mass. Spillway anti-filtration protective measures were implemented in accordance with the VODGEO scheme, in the form of a concrete slab on the front of the spillway, a watertight curtain under the concrete-reinforced walls, anti-filtration spurs in sections of onshore abutments, plus drainage structures under the slab and in the rapid flow zone. The main goal of these structures is to minimize the hydraulic gradient of seepage under the dam in order to ensure preservation of the bearing capacity of the foundations in the presence of a significant amount of gypsum. Drainage under the slab is designed to intercept a significant and most aggressive part of the filtration flow saturated with dissolved gypsum. Observations have shown that, due to the design adopted and silting of the reservoir, the hydropower unit is functioning favorably with regard to seepage and structural deformation hazards.

132

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.23 Baipazinsk Dam, Tajikistan. Geological cross section following the dam axis (from Lykoshin et al. 1992)

Fig. 9.24 Sangtuda Dam, Tajikistan

Sangtuda Dam (Vakhsh River, 2009) is situated on a site with limestone and gypsum/anhydrite rocks (Fig. 9.24). Dam height is 75 m and length 517 m. Due to different structural, lithological, and geomorphological properties, the density and distribution of karst forms on the right and left riverbanks are also different, the density of features on the right bank (Upper Cretaceous–Lower Paleogene) being much less than on the left bank (Lower Paleogene), where the recorded karst forms are in syngenetic calcareous breccias. Density of all karst forms in that area is 1008 per square kilometer, with the density of shafts (funnels) being 545 per square kilometer.

9.2.5 Dams and Reservoirs in Iran Gotvand Dam (Iran, Karun River) is a 178-m-high rock-fill structure with clay core. Crest length is 760 m. The maximum reservoir level is 232 m a.s.l. The reservoir length is 90 km, and storage volume is 4.5 billion m3. The dam is situated in the Bakhtiary conglomerate (Pliocene/Pleistocene) with thin gypsum veins. Large cracks observed in galleries in the right dam abutment are empty or filled with clay (Fig. 9.25). Approximately 4 km from dam, the left bank of the reservoir is in direct contact with the approximately 3-km-long and 120-m-thick Gachsaran Formation

9.2 Selected Case Studies

133

Fig. 9.25 Gotvand dam site, right bank. a Bakhtiary conglomerate with thin discontinuous gypsum veins. b Empty crack. c Crack filled with clay (Photographs P. Milanović)

(Miocene), i.e., in direct contact with a large evaporite structure that consists of a thick salt formation intercalated with anhydrite and layers of plastic clay and marl (Ambal salt structure). Each salt bed is 10–15 m thick. Edgell (1996), Milanović (2010), Barjasteh (2012), and Raeisi et al. (2013) described the Ambal salt plug as an active diapiric structure. Locally, massive and compact salt diapiric structures penetrate through the upper beds of the Gachsaran Formation. Twelve of 16 investigation boreholes intercepted cavities occupying up to 16% of borehole depth. The largest cavities are detected mostly at the borehole bottoms. The drill bit drop in some boreholes was between 7 and 22 m. Some of the caverns are deeper than the river level. Surprisingly, groundwater level in the Ambal block is 80–90 m higher than water level in the Karun River. The difference of groundwater level in neighboring piezometers in these two groundwater basins is unusually high, 15–40 m. These data indicate that hydraulic connection between river and the Ambal block, and inside the block, is negligible in spite of the fact that this evaporite block is extremely karstified (Milanović 2004–2010). Water in boreholes drilled in the salt contained on average: Na—4337 (meq/L); Cl—5000 (meq/L); SO4—31.7 (meq/L); EC—184,080 µS/cm2; and TDS—337,832 (mg/L): i.e., these are essentially saturated solutions. pH ranged between 5.4 and 7.1. Due to the very effective dissolution of the salt, there appear to be three crucial problems: leakage from the reservoir, stability of the reservoir banks, and pollution of

the water. Because this water has to be used for irrigation, its high chloride content creates many problems. In flood conditions, dissolution has been very rapid. The horizontal notches seen in Fig. 9.26 are consequences of salt dissolution during one flood with rapid river flow based on experiments (physical models) by site staff at the Gotvand dam site, and the estimated dissolution rate is approximately 10 cm/day on exposed salt cliff faces. Based on this rough dissolution estimate, it was concluded that 11.5 million ton of salt would be dissolved during the first year of reservoir operation. According to detailed hydrogeological analysis, off-site leakage through the evaporite strata is unlikely, however. To minimize the intensity of direct contact between the reservoir water and the halite outcrops, cover with different materials was tested: geomembrane, polyurethane, asphaltic materials, concrete blocks, and clay blanket. Finally, clay blankets and riprap protection were applied as covers. Large collapses were filled with clay slurry (Mahjoob et al. 2014). During the first year after filling the dissolution rate was higher than expected from the numerical models. However, in the following year the solution rate decreased. A 30-m-deep layer of dense, supersaturated water has been formed on the reservoir bottom. This water is distinctly separate from the overlying water where TDS is similar to that of the river in its previous natural conditions (see Sect. 5.5). In the upper 70 m of the reservoir, the chemistry of the water is in accordance with the standards used for irrigation. Three additional low elevation intakes (at 90, 110,

134

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.26 Gotvand Reservoir, Iran. Effects of salt dissolution on cliffs in the reservoir bank following a few hours of high flood (upper fresh scars) and a horizontal corrosion notch (the cavern-like feature) formed at the waterline over approximately ten days (Photograph P. Milanović)

and 158 m a.s.l.) have been constructed to drain away the oversaturated water (Fig. 9.27). This brine can be used in the oil industry or transported by surface canals to the sea. The 15 Khordad Dam (Iran, 1995) consists of an earth-fill dam 96 m high and a dike on the right bank. The reservoir is for water supply and irrigation. The dam itself is built on granite. However, the floor of much of the large reservoir behind it consists of karstified limestone, partly covered with alluvium and evaporites. To prevent overflow or leakage through a ridge on the right bank, a grouted dike was constructed (Fig. 9.28). In sequence downward, the geological formations beneath the dike foundation consist of: alluvium; red beds; sedimentary breccias; limestone; and the granite base. Karst features are detected in the limestone and sedimentary breccias. Locally, water pressure tests gave large values of more than 100 Lu—implying high permeability. During site

investigations, a few collapses and a number of cavities were noted and some investigated with tracer tests. According to results of chemical analyses, water pollution could be a crucial problem (see Chap. 5). After filling the reservoir (1993), the pollution (TDS, due to solution of the evaporites) increased tremendously, well above acceptable limits. To improve the water quality, the most acceptable solution is to dilute the reservoir water with other water from the Dez Reservoir (transported through a long tunnel) and with freshwater from deep wells. Khersan III Dam (Iran) is a concrete double arch structure currently under construction on the Khersan River. Dam height from the foundation is 195 m, and crest length is 421 m. The bedrock at the site (Asmari Formation) consists of limestone, marly limestone, and marlstone. A small part of the reservoir is also in direct contact with the Gachsaran Formation (marlstone, gypsum, gypsiferous marlstone, and

9.2 Selected Case Studies

135

Fig. 9.27 Gotvand Dam. The salinity stratification of the reservoir water relative to the depth, 2013–2017, and showing accesses to the water at different levels (from Mashkat et al. 2018)

Fig. 9.28 15 Khordad dam site, layout and cross section (Milanović 1989, unpublished)

marly limestone). The reservoir is separated from the downstream Einakak River by a ridge composed of Gachsaran strata. In some deep boreholes, salt deposits up to 20 m thick were recorded. This formation is intensively tectonized. The distance between the reservoir area and a spring zone in the adjoining Einakak valley is only about

800 m, with the difference in elevation being 52 m. The electrical conductivity of water in boreholes ranges between 6000 and 11,000 ls/cm. As evidence of risk is rapid dissolution of gypsum on the ridge. Figure 9.29 shows the large shaft created by water leaking from a small pipeline (diameter 2 in.) over a period of approximately 10 years.

