Page 1

Front. Archit. Civ. Eng. China 2010, 4(1): 40–46 DOI 10.1007/s11709-010-0006-5


Jinsheng JIA

Several issues to be considered for long-term better behavior of concrete gravity dams

© Higher Education Press and Springer-Verlag Berlin Heidelberg 2010

Abstract Along with economic, social quick development and urbanization, dams and reservoirs are of strategic importance for flood control, water supply, electricity production, irrigation, etc., both for developed countries and for developing countries. Climate change is a new challenging issue to be considered which will speed up the development of hydropower in developing countries. More and more attention will be paid on the long-term better behavior of dams to guarantee the safety of the people involved and the better development of the world. There are about 50000 old dams in the world and a lot of them have been completed and operated for more than 50 years. However, how do we evaluate the dams’ safety? How do we make the decision to do rehabilitation work or to rebuild a new dam based on evaluation results? The life span and the real safety status of old dams becomes a challenging task for the dam society, especially for China because it has more than 6000 dams to be evaluated and rehabilitated within the next few years. Based on the investigation of the Fengman gravity dam, which is 91.7 m high, operated since 1943 and suffered uplift pressure, freeze and thaw problems, etc., discussions on the life span evaluation of old concrete gravity dams have been made. The reasonable coefficient of dam safety has been discussed. The social decision for the final choice after comprehensive studies has been introduced. Keywords dam safety, economic life span, structural life span, environmental life span, rehabilitation, social decision

Received July 17, 2009; accepted November 6, 2009

Jinsheng JIA ( ) China Institute of Water Resources and Hydropower Research, Beijing 100038, China E-mail:

1 Life spans of old concrete gravity dams and comprehensive evaluation suggestions for a long certain period It is very important for society to understand the concept of the life span of dams because there are clear life spans in the design of bridges, houses and other structures in many countries. However, it is really difficult for the dam society to make a very clear definition of the life span of dams because there are so many complicated issues to be considered and there is a short history of dams in operation made by modern techniques. To make it easy, we can discuss the life span of old concrete gravity dams for better understanding of the complicated situations. For concrete gravity dams built in the 20th century, there are no standards for the normal working period (structure life span) in China or in other countries. According to literatures, a large number of dams were built 1000 years before with a height lower than 30 m. Few of them exist nowadays. Most of these dams have failed and their structure life span was short. The main reason is that the level of design, construction, and reinforcement was very low. For example, flood control standard was low, or there were obvious shortcomings in dam structures and construction quality, or the technology of operation and maintenance were limited. Since the 20th century, the design and construction of concrete gravity dams have been standardized and technologies in reinforcement have become more and more advanced. Because of these, most of the concrete gravity dams in existence are still in operation and have longer structure life span. The life span of a dam not only depends on its quality but also on the environment and the needs of society. At the same time, it also has a close relationship with the reinforcement. The life span of concrete gravity dams can be divided into structure life span, environmental life span and economic life span. Structure life span mainly lies on the conditions of dams. Because of structures, materials, earthquakes, floods or other reasons, some dams may become unsafe and need to be rebuilt, disused or removed,

Jinsheng JIA. Several issues to be considered for long-term better behavior of concrete gravity dams