136

Fig. 9.29 Khersan III Dam, Iran. Karst shaft in gypsum created by water dripping from a 5.1-cm pipe (Photograph P. Milanović)

Large areas of the evaporites are covered with very pervious overburden, and there are also some open outcrops of gypsum in the reservoir area that will be directly exposed to solution. After such solution starts, karstification of the gypsum and gypsiferous marl must progressively increase. In addition to leakage from the reservoir, pollution by the solute chemicals is one more potentially important problem. To prevent progressive solution and the rapid creation of solution voids and channels, construction of a cut-off wall

9

Overview of Dams and Reservoirs in Evaporites

must take priority over other waterproofing measures. The lower contour of cut-off wall should be deeper than lower groundwater level at the ridge area. The Marun Reservoir is created by the rock-fill Marun Dam (165 m high) and has a capacity of 1200  106 m3. At the dam site, there is stratified Asmari limestone dipping in the upstream direction. From a karstification point of view of great importance are: (1) a major joint system parallel to the geological structure and prominent penetrable bedding planes, and (2) A thin layer of vuggy limestone in the Middle Asmari limestone (Fig. 9.30a). Two-row grout curtains were constructed in both abutments. Prior to the reservoir impoundment, a concrete plug was inserted into the middle section of the second diversion tunnel (Fig. 9.31). During the first reservoir filling, old karst channels along the vuggy zone cut through by the second diversion tunnel were reactivated and leakage occurred. At a reservoir level of 40 m, leakage of 40 L/s was recorded and there other discharges in the second diversion tunnel and access gallery. As a consequence of washout of fill in the (now-submerged) old karst channels, leakage increased to about 10 m3/s. The largest part of the water bypassed the concrete plug and discharged from the tunnel and gallery. No new leakage springs were seen downstream in the river gorge. Speleological investigations of the washed-out cave system were organized; locally, the channel was more than 2 m in diameter (Fig. 9.30b).

Fig. 9.30 Marun Dam, Iran. a Vuggy zone in the Asmari limestone. b A major karst solution channel (Photographs P. Milanović)

9.2 Selected Case Studies

Fig. 9.31 Simplified sketch of the Marun dam site (Milanović 2004— based on Banihashemi, R., Binazadeh, K and Sheikhi, A. personal communication 1998)

Remedial works included a one-row grout curtain extension up to watertight shaly interbeds (Pabdeh Formation), and construction of concrete plugs in accessible sections of the

137

karst channel (Milanović 2004). No more seepage has been observed. According to Barjasteh (2012), nearly two-thirds of the 18-km-long Marun Reservoir is on the Gachsaran Formation (gypsum, anhydrite, marlstone, siltstone, and shale). The main problem with the Gachsaran beds is that their erosion and weathering cause high rates of siltation in the reservoir and also has some effects on the water quality. Seymareh Dam (Iran) is a concrete arch dam 135 m in height (185 m from the foundation) situated on karstified limestone. It is located in the northern limb of the regional Ravandi Anticline. A huge landslide that occurred downstream approximately 15,700 years ago has had a crucial influence on the intensity and depth of karst evolution inside the large mass of Asmari limestone. Before the slide, the local erosion base level was 40–50 m lower than at present (*600 m a.s.l.) A number of caverns have been developed along faults parallel to the anticline axis. Clearly, the entire fault zone is weak and prone to erosion. Evaporite rocks (the Gachsaran Formation) in the left reservoir bank just upstream of the dam site consist of folded marls and gypsum beds *20 m thick. From the karstification and leakage viewpoints, the gypsum can be declared as a weak zone with considerable potential losses (Fig. 9.32).

Fig. 9.32 Seymareh, Iran. Gypsum deposits in the reservoir left bank (Photographs P. Milanović)

138

An inverse thermal gradient that was discovered in some boreholes in the left abutment indicates that there is active groundwater flow 10–20 m lower than the present river bed. Further, cavities filled with clayey–silty material and in some cases with sand were detected in some boreholes at depths approximately 100 m lower than the river, i.e., about 50 m lower than the elevation of the fossil river bed; during the karst evolution of this part of the aquifer, there were siphon (phreatic looping) conditions. Due to the findings underground, the grout curtain was redesigned in direction and length. The impounding of the reservoir displayed all the complexity of dam sites in karstified rocks. For example, impounding began before the watertight sealing works were completed. A key seepage zone (“hydrogeological window” of about 10,000 m2) at the end of the left bank grout curtain was not completed. Grout mix consumption in the boreholes already grouted was between 1.0 and 30 tons/m’. Immediately after impounding began, leakage increased rapidly. After the reservoir level had risen up to half of the design depth, leakage had increased from 4 m3/s in 2013 to more than 8 m3/s. Kangir (Tang-E-Shemiran) Dam (Ilam, Iran, 1991– 2013) is an earth-fill structure with clay core, high 42 m, with a crest length of 742 m. The dam site is situated in a gorge on the Kangir River. Storage capacity is about 20 million cubic meters. The main purpose of reservoir is water supply and irrigation. The dam and reservoir area are located within a typical Zagros Mountains folded structure with outcrops of the Gachsaran Formation and karstified limestone of the Asmari Formation (Fig. 9.33). At the dam site, the Asmari limestone is intensely karstified. Some thick limestone layers have dense vuggy porosity and locally spongy structure (Fig. 9.34a). The area covered by the reservoir is composed mostly (95%) of Gachsaran rocks: anhydrite, gypsum, halite, red marls, mudstone, and thin layers of limestones and sandstones (Milanović 1999). In the gypsum on the reservoir floor, there are a few ponors (swallow holes), some of which have been plugged by local people (Fig. 9.34b). About half a kilometer upstream of the dam, an active ponor with permanent water inflow was recorded in an irrigation canal (Fig. 9.35). According to Rezaei et al. (2017), the thickness of the formation containing evaporites is around 60 m. Geotechnical measures to achieve a watertight dam site and reservoir consist of: a grout curtain at the dam (Asmari limestone), and a surface blanket and cut-off wall in the gypsum (Gachsaran formation). Two crucial problems have appeared during the past five years of reservoir operation: first—decreases of the Kangir River flow from an average Q = 1.53 m3/s to Q = 0.68 m3/ s, followed by decrease of the groundwater table upstream; and second—deterioration of water quality due to dissolution of gypsum and halite.

9

Overview of Dams and Reservoirs in Evaporites

Borehole data and monitoring results suggest that there are two separate aquifers. An upper one in the Gachsaran Formation that is much higher than a lower one in the karstified Asmari limestone (Karimi and Pakzad 2009). A prolonged dry period and intensive overexploitation of groundwater in upstream parts of the catchment are, most probably, the reasons for the temporary drying up of the Kangir River and some springs. Seepage water from the Kangir Reservoir catchment flows mostly to the Sorkhejo Spring (Fig. 9.32, S6 and S9) downstream of the dam. Part of the groundwater from the Asmari anticline flows directly to a downstream spring zone along the Zarhen-Siaghel fault zone; i.e., it never discharges into the reservoir. As is common in karstified rocks, the orographic and hydrogeological (real) catchment areas differ from each other. Water samples were collected at more than 40 boreholes and springs for hydrochemical analysis. The groundwater EC values vary between approximately 400 and 2450 lS cm−1. The highest groundwater EC value is in reservoir area (chloride water type). Results of the chemical analyses show that, due to the increasing water level, the flooded area of reservoir floor and banks became larger and water quality deteriorated. Polluted water in the Gachsaran Formation has a deleterious effect on the Asmari limestone karst aquifer also. Because the purpose of the reservoir is to supply about 70,000 inhabitants with potable water and the irrigation of 2500 ha of arable land, deterioration of water quality becomes a serious problem. One of the conclusions of Rezaei et al. (2017) is that in the case of construction of dams and reservoirs in evaporites, the hydrogeological properties, regional and local, have to be determined in much detail before any construction is begun. Jareh Dam is a concrete-faced earth-fill dam located within the Pleistocene Bakhtiary conglomerates. Due to the calcareous or argillaceous matrix, the Bakhtiary conglomerates are impervious or of low permeability. The largest part of the reservoir is in the Rud-e Zard River valley, which is floored with beds of the Gachsaran evaporites: anhydrite, red and gray marl, salt and sandy limestone. Potential leakage from the reservoir is thus of concern when reservoir water levels rise to the design maximum (Barjasteh 2012). Gheisaragh Dam (2005) is an earth-fill structure in the East Azerbaijan Province. The dam is 18 m high, 1000 m in length, with a reservoir volume 2.6  106 m3. The purpose is for irrigation. The rock foundations at the dam site area include claystone and marl with gypsum pockets and veins of various thicknesses (Fig. 9.36). During the reservoir impounding and later, leakage through the foundation and abutments was recorded. Due to piping processes, typical piping outlets (point outlets) and swampy discharges were created downstream (Moradi and Abbasnejad 2011). Chemical analysis of the main leakage zones showed increase of SO4 up to 49.6 meq/L and TDS up to

9.2 Selected Case Studies

139

Fig. 9.33 Kangir Dam and Reservoir, Iran. Hydrogeological map of the study area. A-B cross section perpendicular to the structure direction (from Rezaei et al. 2017)

3430 mg/L. The content of Mg++, Ca++, and SO4—in discharge water—is six to twelve times greater than in the reservoir water. These results clearly show that there is active dissolution. To reduce the risk of failure, different remedial measures were considered, including a grout curtain, a cut-off wall from the dam crest, a concrete slab on the upstream face coupled with a cut-off wall, or a clay blanket on the reservoir floor. The chief factors for a final choice were geological, technical, economic, and the amount of time required for construction. The chosen solution was a cement–bentonite cut-off wall along the heel of the dam. To prevent dissolution and piping, the cut-off wall was required

to cross through the formation containing gypsum. The cut-off structure is 400 m long and 15 m deep. Khoda Afarin and Gyz-Galdasi dams are located on the Aras River, the border between Iran and Azerbaijan. The Khoda Afarin Dam is earth-fill embankment structure 64 m high and 380 m in length (Fig. 9.37). Construction started in 1988 and was completed in 2008. The dam foundation consists of gypsiferous argillites and sandstones. To prevent dissolution of the gypsum, complex protective structures (a gallery in the dam foundation) and treatment measures were applied (Demyanova 1986).

140

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.34 Karstified rocks at Kangir dam and reservoir area, Iran. a Spongy Asmari limestone and b sinkhole in gypsum on the reservoir bottom (Photographs P. Milanović)

Fig. 9.35 Tang-E-Shemiran Reservoir with swallow hole (ponor) at bottom (Photo P. Milanović)

9.2 Selected Case Studies

141

Fig. 9.36 Gheisaragh dam site, Iran. Geological cross section along the dam axis (from Moradi and Abbasnejad 2011)

Fig. 9.37 Layout sketch of Khoda Afarin Dam. From ‘A General View of Iranian Dams’ (from IRCOLD publication, 1993)

9.2.6 Dams and Reservoirs in Iraq The Mosul Dam (Iraq) is located on the River Tigris approximately 60 km upstream of Mosul, the second city of Iraq. The construction work started in January 1981 and finished in July 1986. It is a multipurpose project for irrigation, flood control, and hydropower generation. It is an earth-fill dam, 113 m in height, 3.65 km in length, 10 m wide at its crest. The design surface area of the reservoir at the beginning of the undertaking was 380 km2, with a storage capacity of 11.11 billion m3. Dissolution processes and leakage were observed and monitored during and after the reservoir first filling. In spite of massive remedial works over the past few decades, dissolution processes are still active. A number of articles have analyzed the Mosul

problems. The most comprehensive and valuable data are presented by: Anagnosti (1987); Guzina et al. (1991), and in a series of eight linked articles “Mystery of Mosul Dam the most Dangerous Dam in the World” by Adamo et al. (2015). The dam is built on bedrocks of the 250-m-thick Fatha Formation (Middle Miocene), which consists of cyclically deposited gypsum, gypsum breccia layers, and anhydrite beds, alternating with soft marl layers and cavernous limestones. The thickness of the gypsum breccia layers ranges between 8 and 16 m. Initial permeability in the massive gypsum and anhydrite, measured by water pressure tests, was close to zero. High initial permeabilities, ranging from 10 to 50 Lu units or more, were measured in the fissured limestones and brecciated strata. The highest measured permeability was found at the boundaries of, or within, the

142

tectonized and brecciated gypsum and anhydrite beds. Depths for the grout curtain were selected mainly on the basis of initial permeabilities. The dam design was based on the assumption that impermeability of the foundations and prevention of any dissolution of gypsum and anhydrite would be achieved by implementing a deep grout curtain strengthened by a grout blanket under the clay core (Guzina et al. 1991). The grouting blanket (consolidation grouting) consists of 20 rows. Depth of the internal rows is 20 m and of the external, 10 m. The grout curtain consists of four sections: – Left bank, 1560 m long, 1 row, from ground surface; – Saddle dam section, 1152 m long, 2 rows. The second row was added after the first springs appeared; – Main Dam section, 2379 m long, 3 vertical rows, 80– 100 m deep, plus 2 inclined rows 25 m deep, one upstream, one downstream of the vertical rows; – Right bank, 408 m long, 2 rows close to dam, and one row further toward the end section. Depth of the first of the two-row curtain varies from 125 to 135 m, and depth of second row from 60 to 90 m. For grout curtain construction, 732,000 m3 of grout mix was used. However, during excavation of the dam foundations, a number of subvertical ruptures were discovered (Guzina et al. 1991). They were intensely karstified, even in the foundation rocks, and caused intensive gypsum and anhydrite dissolution and seepage. This has created numerous problems during the construction, the impounding and subsequent operation of the dam.

Fig. 9.38 Mosul Dam, Iraq. Layout with seepage measurement locations (from Guzina et al. 1991)

9

Overview of Dams and Reservoirs in Evaporites

The first partial impounding of the reservoir began in February 1986. Shortly after it started, seepages were observed through a 1000-m-long section of the dam foundations (Fig. 9.38). By the month of June, 1986, leakage through the dam area had increased to 1400 l/s. The approximate leakage paths, at that time, were located at average depths between 60 and 70 m. The increase was a consequence of enlargement of flow routes due to intensive dissolution of the gypsum and anhydrite. By measuring the total dissolved solids, average amounts of gypsum solution of 42–80 t/day were found (Guzina et al. 1991). During the six-month observation period, 13,000 t of gypsum and anhydrite were dissolved beneath the Mosul foundations (Fig. 9.39). Chemical analyses of seepage water showed that increases in dissolved solids were consequences of fluctuations in the reservoir level, i.e., consequences of the varying pressure and consequent responses in the velocity of groundwater flow (Anagnosti 1987; Fig. 9.40). As consequences of such rapid dissolution, large voids were created, mostly between the evaporites and the less soluble (limestone) or insoluble rocks. However, the limestone, with its well-developed karst features, is considered to be hazardous by many authors as well, so that its role in the entire process of rock mass degradation should not be underestimated. Locally, in tectonically disturbed places, there were surface settlement and formation of solution collapses. Over the period, 1992–1998, four collapses formed approximately 800 m downstream, near the regulating pool, and later (2003–2005) in different parts of the dam site (see Sect. 5.2).

9.2 Selected Case Studies

Fig. 9.39 Mosul Dam, Iraq. Cross section perpendicular to the axis of the dam. (1) Cofferdam; (2) deep grout curtain; (3) piezometer; (4) river alluvium; (5) marly breccia; (6) gypsum/gypsum breccia/anhydrite;

Fig. 9.40 Mosul Dam, Iraq. Relationship between reservoir levels and dissolved solids in the seepage water (from Anagnosti 1987)

The quantity of different grout mixes consumed between 1986 and 2015 reached 95,657.43 tons (Adamo et al. 2015). According to these authors, “massive grouting did not prevent the dissolution of gypsum and anhydrite altogether and it seems that it is not likely to do so in the future.” However, the seepage losses due to the dissolution of gypsum and anhydrite beds under the foundations of the dam could not be stopped. This caused great concern about the possibility of the dam failing. The capital of Iraq, Baghdad (638 km downstream of Mosul Dam), and the very large populated area in between are endangered. Due to the