which can be considered to have reached their structure life span. For those dams that are breached when the loading (overtopping flood, flow through or beneath the dam, earthquake loading, etc., but excluding wars or terrors) exceeds the resistance against overtopping, internal erosion or piping, slope instability, sliding/overturning, etc., they can also be considered to have reached their structure life spans. For example, more than 3000 dams in China and 1300 dams in the United States were breached and reached their structure life span. One gravity dam with a height over 50 m in Canada was reported to have developed serious problems after operating for more than 50 years; the cost of reinforcement was deemed to be higher than that of rebuilding and the final decision was to rebuild it in 2003. This can also be considered for the dam to have reached its structure life span. Considering structures with thin thickness affiliated with the dams, usually they need to be rehabilitated frequently or rebuilt within about 50 years, including some metal structures. It is different with the definition of life span of concrete gravity dams. However, it sometimes has a big influence in social decision on whether or not to remove or rebuild the dam. When a dam needs to be demolished because of reservoir silt or the needs of environmental protection or the changing of dam purpose, it can be considered that the dam has already reached its environmental life span. About 500 small dams with an average height of 5 m in the USA have been removed because of social or environmental purpose changes. For example, some of them were built to supply water to factories and the factories have been stopped. Therefore, it is no longer necessary to keep them. There are similar cases in other countries. Up to now, there are very few reports of removing large dams over 30 m for social or environmental reasons. However, there are some reservoirs designed for 100 year’s application considering heavy silt problems. For concrete gravity dams, there is a yearly maintenance work and evaluation requirement every 5 or 6 years in order to guarantee the dam safety based on practices in several countries. However, it is also possible that after a period of operation, the security and functions of a dam may be far below the standards compared with new dams. In this case, we can consider that the dam has already reached its economic life span. During the structure life span of a dam, there could be several or many economic life span cycles. Different dams have different economic life spans. The normal service period of a concrete structure is about 50 years as defined by some countries, but according to the operation status of concrete gravity dams around the world since the 20th century, the economic life spans of gravity dams may be over 50 years for most of the cases. To make it easier and safer and based on the definition of the life span of other concrete structures, we can define that the economic life span of a concrete gravity dam is 50 years. The main reasons are as


follows: 1) It is necessary to make dams as safe as a new dam after 50 year of operation. Considering the related technology and the behavior of a dam changing obviously with time, it is of practical significance to carry out comprehensive evaluation and studies on dam safety at their economic life span and to make their safety similar to the level of a new dam. Technique will be definitely improved after 50 years development. Many design standards and design parameters including earthquake parameters can be quite different for building a new dam after 50 years. 2) It is necessary to be clear to society, especially for the government and the people downstream. It is very important for the people downstream to better understand the safety situation of the dam. They are not professionals but they have a big influence on the final decision whether or not to keep or remove the dam. Another factor that affects reinforcement is the environmental life span of dams. Sometimes, the environmental life span may be obviously shorter than the structural life span and it will also have impact on the economic life span of the dam. For example, many rivers in the world are sediment-laden rivers and the sediment in reservoirs will reduce the capacity of the reservoir directly. Although measures have been taken to alleviate siltation to some extent, the life span of such reservoirs is sometimes limited, even shorter than 100 years, which will affect the schemes of reinforcement. The closing down of the reservoir will cause the ending of the economic life span of the dam. There is a good example to understand the life span of concrete gravity dams. The Dujiangyan irrigation project has been used for more than 2600 years. It is a very large project with weir for flood control, water supply and farm irrigation. It is so famous that the city nearby has the same name. When we investigate the life span of this project, we can find that only three parts of the project have lasted more than 2600 years: 1) Foundation rock. The base rock of the project has not been changed; 2) The layout of the project. In order to keep the project safe and in good condition for operation, there are very strict regulations for maintenance and rehabilitation yearly. Some parts of the project are carefully checked and rebuilt to maintain a better behavior for a certain period of about 10 years. However, the main parts have been destroyed several times and have been rebuilt in a similar layout. 3) Social requirement. Although it has lasted more than 2600 years, the social requirements remain the same for this project. The economic life span for the Dujiangyan project may be 10 years and the structure life span on average may be 100 years. The environmental life span is more than 2600 years. For most concrete gravity dams over 30 m, the structural


Front. Archit. Civ. Eng. China 2010, 4(1): 40–46

life span should be longer than 500 years considering the materials, the design principles and the advanced techniques used for rehabilitation. Social decision making is more and more complicated these days. If there is no better understanding of society, the structural life span may be much shorter than it should be. The professionals should be responsible for the economic life span, structural life span and better understanding of society. 1.1