143

(7) marl/marly limestone; (8) chalky series (vuggy limestone); (9) limestone (hard and dense) (from Guzina et al. 1991)

adverse geological properties and intensive dissolution in the foundations, Mosul Dam has been declared to be one of the most dangerous dams in the world. Mosul Dam, Intake Structure. The intake structure for Mosul Dam including a 1000-m-long section of the tailrace tunnel is constructed in gypsiferous rocks—gypsum and marl occasionally interstratified by thicker marl units. The intake structure is situated between the upstream coffer dam and the main dam body (Fig. 9.41). The sealing technology was explained in detail by Guzina (1992). When excavation of the foundation pit reached 10 m below the river level, seepage inflow of 150 L/s was measured, with the content of dissolved sulfates reaching 1600 mg/L. When the inflow increased to 1750 L/s, the foundation pit was flooded. Solution conduits discharging water into the pit increased from about one centimeter in diameter to about one meter and the sulfate concentration dropped to 1100 mg/L. A large sinkhole was created in the right-hand riverbank 150 downstream. Immediate infilling of this sinkhole did have not any positive effect on the karstification. Because the extensive grouting campaign was without result, the grouting work was suspended. About 10,000 tons of dry grouting material (mostly cement) had been used. Sand was added only when the consumption of cement per hole exceeded 18 tons. After the cut-off wall was completed, its bottom was made waterproof by a three-m-thick grouting blanket created by 126 additional grouting holes. For consolidation of the rock mass beneath the cut-off box, 10,000 m of boreholes was drilled and 1200 tons of dry mass was used. Nine months were needed to construct the cut-off wall box. Due to problems with water ingress in the gypsiferous rocks, the construction time for intake was extended by 19 months, with an additional drilling of 27,000 m of holes,

144

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.41 Mosul Dam, intake structure. Box around the intake structure constructed by application of cut-off walls (from Guzina 1992)

consumption of 14,000 tons of dry components, and construction of the cut-off wall box. Construction of cut-off wall is explained in paragraph 6.1.2.2. Haditha Dam is an earth-fill dam with an asphaltic concrete core, located on the Euphrates River and constructed in 1982. It is 9064 m in length and 57 m high (Demyanova 1986; Lykoshin et al. 1992; Kislev 2004). The

dam site is on Oligocene–Miocene carbonate rocks (limestone and dolomite with intercalations of breccia, conglomerate, shale, and clay). Karstic features are well developed at the surface and underground, particularly in the right bank (Fig. 9.42). On the left bank, about two km from the river, these formations are covered with clayey–carbonate sediments

9.2 Selected Case Studies

145

Fig. 9.42 Haditha Dam, Iraq. Cross section of the dam site area. V–VII—mostly limestone and limestone breccias; VIII–XIV—clayey-carbonate rocks with gypsum and anhydrite (from Kislev 2004)

containing deposits of gypsum and anhydrite. The evaporite beds are 7–8 m in thickness, and the proportion of gypsum in this cover formation in the dam area is 50–90%. Diameters of sinkhole features in the gypsum average 2 m and can be seen to have depths of 6–7 m. A number of different karst features are developed in the gypsum area. Karstification is closely related to the density and location of discontinuities exposed at the surface. Gypsum is karstified at the same intensity at the surface and underground. The thickness of carbonate rocks beneath the gypsiferous strata is from 17 to 24 m. More than 120 sinkholes (about 100 of them in the right bank) were found in a broad area around the Haditha site (Fig. 9.43). Sizes of surface openings were up to 150  70 m, with depths of 3–40 m (Kondratiev 1979). A number of caverns and horizontal solution channels were discovered, mostly a few meters above the water table. The largest caves in the dam site area were 60 m wide, 5 m high, and 50 m long. Investigation boreholes crossed 50 cavities 0.1–1.7 m in height developed along joints. The principal breccia layer had a clay–shale matrix and was accepted as being sufficiently impermeable to serve as the base for emplacement of a watertight structure in the right bank. It was necessary to protect the dam from the negative impacts of the extensive karstification on the bearing capacity of the foundations and the watertight capacity of the reservoir. When preparing the foundations, there were additional removals of gypsum in different places and individual open cavities were filled by concrete.

Fig. 9.43 Haditha Dam, Iraq. a Simplified geological layout and b cross section A–A′ (Kondratiev 1979, from Lykoshin et al. 1992)

146

Protective measures include the installation of a gallery at the base of the dam along all of the alignment, and a two-row, cement-based grout curtain. Particular attention was paid to the end section of the left dam bank. There were tracer tests along the edge of gypsum outcrops on the banks. An asphalt concrete cut-off wall was constructed in the central dolomite unit in the dam foundation. The base of the cut-off wall was reinforced by a concrete plate with a grouting “tooth” below it. Darbandikhan Dam is embankment rock-fill structure on the Diyala River, 128 m high and 445 m long. The dam was constructed between 1956 and 1961. Schistose and marly rocks in the foundations contain gypsum, but the permeability was relatively low. After reservoir filling, due to the satisfactory performance of the grout curtain installed, there was no leakage. However, in 1967 the crest of the dam settled due to slope failure.

Fig. 9.44 Huoshipo Dam, China. a Layout of dam site area and part of reservoir. b Cross section through from the collapse zone in the reservoir to the sand spring (from Lu and Cooper 1996)

9

Overview of Dams and Reservoirs in Evaporites

9.2.7 Some Other Dams in Evaporites Huoshipo Reservoir (China, Guizhou Province) is constructed on a karstified limestone with 48 gypsum interbeds (Hu 1988). The height of this earth-fill dam is 23 m, and length is 200 m. The foundation consists of karstified, mixed gypsum and limestone in the Triassic Guanling Formation (Fig. 9.44). No investigation of underlying geology was undertaken before the dam was constructed (Lu and Cooper 1996). During the grouting program, a cavity 60–70 cm high was found. During the first filling, there was considerable leakage (237 L/s), the water appearing approximately 400 m downstream as an artesian outflow at sand spring. The rate of discharge correlated with the hydraulic head in the reservoir water (Fig. 9.45) and the water contained sand.

9.2 Selected Case Studies

147

Fig. 9.46 Huoshipo Reservoir, China. Recession curve of sand spring (from Hu 1988)

Fig. 9.45 Huoshipo Reservoir seepage. Correlation between the amount of rainfall, reservoir level, and sand spring discharge (Information from Jiang Defu, from Lu and Cooper 1996)

A number of fresh collapses opened in the reservoir bottom; evidently, laminar flow seepage had been replaced by turbulent flow. Any time after reservoir dewatering, there were new collapses about 100 m upstream of the dam. According to Hu (1988), the recession curve of sand spring showed an upward convex form when conduit flow took place as a result of the sinkhole collapse in the reservoir bottom (Fig. 9.46). The collapses were filled with rock and covered with plastic membranes and clay. According to information from Jiang Defu (from Lu and Cooper 1996), the water in the dam contains 29–38 mg/L of sulfate, but at the downstream spring it contains about 200 mg/L. Because the reservoir capacity is only 4.7 million cubic meters, the reported leakage was declared to be too high. With additional sealing work (grout curtain, plastic membranes,

clay cover), seepage was reduced to 80 L/s; however, the dissolution has not been completely eliminated. Mingachevir Dam on the Kura River (Azerbaijan) is a concrete structure, 80 m high and 1550 m long (Fig. 9.47). The valley of the Kura River is situated in clastic deposits of the Absheron stage that have deep, exposed weathering cracks with crystalline precipitates and veins filled by secondary gypsum (Fig. 9.48). In order to increase the stability of the slopes, they were drained by a number of near-horizontal wells drilled from the dam excavations into the clay layers. To prevent degradation of the sandstones, these were protected on both sides by a special cement grout curtain in the drained area. The local groundwater is highly aggressive with respect to the cements found in all cracks, closed or open. To prevent dissolution in the foundation rocks, an anti-filtration grout curtain was designed. A bitumen emulsion was used to construct an impervious underground screen, applying it as a slurry. However, during operation of the dam the bitumen was washed out in some sections, and as a result the bitumen curtain was re-grouted using a cement-based mixture. Farkhad Dam (Syr Darya River, Uzbekistan) is a concrete gravity structure (Fig. 9.49). Construction started in 1942 and operation in 1948 (Rodevich 1989). The hydraulic unit consists of a pressure basin with a frontal wall and outlets in five sections, turbine conduits, a station building, and an outflow canal. At the base of the wall, there are two anti-filtration structures. The foundations are a sequence of weakly permeable Middle Quaternary loess-like soils interbedded with lenses of gravel and pebbles, with a total thickness of 40 m. The dense loams contain fissures and veins with mineralized waters (up to 20 g/L). The soils have significant carbonate

148

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.47 Satellite image of Mingachevir dam site and part of the reservoir (from Google Earth)

Fig. 9.48 Geological section along the axis of the Mingechaur HPP. Quaternary deposits: 1—alluvial loams, gravel pebbles and sand, deluvial and proluvial–deluvial loams; 2—fragments of bedrock with loamy aggregate (landslide soils). Absheron deposits: 3

—silty sandstone; 4—sandstone–clayey silt; 5—silty clay; 6—intermittent strong sandstones, plus clays, aleurites; 7—intermittent weak sandstones, plus clays, silts; 8—tectonic breccia (from Geology and Dams 1959, in Russian)

and sulfate content: The carbonate averages 2% but can be as great as 38%; gypsum—2–4%, maximum 16; and soluble salts between 2 and 0.7%. Almost from the beginning of operations, there was rapid sedimentation in the reservoir (30–40 mm/year) and a high decrease of salinity including water loss in foundation of the dam. Observations of the hydrochemical regime were begun to monitor the chemical degradation. It was concluded that there was strong mineralization, with conditions favorable for deformation to occur in the foundations due to intensive loss of the salts. This led to the creation of a gap between the bottom of the wall and the settled soil, allowing for the possibility of a failure emergency arising.

Toktogul Dam, Kyrgyzstan, is a concrete gravity structure, 215 m high and 295.5 m long. It is a part of the Narun— Syr Darya cascades, constructed in 1960–1970. At the end of the reservoir, a few meters of salt, between clayey layers, are exposed to the reservoir water along a frontage of 7.5 km. Mineralization at a leakage spring is up to 257 g/L, implying that up to 130 kg of salt discharges from this spring every hour. A number of fresh sinkholes have appeared in the surrounding reservoir banks (Lykoshin et al. (1992). Erevan Dam on the Hrazdan River in Armenia is a rock-fill structure, 20 m in height, creating a reservoir depth of 18 m. It was constructed in 1962. The foundation consists of gypsiferous clay and shaly limestone with a high content

9.2 Selected Case Studies

149

Fig. 9.49 Farkhad Dam, Uzbekistan

Fig. 9.50 Erevan Dam. Armenia. Cross section along the dam axis (from Tolkonnikov 1974)

of gypsum (Fig. 9.50). The river valley at the dam site consists of different lithologic formations of Miocene and Upper Pliocene age, succeeded by Quaternary effusive rocks, and later alluvial, diluvial, and colluvial deposits. The Miocene strata consist of lower gypsiferous and upper claystone sections. In the deepest horizons, thickness of the gypsum deposits is up to 10–15 m and total thickness of gypsiferous deposits is 300 m. At the dam site (foundation area), the upper gypsiferous beds are at the level of the river bed (Tolkonnikov 1974); Lykoshin et al. 1992). The preventive waterproofing structure consists of a deep cut-off wall in the dam foundations. This is connected to a cement-based grout curtain along the axis of the dam and a lateral grout curtain in the well-jointed basalt of the right bank. In the dam section at river level, there is 6 m of crushed gypsum that dips steeply beneath the right bank. The

cut-off wall in that section is taken to a depth of 30–40 m. It serves as a positive cut-off because its base is dug tight into the Miocene clays. This has prevented seepage beneath the cut-off wall and in the right bank. After 30 years of dam operation, there have been no indications of possible disturbance of dam stability and no leakage from the reservoir has been recorded. Tbilisi Dam (Georgia) (Demyanova 1986); Lykoshin 1992; Maksimovich 2006). To improve reservoir capacity threatened by sedimentation, the construction of four concrete dams was necessary. The foundations were gypsiferous clayey sandstones and shales incorporating thin gypsum layers. Percentages of gypsum in an upper weathering zone are up to 7%. In the design phase, the presence of gypsiferous rocks in the dam foundation was not considered to be a problem. Cut-off grouting structure

150

9

Overview of Dams and Reservoirs in Evaporites

Fig. 9.51 Hessigheim Dam, Germany. Location map of the weir, locks, powerhouse, and remedial measures (from Wittke and Hermening 1997)

beneath the dam was designed to prevent destructive seepage, however. After reservoir impoundment began, leakage and local damage to the dam were observed, endangering its stability. Chemical analysis of the seepage water (1–2 L/s) shows increase of sulfates from 0.34 g/L up to 2.6 g/L. After the reservoir level was lowered, a row of boreholes was drilled along the dam and a grout curtain deep 20–30 m was constructed. However, chemical analysis of seepage water confirmed that significant solution continued. Ten years after these additional measures were executed, the reservoir level had risen to the design level. After two years of operation at this level, the leakage had increased by eight times. After 27 years of operation, two new discharge points were recorded. After 33 years of reservoir operation, the new discharge points with concentrated discharge become active. A leakage intensity theoretically expected after 100 years occurred after only 30 years in reality. Hessigheim Dam on the Neckar River, Germany, was constructed between 1949 and 1958 (Wittke and Hermening 1997). Storage height is 6.20 m. The structure consists of three main components: a weir in the river, a power plant on the left bank, and a pair of locks on the right bank, 12 m wide. The main dam is 22.6 m in width (Fig. 9.51). At the dam site, there are a few meters of alluvium (3– 6 m sand, gravel) deposited over a clay–silt with gypsum fragments that is 1.6–9.8 m in thickness. Beneath this clay are two sequences of Middle Triassic gypsum separated by dolomitic limestone beds (“dolomite horizons”). Salt laminae were also detected. Before and during the foundation excavations, sinkholes, subsidence collapses, and caverns

were observed the latter at the contact between the proposed structure and the soil, in the residual soil and in gypsum. Average height of the caverns ranged between 1 and 2 m. After these discoveries, the original design for a shallow sheet pile cut-off wall was extended down to the contact between the residual soil and the upper gypsum (Fig. 9.52). During the weir operation, from 1950 and 1984, new sinkholes were recorded, some of them as much as 8 m in diameter. Remediation works were undertaken between 1988 and 1994. Beneath the foundations, two groundwater levels, separated by the residual soil, were detected. Rates of flow in seepage water in the lower residual soil and upper gypsum were few decimeters per second. Content of sulfates in the groundwater was 1067 mg/L and chloride about 60 mg/L. Caverns beneath the foundations of the locks, weir, and powerhouse were filled by a thick grout mix (W/C = 0.45 and stabilizer) through boreholes spaced 4 m apart. The boreholes crossed the “dolomite horizons” to search for any likely cavern development beneath. The grout curtain is constructed on the upstream side of the weir, along the power plant foundations, and between the lower gates and locks. Vertical and inclined grouting holes, spaced 1 m, were added to intersect cavities and vertical joints along the vertical grout curtain. Drilling was organized from floating pontoons. The role of the grout curtain was to connect the toe of the cut-off wall to effectively impermeable rock deeper below. A thinner grout mix (W/C = 0.8) was used for grouting the bedrock mass. In total, 10,600 t of cement was used—6600 t for cavern filling. “The efficiency of the emplaced grout curtain is assessed to last for a period of 30–40 years” (Wittke and Hermening 1997).

9.2 Selected Case Studies

151

Fig. 9.52 Hessigheim Dam. a Stratigraphical sequences and remedial measures underneath the weir. b Drilling for construction of the grout curtain (from Wittke and Hermening 1997)

Fig. 9.53 Foum El Gherza, Algeria. Geological properties of catchment area (from Benfetta et al. 2017)

Foum El Gherza Dam in Algeria is a 65-m-high concrete arch dam situated in a 200-m-thick Cretaceous limestone lying over marly shale (Fig. 9.53). At Fig. (9.54) is presented layout of grout curtain and monitoring piesometric system.