Dam safety in China

Based on statistics, among 87076 reservoirs in China in 2009, there are about 37800 with safety problems. To ensure the security, studies and reinforcement have been carried out in recent years. In the new schemes (2007– 2009), the reinforcement of 6240 dams will be conducted. Some of them are very difficult to deal with. The main problems are as follows: 1) Flood control problems. Because of longer hydrology data, the flood control standard of reservoirs cannot meet the new operation conditions and the discharge ability of reservoirs becomes insufficient. If we consider climate change and more strict requirements on dam safety with the development of urbanization, it would be more important to carefully re-evaluate the requirement for flood control for many dams. 2) Seismic problems. According to the “Seismic Parameter Distribution Map of China” (GB18306-2001) and the current specification, the safety considering seismic loading cases of many reservoirs cannot meet with the current requirements. For the area affected by the May 12 Wenchuan Earthquake in 2008, more careful investigation has to be done. There are 4 dams higher than 100 m in this area, namely Shapan RCC arch dam (H = 132 m), Zipingpu CFRD dam (H = 156 m), Bozhusi concrete gravity dam (H = 120 m) and Bikou rockfill dam with clay core (H = 105 m). They are very close to the epicenter of the earthquake. The designed accelerations for Shapa and Zipingpu are much lower than the real ones. The four dams were kept safe after the strong earthquake. A lot of new tasks related to this earthquake have to be done in the future. 3) Stability of dams. Because of insufficiency in dam section or cracks existing or joints opening, many dams have to be evaluated and rehabilitated. 4) Leakage and uplift problems. Some dams suffered a lot with leakage. Most of them are earth dams and usually new clay cores have to be built to decrease the leakage. For old concrete gravity dams, uplift for some projects is usually higher and higher. It is not easy to find a reasonable and reliable solution. Grouting is sometimes used but it is not so reliable to decrease the leakage and uplift for the old concrete gravity dams. 5) Crack and aging problems. Structures with thin thickness affiliated with the dams suffer a lot of cracks and

aging problems. Some of the dams have similar problems and need to be rehabilitated carefully in dry condition or under water. 6) Metal structures and electrical equipment problems. Metal structures and electrical equipment are aging or seriously eroded that they can hardly be operated normally, which have seriously affected the safety of reservoirs. 7) Management facilities and monitoring, observation equipment are not in good condition. 8) Reservoir silt and landslide. 9) Freezing and thawing problems. 10) Others. Large-scale construction and management of hydro projects have been conducted for more than 50 years in China and many effective methods and experiences on reinforcement and heightening have been accumulated. However, problems related to dam safety still are challenging issues to be investigated to achieve a better long term behavior of dams. The government of China has paid much attention these years to be sure about dam safety. There are fewer dam failure cases every year. There are two main reasons. One is high quality in dam design and construction. Another is that the governors are usually the main persons for the safety of the dams especially in flood seasons.

2 Main problems after new comprehensive evaluation on Fengman Dam The Fengman concrete gravity dam is situated at the main stream of the second Songhua River, 24 km downstream of Jilin City, Jilin Province. It is situated in a severe cold area. Its mean annual temperature is 5.3°C. The highest mean monthly temperature is 24.3°C and the lowest is – 19.7°C. The maximum height of the dam for its original design was 90.5 m and the dam crest elevation is 266.5 m. Dam construction started in Apr. 1937 and water impounding started in Nov. 1942. The project was completed and operated in Oct. 1953. By the end of reinforcement in 1996, the maximum dam height was 91.7 m and the dam crest elevation was 267.7 m. The dam crest is 1080 m long, divided into 60 dam sections, each of which is 18 m long. Arranged from the left to the right bank, the 9th to 19th dam sections of the dam are overflow dam sections, the 21st to 31st are intake sections for power generation. The upstream slope of the dam section is 0.05 and downstream slope is 0.78. During the construction, the dam cross section was divided into A, B, C and D blocks by the longitudinal joints. The typical cross section after reinforcement is as shown in Fig. 1. In order to guarantee the safety of old concrete dams over 50 years, it is really necessary to do a comprehensive evaluation based on studies on the Fengman Dam in China. The main problems and main achievements from evaluation up to now for the project are as follows [1]:

Jinsheng JIA. Several issues to be considered for long-term better behavior of concrete gravity dams


Fig. 1 Typical cross section of dam

1) Problems of stability related to seismic assessment Original conclusion: With weak longitudinal joints and sub-longitudinal joints, the stress level of some points of the Fengman Dam is higher than the allowable value and safety cannot be guaranteed for seismic load and some parts of Fengman Dam may be destroyed based on earthquake parameters. Based on results of the original analyses, anchoring has to be installed and had been installed before 1997 from the top to the foundation and the dam had been increased to 91.7 m high (1.2 m higher than the original dam) to improve the stability of the dam. Considering the reliability of anchoring and other issues, it is still a safety problem under seismic loading cases. Evaluation on dam safety should be made every 5 years according to Chinese regulations. The seismic acceleration coefficient adopted for rehabilitation design and safety evaluation is 0.161. However, some experts suggest that the value should be higher than 0.161 compared with similar projects. New conclusion: Considering new progresses made in the past 20 years, seismic parameters have been comprehensively re-evaluated based on current standards. It is found that the acceleration coefficient should be decreased from 0.161 to 0.131 and not to 0.22 or even higher as earlier estimated. With this conclusion, the dam safety under seismic loading case is much better than before. 2) Flood control Original conclusion: The spillway consists of eleven 6 m by 12 m orifice sluice ways. Discharge capacity of 1300 m3/s by the 10 turbines in the powerhouse is used for flood

control before the reservoir level reaches El. 267.7 m. New conclusion: The discharge capacity of the plant would not be allowed for use when the reservoir level is at 267.7 m to guarantee the dam safety based on current regulations. The flood control problems will be more critical for the current design standards compared with the original ones in history. Some experts suggest that it could be carefully evaluated to use the possibility for predischarging based on reliable hydrological monitoring and prediction system. However, it is not allowed by regulations. Further measures to increase the capacity for flood control have to be investigated and evaluated. 3) Serious leakage of the dam and high uplift pressure Original conclusion: With poor integrity and defects, such as cracks and honeycomb structures, serious leakage from the dam body and joints occurred after impoundment, which affected the integrity and durability of the dam. When the reservoir water level reached El. 255 m in 1950, leakage measured in the galleries reached 16380 L/min and the wet area in the downstream surface was about 24947 m2. After rehabilitation for many times, leakage today reduced to 39 L/min in total and wet area in the downstream surface at El. 256.55 m was about 440 m2 in 2004, most of which was located at spillway sections. The average uplift pressure coefficients for blocks of 8, 14, 22, 28, 35, 40, and 47 at different positions monitored in 1996 were 0.84, 0.48, 0.63, 0.16 and 0. The distances between test holes to the dam axis are 2.9 m, 6.5 m, 12 m, 39 m and 51.6 m, respectively. Data at block 15 in 2005 gave similar results. Grouting measures could be acceptable and have been


Front. Archit. Civ. Eng. China 2010, 4(1): 40–46

done many times after first impounding to decrease the leakage and uplift. The leakage is decreased obviously and successfully but the uplift is still a problem. New conclusion: High uplift pressure is still a big problem to be solved in the future to reach the design specification. Grouting measures can be used but not enough especially for decreasing the uplift in the body. 4) Poor quality of concrete Original conclusion: The strength of dam concrete of 90 d in the original design should be 15 MPa, but actual strength of concrete was only 12 MPa, 9 MPa or even 5 MPa at different parts of the dam. Aggregate and cement used were not of good quality. Water reducing admixture and air-entraining admixture agent were not used and there was no freezing resistance consideration for all the concrete used even though the dam locates at an extremely cold area. Concrete mix proportion in construction had problems and there was no temperature control during construction. Outer concrete of 0.6 m thick has been replaced in recent years during rehabilitation in order to improve the situation. Some parts of the dam concrete in the upper and downstream side have very low strength of about 5 MPa. Grouting measures cannot improve the situation. New conclusion: Comprehensive rehabilitation has to be done to solve this problem. 5) Longitudinal joints of the dam Original conclusion: Fengman Dam was divided into blocks by transverse joints, longitudinal joints, subtransverse joints and sub-longitudinal joints. No special treatment was done on most of these joints. No shearing resistant structure and joint grouting was made in AB longitudinal joints from EL. 220 m to EL. 242 m (see Fig. 1). Although grouting was carried out during operation for many times, unbounded longitudinal joints of the dam body are still a problem. Model test, static and dynamic analyses made previously show that the joints have a big influence on stress distribution and stability. Strong anchoring measures should be carried out considering possible higher seismic load. New conclusion: Strong anchoring measures are not as important as before and it is difficult to be sure about their reliability because it is difficult to get a clear monitoring result from time to time for evaluation. 6) Freezing and thawing problem Original conclusion: With poor quality of concrete and serious uplift pressure of the dam, some surface concrete was destroyed by freezing and thawing condition, especially for spillway sections. In 1986, the damaged concrete of the upstream surface above elevation 245 m and of the downstream surface from top to ground level were excavated 0.4 m and were covered by 1 m reinforced concrete with high strength and frost resistance. The dam crest was heightened by 1.2 m. For the spillway, the downstream surface was replaced by new concrete with 1.5 m thickness and was fixed with lots of anchorage (3.5 m in depth).