Reservoir volume is 47 million m3. Limestone formations at the dam site are in direct contact with marine gypsum that is observed in outcrop at a few places in the catchment area. The dam was constructed in 1952, and after the first filling, a

152

Fig. 9.54 Foum El Gherza Dam, Algeria. Simplified layout of dam structure and grout curtain (from Benfetta et al. 2017)

9

Overview of Dams and Reservoirs in Evaporites

number of small new seepages were noted in both banks downstream of it (Hocini and Moulla 2005: Benfetta et al. 2017). Maximum water losses of 20.7 Mm3 from the reservoir were recorded during the 1981/82 season. Collapses at the surface were also reported. A program of tracer tests and monitoring of conductivity and temperature in piezometric boreholes were begun. Results of the tracer tests showed that seepage into the reservoir banks needed only two days to appear at the downstream right bank and one week at the left bank. Examples of thermal measurements in sample boreholes are presented together with conductivity graphs (Fig. 9.55). All of the thermographs showed an inverse temperature

Fig. 9.55 Foum El Gherza Dam, Algeria. Schematic layout and temperature/conductivity graphs in piezometers S24 and S26, by depth (from Hocini and Moulla 2005)

9.2 Selected Case Studies

gradient, i.e., a negative thermal anomaly in response to a concentrated zone of comparatively cool underground flow. It is clear that the zone of groundwater leakage is much deeper downstream of the grout curtain than upstream. A sharp increase in conductivity from 1000 to almost 2000 µS is, most probably, consequence of dissolution in the sections with gypsum content. Since the first filling was completed in 2002, the seepage rate has been mostly less than 150 L/s. In a separate investigation, conductivity was found to have increased to close to 2000 µS. This indicates the minor influence of nearby gypsum formations. However, “the mineralization of underground waters of Foum El Gherza Dam which is generally raised, presents a big risk of salinization and water pollution” (Benfetta et al. 2017). Jedra Dam (Algeria) is 60-m-high dam. In the foundations, there are cracks filled with gypsum in the left dam abutment. Laboratory experiments were undertaken to estimate the potential risks of solution. Water was allowed to flow over plate-shaped samples of gypsum taken from the joints for 24 h. It was found that the very low natural velocities cannot substantially influence (?) the state of the gypsum; after initial dissolution, the boundary layer

153

establishes an equilibrium between solution and precipitation on the surface of the plate. The field measurements show that discharge capacity of some permanent springs was not changed significantly (Anagnosti 1987). Joumine Dam (Tunisia) is a 57-m-high embankment structure constructed in 1983. The dam site consists of Cretaceous limestone underlain by Triassic formations. A gypsum bed was detected at the site. During the first impounding, seepage began at 0.5 m3/s and, when the reservoir reached its maximum level, increased to about 1.0 m3/s. After a few additional stages of grouting, this was reduced to 90 L/s (2013) but with an increasing amount of sulfates in the seepage water (Sari 2013). Kavşak Dam Site (Turkey). The concrete-faced rock-fill dam rises 88 m from its foundations (73 m above the river channel level). The crest length is 186 m and the reservoir volume V = 30 million m3. More than 30 years of intensive investigations and analyses were undertaken before dam construction started. It was completed in 2013. From the general geologic and hydrogeologic viewpoint, the dam site is located in a Permian limestone which is well known to be prone to karstification (Fig. 9.56). Below the limestone, thick Carboniferous shale plays the role of a

Fig. 9.56 Kavşak dam site, Turkey. At left corner—cavern in the right bank is situated above elevation of dam crest Photographs P. Milanović

154

basal hydrogeological barrier. The thin transition zone between limestone and shale consists of dolomite and quartzite. A number of cavities are developed along fault zones (tectonic breccia) or subvertical joints. According to borehole data, all caverns were naturally plugged with clayey/sandy material that included some gravel and limestone breakdown. The largest karst feature at the surface is a subvertical shaft in the left bank of the dam at an elevation of *363 m a.s.l. (40 m above the reservoir level). Its depth is not known. An attempt to plug the shaft by filling with cement mix (undertaken before 2009) was not successful in spite of the many tons of cement mix poured in. An additional problem is the presence of gypsum beneath the dam foundations that was recorded in three downstream boreholes. Maximum thickness was 30 m (between elevations of 167 and 197 m a.s.l.). This is substantially below the river level and close to the base of the grout curtain. The presence of gypsum, despite being relatively deep, requires very careful and permanent groundwater level monitoring and water chemical analyses of nearby springs, particularly any new ones that appear during reservoir operation. Tannur Dam, Jordan, a RCC gravity structure, 69 m high and 250 m long, situated in Wadi al-Hasa, was built in 2002. Purpose of water storage is water supply and irrigation. Local geological formations range from Upper Cambrian to Quaternary. A large proportion of the dam and reservoir site consists of gray marl, calcareous siltstone, mudstone with gypsum layers, and dolomitic limestone; the aggregate thickness of this unit is about 20 m (Gouzadeh and Al-Shabatat 2013). During investigations and geochemical analyses, some indication of karstification was found. As a preventive measure, a two-row cement-based grout curtain 55 m deep was constructed. Concentration of salinity of seepage water through the dam site is different at three discharge points: 1300–2800 µS cm−1 (reservoir and the right side of the central gallery); 4400 µS cm−1 cm (left side of central gallery); and 8500 µS cm−1 (right exist adit) (El-Naqa and Al Kuisi 2004). In the Lisan Peninsula in the Dead Sea, Jordan–Israel, a dike 12 kilometer long was built with the aim of creating a 95  106 m3 evaporation pond for potash salt production. Due to the rapid modern lowering of the level of the sea and the consequent water table decrease, the pond was heavily impacted by development of sinkholes. The original dike trend was abandoned and a new alignment selected. Subsidence of an area approximately 600 m across is compromising the integrity of the new dike and leading to the loss of a significant proportion of the planned evaporation pond (Closson and Karaki 2015). Poechos Dam, Peru, is the key structure in the large Chira-Piura multipurpose water management project. This

9

Overview of Dams and Reservoirs in Evaporites

embankment dam with a clay core is 48 m in height and 9500 m long. The presence of swelling bentonite and secondary gypsum deposited locally in joints appears to be the most important questions. Very detailed analysis of the gypsum at the site and in a laboratory was arranged. The final conclusion was that the gypsum will not endanger the structure and water tightness of the reservoir.

References Adamo, N., N. Al-Ansari, E.I. Issa, V.K. Sissakian, and S. Knutson. 2015. Mystery of mosul dam the most dangerous dam in the world: Karstification and sinkholes. Journal of Earth Sciences and Geotechnical Engineering 5 (3): 33–45. Anagnosti, P. 1987. Prediction and control of seepage in soluble grounds. In IX European Conference. Dublin: International Society of Soil Mechanics and Foundation Engineering. Barjasteh, A. 2012. Salt tectonics impact on dam construction in Khuzestan Province of Iran. Kyoto, Japan: International Symposium on Dams for a changing world. Benfetta, H., A. Ouadja, N. Hocini, S. Boudina, and A. Moulla. 2017. Updating and Improving the Study of Water Leaks in the Algerian Dam of Foum El Gherza. Manuscript. Closson, D., and N.A. Karaki. 2015. Earthen dike leakage at the Dead Sea. In Engineering Geology for Society and Territory, vol. 5, ed. G. Lollino, A. Manconi, F. Guzzetti, M. Culshaw, P. Bobrowsky, F. Luino, 461–464. Dordrecht, The Netherlands: Springer. Cooper, A.H., and R.C. Calow. 1998. Avoiding gypsum geohazards: Guidance for planning and construction. British geological survey. Technical report WC/98/5 overseas geological series. Cox, E.R. 1967. Geology and Hydrology Between Lake McMillan and Carlsbad Springs, Eddy County, New Mexico: Geological Survey Water-Supply Paper 1898. Washington. Craft, D., and R. Pearson. 2002. Horsetooth Seepage Chemistry: Analysis of the Seepage Water Chemistry before and during Safety of Dams Modification Construction at Horsetooth Dam, Colorado-Big Thompson Project, Bureau of Reclamation, Technical Memorandum D-8290-02-001, Ft. Collins, Colorado. Demyanova, E.A. 1986. Dissolution and transport of gypsum by filtration flow beneath dam foundation. Engineering Geology 6: 23– 33. (in Russian). El-Naqa, Ali, and M. Al Kuisi. 2004. Hydrochemical modelling of the water seepages through Tannur Dam, southern Jordan. Environmental Geology 45 (8): 1087–1100. Geology and Dams. 1959. Moscow; Leningrad, Gosenergoizdat, vol. 1. p. 182 (In Russian). Gorbunova, K.A., N.G. Maximovich, V.P. Kostarev, and V.N. Andreichuk. 1991. Tectogenic impact on the Karst in Perm Region, Newsletter, UNESCO, IGCP Project 299, 85. China: Guilin. Gouzadeh, M., and A. Al-Shabatat. 2013. Geological and geotechnical properties of soil materials at Tunnar dam, Wadi Al Hasa, South Jordan. Journal of Taibah University for Sciences. www.elsevier. com/locate/jtusci. Gunnar, B. 1965. Anhidrite and Gypsum Problems in Engineering Geology. American Geological Institute, vol. 7. Gutierrez, F., M. Desir, and M. Gutierrez. 2003. Causes of the catastrophic failure of an earth dam built on gypsiferous alluvium and dispersive clays (Altorricon, Huesca Province, NE Spain). Environmental Geology 43: 842–851. Gutierrez, F., M. Mozafari, D. Carbonel, R. Gomez, and E. Raeisi. 2015. Leakage problems built on evaporates. The case of La Loteta