New conclusion: The freezing and thawing problem is still a big and challenging problem up to now and it will decrease the safety obviously. The original measures are not enough and concrete with thickness of more than 4 m should be put on the surface downstream to achieve similar safety standard as a new dam based on FEM analyses. Fengman Dam has been operated safely under all the loading cases except earthquakes from the first impounding. Many measures have been taken to keep it in a better behavior. Main rehabilitation work finished before 1997 is as follows [2]: 1) Bituminous concrete lining with a thickness of 10 cm was placed on the upstream surface between El. 245 m to El. 226 m before the flood season in 1990. 2) Grouting to dam body and foundation was one of the main measures adopted for reducing the seepage and uplift pressure. Grouting was carried out in 37 dam sections in different years. 3) Pre-stressed anchors were installed to improve the dam safety under earthquake loading cases. 378 prestressed cables in total with different loading grades were installed from No.7 to No. 49 dam sections, in which 361 cables were installed in the dam body (excluding the test anchors) and 17 cables, in the dam foundation. 4) Heightening of dam. According to the dam reinforcement design, the dam crest has been raised by 1.2 m to improve the stability of the dam and increase flood control ability. The work above is reasonable but not enough to achieve a new economic life span for the old dam.

3 Safety evaluation based on numerical analysis by FEM for Fengman Dam To truthfully reflect the practical safety situation of the Fengman Dam and achieve clear results compared with the current dam, simulation analyses have been made with consideration of the construction processes and joints including longitudinal and construction joints. Whole course simulation analysis of BL47 from the construction period to 2005 is conducted. A three-dimensional FE model is built, which takes into account all factors that will affect the stressing and deformation of the dam [3]. In simulation analysis process, the construction process is simulated according to the construction data recorded, using measured data of air temperature and water level as boundary conditions. The calculation period is from Oct. 1941 to Dec. 2005. In the concrete construction period, the minimum calculating step is 0.5 d, while the maximum calculating step is 5 d. In operation period, calculating step is 10 d. There are 2707 calculating steps in the whole simulation analysis process in total. In the whole course simulation, the adiabatic temperature rise of the concrete material in simulation analysis is determined according to the practical mixing ratio of the

Jinsheng JIA. Several issues to be considered for long-term better behavior of concrete gravity dams

concrete and the maximum internal temperature observed on site. The final elastic modulus, coefficient of linear expansion and coefficient of temperature conductivity are determined according to observed data of displacement and temperature by back analysis. Main parameters can be seen in Table 1. According to the whole course simulation calculation, temperature field, stress field and displacement field of BL47 at any time can be obtained. Comparing the horizontal displacement at the dam top obtained by simulation analysis with measured data (see Fig. 2), it can be found that these two are consistent with each other, which indicates that the simulation analysis can reflect the practical operating condition of the dam. According to the envelope diagram of minimum temperature field during operation from 1990 to 2005 (see Fig. 3), negative temperature may occur in concrete of the dam top and downstream side during winter, which may cause frost thawing fracture due to sever seepage in the dam concrete. The depth of the negative temperature zone of the downstream side is about 4.5 m. According to the contour diagram of σ1 in winter (see Fig. 4), tensile stress with maximum value of 0.8 MPa is distributed in the area of the dam shell, in most part of which concrete has low tensile strength of about 0.5 MPa. The depth of the tensile stress zone of the downstream side is about 3.2 m. Furthermore, tensile stress of 0.2–0.8 MPa is distributed in the dam heel with width of 6 m or so. Calculated results show that the value and distribution area of tensile stress in the dam heel may decrease, while tensile stress distributed in concrete of the dam center may cause longitudinal joints and construction joints to open. Results of simulation analysis show that most part of the Table 1