References Dam (NE Spain), a reservoir in a large karstic depression generated by interstratal salt dissolution. Engineering Geology 185: 139–154. Guzina, B.J., M. Sarić, and N. Petrović. 1991. Seepage and Dissolution at Foundations of Dam During the First Impounding of the Reservoir, 1459. Vienne: Commission Internationale Des Grandes Barages. Guzina, B.J. 1992. Power water ingress through Gypsiferrous Rock into foundations during excavation for a pumped storage scheme. Tunneling and Underground Space Technology 7 (2): 141–144. Hocini, N., and A.S. Moulla. 2005. Detection of water leakage in Foum el-Gherza Dam (Algeria). In Ninth International Water Technology Conference, 581–589. IWTC9, Sharm El-Sheikh, Egypt. Hu, Wuzhou. 1988. A study on the formation of Triassic “gypsum-dissolved strata” in Guizhou Province and the seepage prevention for reservoirs. In Proceedings 21st IAH Conference Karst Hydrogeology and Karst Environmental Protection, ed. D. Yuan, vol. 2. IAH-AISH Publication. Guilin, China. James, A.N., and A.R.R. Lupton. 1978. Gypsum and anhydrite in foundations of hydraulic structures. Geotechnique 3: 249–272. Jarvis, T. 2003. The money pit: Karst failure of Anchor Dam, Wyoming. In Evaporite Karst and Engineering/Environmental Problems in the United States, ed. K.S. Johnson, and J.T. Neal, 271–278. Oklahama Geological Survey Circular 109. Johnson, K.S. 2003. Gypsum karst and abandonment of the Upper Mangum Damsite in southwest Oklahoma. Evaporite Karst and Engineering/Environmental Problems in the United States, ed. K.S. Johnson, and J.T. Neal, 85–94. Oklahoma Geological Survey Circular 109. Johnson, K.S. 2004. Problems of dam construction in areas of gypsum karst. Karstology—XXI century: theoretical and practical significance. In Proceedings of the International Symposium. Perm, Russia. Johnson, K.S., and J.M. Wilkerson. 2013. Gypsum karst causes relocation of proposed Ceder Ridge Ram, Throckmorton County, Texas. In Proceedings of the Thirteenth Multidisciplinary Conference, ed. L. Land, L.D.H. Doctor, and J.B. Stephanson. New Mexico: National Cave and Karst Research Institute, Karlsbad. Karimi, H., and A.A. Pakzad. 2009. Water tightness evaluation of the Kangir dam reservoir, Ilam Province. In Iran ICWR 2009 International Conference, Shahrood, Iran. Kiselev, N.P. 2004. Karst collapses at area of Khadita Dam construction, Iraq. Karstology—XXI Century. Theoretical and Practical Significance. Perm, Russia, pp 266–270. Kondratyev, N.N. 1979. Characterisation of carbonate karst development at the area of Hadita Dam construction. Scientific Bulletin of Gidroproject, (In Russian). Moscow. Lafuente, R., J. Granell, I. Poyales, V. Florez. 2006. La loteta dam. A strategic location. In Dams and Reservoirs, Societies and Environment in the 21st Century, ed. Berga, L., Buol, J.M., Bofill, E., De Cea, J.C., Garcia-Perez, J.A., Manueco, G., Polimon, J., Soriano, A, and Yagüe, J, 105–112. London: Taylor & Francis Group. Lebedev, A.L., and A.V. Lekhov. 2011. Modelling of chainage permeability in a gypsiferous fractured-porous rock mass. In Geology, Engineering-geology, Hydrogeology, Geocrylogy, No. 1 (in Russian), 63–74. Moscow. Lykoshin, A.G., L.A. Molokov, and I.A. Parabutchev. 1992. Karst and Dam Engineering, in Russian. “Gidroproekt”, Moskva (in Russian). Mahjoob, D.F., A. Sadatifard, H. Hassani, and A. Zia. 2014. Upper Gotvand Dam and hydro power plant dealing with salinity in reservoir. Challenges, remedies and evaluations. In International Symposium on Dams in a Global Environmental Challenges, Bali, Indonesia.

155 Maksimovich, N.G. 2006. Safety of dams on soluble rock (The Kama hydroelectric power station as an example). Publisher “Garmonia”, Perm, Russia, p. 212. Meshkat, T., D. Mahjoob Farshchi, and E. Ebtekar. 2018. Evaluation of evaporate karstic challenge in Gotvand dam reservoir. In Proceedings of International Symposium, KARST 2018, Expect the Unexpected, ed. S. Milanović, and Z. Stevanović, 89–96. Belgrad: Centre for Karst Hydrogeology. Milanović, P. 1999. Tang-E-Shemiran Earth Dam Project. Mission Report. Western Regional Water Authority, Iran. Milanović, P. 2004a. Water Resources Engineering in Karst. Boca Raton: CRC Press. Milanović, P. 2004, 2005 and 2010. Gotvand Dam project. Problem of Anbal salt and gypsum (anhydrite) formation. Mission reports, unpublished. Moradi, G., and A.R. Abbasnejad. 2011. The investigation of extra seepage of the Gheisaragh Dam and remedial method. In 6th International Congress on Civil Engineering, 1–6. Semnan, Iran. Morlans, H., J.L. Moreno, and J.L. Galan. 2005. Estudio de filtraciones en le presa de Caspe II. Espana: Zaragoza. Nedriga, V.P., and E.A. Demyanova. 1986. Construction of dams on rocks that contain evaporates. Construction of Hydrostructures. (In Russian). Moscow, Russia. Osadchy, L.G., and R.I. Bahtiyarov. 1975. Rogunaskaya dam on the Vahsh River. (In Russian). Hydrotechnical Construction. No. 4, Moscow, pp 10–30. Payton, C.C. 1992. Geotechnical investigation and foundation design for the reconstruction of Quail Creek Dike. In Engineering and environmental geology of southwestern Utah, ed. K.M. Harty, 39– 51. Utah Geological Association Publication 21. Payton, C.C., and M.N. Hansen. 2003. Gypsum karst in southwestern Utah. In Failure and Reconstruction of Quail Creek Dike/Evaporite Karst and Engineering/Environmental Problems in the United States, ed. K.S. Johnson, and J.T. Neal. Oklahoma Geological Survey Circular 109. Pechorkin, I.A., and A.I. Pechorkin. 1979. Theoretical aspects and engineering-geological forecast on the on the shores of karst reservoirs. Bulletin International Association Engineering Geology 20. Raeisi, E., M. Zare, and J.A. Aghdam. 2013. Hydrogeology of gypsum formations in Iran. Journal of Cave and Karst Studies. Iran 75: 68– 80. Rodevich, I.V. 1989. Investigation of the hydrogeochemical regime of loess bases of hydraulic structures for the purpose of controlling chemical suffusion. Izvestiya VNIIG. (In Russian), pp. 40–46. Rezaei, A., H. Karimi, and H. Zhan. 2017. The importance of understanding the hydrogeology and geochemistry of karst terrains for safely sitting dams. Journal of Cave and Karst Studies 79 (1): 48–58. Rogers, J.D. 2007. The 1928 St. Francis Dam Failure and its Impact on American Civil Engineering. American Society of Civil Engineers. Rogers, J.D., and K.F. Hasselmann. 2013. The St. Francis Dam Failure: Worst American Engineering Disaster of the 20th Century. In AEG Shelmon Specialty Conference: Dam Failures and Incidents. Denver: Association of Environmental and Engineering Geologists. Sari, S. 2013. Détection des fuites d’eau dans le barrage de Joumine et étude de la sédimentation dans le barrage de Ghezela par la méthode nucléaire. Rapport interne INIS-TN–193–Université de Jendouba, Tunisie, pp. 20–40. Tolkonnikov, I.S. 1974. Erevanskaya Dam at Razdan River. Dam Geology, M. Engineria, T.VII. pp. 76–96. Turcev, A.A. 1938. Analyse of joint systems of the Perm formation during construction of Kama Dam (in Russian) Bulletin of Petrigraphy Institute – L. Question 12: 173–202.

156 Wittke, W., and H. Hermening. 1997. Grouting of cavernose gypsum rock undernith the foundation of the weir, locks and powerhouse at Hessigheim on the River Neckar. In Proceedings of the 19th Congress of the ICOLD, Florence, Q75, R.44, pp. 613–626.

Bibliography Balamirzoev, A.G. 2005. Method of calculating the filtration safety of hydraulic structures on a fractured plastered base // News of universities. North Caucasus region. Technical science. No. 4: 86– 90. Buchatsky, G.V., E.V. Zernov, L.A. Evdokimova, V.I Sergeev, S.D. Voronkevich. 1976. Creation of anti-filtration curtains with experimental application of a new chemical oil-well mortar // Hydrotechnical construction. No. 4. (in Russian). 4–6. Calcano, C.E., and P.R. Alzura. 1967. Problems of dissolution of gypsum in some dam sites. Bulletin of the Venezuelan Society of Soil Mechanocs and Foundation Engineering: 75–80. Chokhonelidze, G.I. 1957. To the question of assessing the physico-technical properties of saline soils used in hydraulic structures. Proceedings of the GruzNIIIP. Tbilisi. Issue 18–19: 23–43. (In Russian). Christoffersen, J., and M.R. Christoffersen. 1976. The kinetics of dissolution of calcium sulphatedihydrate in water. Growth 35 (1): 79–88. Craft, D., C. Cain, C. Sullivan. 2006. Seepage Geochemistry and Mineral Dissolution at Clark Canyon Dam. Department of the Interior Bureau of Reclamation Technical Service Center Dam Safety Office. Denver, Colorado. Davydov, V.A., and G.A. Tsai. 2018. Study of dangerous natural and manmade geological processes using geophysical methods. News of the USMU 2 (50): 65–71. (in Russian). Drogue, C. 1992. Hydrodynamic of karst aquifers: Experimental sites in the Mediterranian karst, south France. In Hydrogeology of Selected Karst Regions, vol. 13, ed. H. Paloc, and W. Back, 133. IAHS, Heise, Hannover. Edgell, H.S. 1966. Salt tectonism in the Persian Gulf Basin. In Salt Tectonics 100, ed. G.I. Alsop, D.J. Blundell, and I. Davison, 129– 151. Geological Society Special Publication. Eppelbaum, L.V. 2015. Quantitative interpretation of magnetic anomalyes from bodies approximated by thick bed models in complex environment. Environmental Earth Sciences 47 (74): 5971–5988. Ezersky, M. 2011. Imrovement of the seismic refraction methods for mapping of the buried salt layers along the Ded Sea shoreline. GII Report No. 211/486/09, Lod, Israel, p. 24. Ezersky, M., and Frumkin. 2017. Evaluation and mapping of Dead Sea coastal aquifers salinity using Transient Electromagnetic (TEM) resistivity measurements. Comptes Rendus Geosciences, 349(1): 1– 11. Frumkin, A. 1992. The karst system of the Mount Sedom salt diaper, Ph.D. dissertation (in Hebrew, English abstract), Hebrew University, Jerusalem, 133 pages. Frumkin, A., L. Kofman, and M. Ezerski. 2009. Improvement of the reliability of subsurface void detection, including sinkhole development, at the Dead Sea shore area by means of Ground Penetration Radar (GPR). Hebrew University—TECHNION Research and Development Foundation LTD—Geophysical Institute of Israel, Report No. MNI-ES-36-2008, 145 p. Gorbunova, K.A. 1979. Morphology and Hydrogeology of Gypsum Karst, 95. Perm: The Perm University. Gorbunova, K.A. 1989. Caves. Mountain encyclopedia. Mocsow, vol. 4. (in Russian).

9

Overview of Dams and Reservoirs in Evaporites

Gorshkov, V.S., V.S. Timashev, and V.G. Savel’ev. 1981. Methods of Physicochemical Analysis of Binders (in Russian). Moscow: Vysshaya Shkola. Guzina, B. 1997. Impact of Impounding on a Karst Reservoir Zield, 489. Congress des Grandes Barrages, Florence: Italy. Henisch, H.K. 1996. Crystal Growth in Gels, 874. New York: Dover Publications Inc. Johnson, K.S. 1998. Evaporite karst in the United States. Carbonates and Evaporites 12 (1): 2–14. Kafri, U., and B. Lang. 1997. Detection subsurface brines, freshwater bodies and the interface configuration in-between by the time domain electromagnetic method in the Dead Sea Rift, Israel. Environmental Geology 31: 42–49. Klizas, P.Y., N.G. Maximovich. 1979. Electrical Modelling of Filtration Related to Chemically Grouted Gypsiferous Rocks at Kama Dam Foundation (in Russian). Scientific workshop, Perm. Kozyreva, E.A. and Y.B. Trzhtsinsky. 2004. Karst and its correlation with other geological processes (with reference to the zone of influence of Bratskreservoir). ed. F. Guo, and D. Tang. World Corelation of Karst Ecosystem, Newsletter, Karst Dynamic Laboratory, Guilin, China. Krawczuk, C., U. Polom, H. Alrshdan, D. Al.-Halbouni, A. Sawarieh, and T. Dahm. 2015. New process model for the Dead Sea sinkholes al Kuznecov, A.M. 1953. About fibroses gypsum in Kungursk area. (in Russian). Bulletin of Scientific Institute of Perm University, T.13. Question 6: 503–518. Lamas, M.R. 1965. Los terrenos yesiferos como element de function de pressa (1a parte). Servocio Geolόgico. Informaciones y studios, boletin No 21, Madrid, pp 22–55. Lu, Y., and H.A. Cooper. 1996. Gypsum Karst in China. International Journal Speleological. 25 (34), Chapter II.13. Maksimovich, N.G., and V.I. Sergeev. 1983. Effect of chemical injection stabilization on gypsum stability in the foundation of hydraulic structures// Power Technology and Engineering 17 (7): 380–384. Mancebo Piqueras, J.A., E. Sanchez Perez, I. Menendez-Pidal. 2011. Water seepage beneath dams on soluble evaporate deposits: a laboratory and field study (Caspe dam, Spain). Bulletin of Engineering Geology (2012). https://doi.org/10.1007/s100640379-2. Springer-Verlag. Maximovich, N.G. 1987. Evaluation of dam foundation at soluble rocks, Comprehensive investigations of geology and hydro-ecology of the Kama Dam Reservoir (in Russian). Perm, pp 114–121. Maximovich, N.G. 2005. Dams on gypsiferous—A possible source of disaster. In International Symposium on Latest Natural disasters— New Challenges for Engineering Geology, Geotechnics and Civil Protection, 65. Abstract book, Sofia. Maximovich, N., and O. Meshcheryakova. 2009. The influence of gypsum karston hydrotechnical constructions in Perm region. In Geological Engineering Problems in Major Construction Projects, Chengdu 2: 604–607. Milanović, P.Т. Geological engineering in karst: Dams, Reservoirs, Grouting, Groundwater Protection. Water Tapping Tunneling. 2000. Belgrade, Zebra, pages 347. Milanović, P., 2010. Gotvand Dam Project, Iran. Evaporite problem. Mission Report. Teheran. Not published. Pearson, R. 1999. Geology and Safety of Dams Case Histories in Gypsum Karst for Horsehooth Dam and Reservoir and Carter Lake Dam No. 2, Colorado Big Thomson Project, Ft. Collins and Loveland, Colorado. U.S. Department of the Interior. Bureau of Reclamation. USBR Technical Service Center Center D-8321, Denver. Plata, A., and L. Arguas. 2002. Detection and Prevention of Leakage from Dams. The Netherlands: A.A. Balkema.

Bibliography Rodionov, N.V. 1958. Engineering-Geological Investigations in Karst Areas, 257. Moscow: Gosgeoltekhizdat, (In Russian). Sergeev, E.M., G.A. Golodkovskaya, R.S. Ziangyrov, V.I. Osipov, V.T. Trofimov. 1983. Gruntovedenie, 392. Moscow: MGU, (in Russian). Voronkevich S.D., S.N. Emelyanov, N.G. Maksimovich. 1986. Finite-element modeling by the influence of post-injection processes on the effectiveness of the filtering curtain. Application of numerical

157 methods to problems of geomechanics (IISS). M. 90–99. (in Russian). Yuan, D. 1991. Karst of China. Beijing, China: Geological Publishing House. Yu, L., and A.H. Cooper. 1997. Gypsum Karst Geohazards in China. In The Engineering Geology and Hydrogeology of Karst Terranes, ed. F.B. Back, and J.B. Stephanson. Roterdam: A.A. Balkema.