longitudinal joints of Fengman Dam have opened gradually even during the construction period. More than 70% of the longitudinal joints have opened, which makes it impossible for the dam to bear load as a whole and has deteriorated the work behavior of the dam. Based on calculations, the safety factor of sliding resistance along the dam base, when considering the effect of longitudinal joints and stress history, is 10%–20% less than that evaluated by traditional methods, in which no such factors are taken into account. On some blocks of Fengman Dam with faults passing through bedrock, potential risk of instability under some work condition may exist. Overloading analysis under different conditions has been carried out to evaluate the safety of Fengman Dam compared with a similar new dam: 1) Case 1: Overloading analysis on a newly built dam which has the same dam section as Fengman Dam and longitudinal joints are well dealt; 2) Case 2: Overloading analysis on a dam with the same dam section and concrete quality as Fengman Dam while longitudinal joints are well dealt; 3) Case 3: Overloading analysis on Fengman Dam under current work conditions. The results can be seen in Table 2. Obviously, poor concrete quality and effect on longitudinal joints are two important factors that have led to the decrease of dam safety. To increase the safety of Fengman Dam, it is an option to add new concrete on the downstream face. When 4-mthick concrete layer is added, overload coefficient will increase to 2.33–2.47 even when no reinforcement is done to longitudinal joints and the safety factor of sliding resistance along the dam base will increase more than 17%, which will meet the requirements of sliding resistance.

Value of material thermodynamic parameters E/GPa

specific gravity/(kg$m–3)

Poisson’s ratio

coefficient of temperature conductivity/(m2$h–1)

specific heat/(kJ$(kg $°C)–1)

coefficient of linear expansion/(10–6°C–1)















Fig. 2 Calculated horizontal displacement compared with observed data at dam top of BL47


Front. Archit. Civ. Eng. China 2010, 4(1): 40–46

4 Main choices for the future of Fengman Dam

Fig. 3 Envelope diagram of minimum temperature during operation /°C

For the long-term better behavior of Fengman Dam, the choices in the future should be based on the following principles: 1) The technique to be used should be sound and possible; 2) The cost should be reasonable; 3) The dam safety after rehabilitation should be achieved without any problems. Two solutions are presented and compared in detail. Solution one: Adding new concrete with 4- to 6-mthickness on the faces of the dam (upper part for the upstream surface and downstream surface) and grouting used to decrease the uplift. Thickening the dam along faces will increase the safety of the whole dam and also make negative temperature and tensile stress transfer to the new high strength concrete, which will improve the safety of the dam to a similar new one. This solution may cost about 500 million in USD. Solution two: Building a new RCC dam and use the old one as coffer dam. This solution may cost about 900 million in USD. The Fengman reservoir with more than 10 billion cubic meters of storage capacity is not allowed to be emptied. According to feasibility studies, several other options have been put forward: 1) Install a geo-membrane on the upstream surface with Carpi tech, which can be done partly underwater; 2) Design and build a special removable coffer dam used to form a dry site so that it is possible to rebuild an upstream surface of the dam; 3) Using new concrete core method to replace concrete of some distance to the upstream face with new concrete part by part to form a concrete impervious wall. Up to now, the author prefers to build a new dam for a long-term better behavior. Acknowledgements This work was supported by the National Natural Science Foundation of China (Grant No. 50879095) and the National Key Technology R & D Program of China (No. 2006BAC14B04).


Fig. 4

Table 2

Contour diagram of σ1 in winter /Mpa

Results of overloading analysis

items overload coefficient

Case 1

Case 2

Case 3




1. Jia Jinsheng, Lu Yihui, Li Xinyu, Xia Shifa. Comprehensive Evaluation on Fengman Dam. Research Report, China Institute of Water Resources and Hydropower Research, Beijing, 2007 (in Chinese) 2. Jia Jinsheng, Lu Yihui, Li Xinyu, Zhang Jiahong. Solutions to Fengman Dam Rehabilitation. Research Report, China Institute of Water Resources and Hydropower Research, Beijing, 2007 (in Chinese) 3. Zheng Cuiying, Zhang Guoxing, Yang Bo. FEM Simulation Analyses on Fengman Concrete Gravity Dam. Research Report, China Institute of Water Resources and Hydropower Research, Beijing, 2007 (in Chinese)

Several issues to be considered for long-term better behaviorof concrete gravity dams  

Download Artikel Riset: