Analysis of earth dams affected by the 2001 Bhuj Earthquake

doi:10.1016/j.enggeo.2005.06.002

Engineering Geology

Volume 80, Issues 3–4, 30 August 2005, Pages 282-291

https://doi.org/10.1016/j.enggeo.2005.06.002 Get rights and content

Abstract

An earthquake of magnitude of 7.6 (M_w 7.6) occurred in Bhuj, India on January 26, 2001. This event inflicted damages of varying extents to a large number of small to moderate size multi-zone earth dams in the vicinity of the epicenter. Some of the distress was due to the liquefaction of saturated alluvium in foundation. Liquefaction was relatively localized for the majority of these dams because the earthquake struck in the middle of a prolonged dry season when the reservoirs behind these dams were nearly empty and shallow alluvium soils underneath the downstream portions of the dams were partly dry. Otherwise, liquefaction of foundation soils would have been more extensive and damage to these dams more significant. Six such dams have been examined in this paper. Four of these facilities, Chang, Shivlakha, Suvi, and Tapar were within the 50 km of epicenter region. These dams underwent free-field ground motion with peak ground accelerations between 0.28g to 0.52g. Of these Chang Dam underwent severe slumping, whereas Shivlakha, Suvi, and Tapar Dams were affected severely especially over the upstream sections. Fatehgadh Dam and Kaswati Dam were affected relatively less severely. Foundation conditions underneath these dams were first examined for assessing liquefaction potential. A limited amount of subsurface information available from investigations undertaken prior to the earthquake indicates that, although the foundation soils within the top 2.0 to 2.5 m underneath these dams were susceptible to liquefaction, Bhuj Earthquake did not trigger liquefaction because of lack of saturation of these layers underneath the downstream portions of these dams. These dams were then analyzed using a simple sliding block procedure using appropriate estimates of undrained soil strength parameters. The results of this analysis for these structures were found to be in general agreement with the observed deformation patterns.

Introduction

An earthquake of magnitude of 7.6 (M_w 7.6) occurred on January 26, 2001. The epicenter of the main shock of the event was located near Bachau at latitude 23.36°N and longitude 70.34°E with a focal depth of about 23.6 km. The event, commonly referred to as the Bhuj Earthquake, was among the most disastrous earthquakes that have affected India.

Bhuj Earthquake affected a large number of small- to moderate-size water-retaining earthen dams and reservoirs, constructed to fulfill the water demand of the area. Most of these dams are embankment dams typically constructed across discontinuous ephemeral streams. Although a number of such structures were within 150 km of the epicenter (Fig. 1), the consequences of the damage caused by the earthquake to these dams and ancillary structures were relatively light. This is primarily because of the low reservoir levels during the earthquake. The nature of damage to the embankment dams within the epicentral region is summarized in Table 1.

The performance of six embankment dams affected by Bhuj Earthquake is examined here. Among these, Chang Dam underwent almost a complete collapse because of liquefaction of shallow foundation soils. Shivlakha Dam was also severely damaged leading the failure of the upstream slope presumably because of liquefaction underneath the upstream portion of the dam. Damages to Suvi, Tapar, Fatehgadh, and Kaswati Dams were relatively less severe and confined near the upstream toe, upstream slope, and dam crest.

Limited subsurface data available from investigations prior to Bhuj Earthquake were analyzed using the simplified procedure for assessment of liquefaction potential (Youd et al. 2001). These analyses indicate a likelihood of widespread liquefaction of shallow alluvium soils underneath Chang Dam, while for Shivlakha, Tapar, Suvi, Fatehgadh, and Kaswati Dams liquefaction could only have occurred underneath the upstream slope because the foundation soils were partially dry at the time of the earthquake or because of overburden pressure due to the dam structure.

The dams were subsequently analyzed using the sliding block method originally developed by Newmark (1965) and the design charts developed by Hynes-Griffin and Franklin (1984) facilitating the use of the Newmark procedure for estimating the magnitude of deformation. The deformations estimated from the sliding block procedure were compared with the observed deformation patterns following the Bhuj Earthquake to check the predictive capability of this simple procedure. The results indicate a reasonable agreement between the deformations estimated from the sliding block procedure and observed distress pattern. However, it should be noted that these analyses are based on limited sub-surface data from investigations undertaken before the occurrence of Bhuj Earthquake and ground motion estimates in a setting where site-specific earthquake records are not available.

Section snippets

Observed dam performance

A brief summary of the performance of the six dams examined in this study is provided in the following subsections. For a more detailed account of the post-earthquake damage survey at dam sites reference may be made to the EERI (2001) Reconnaissance Report.

Assessment of liquefaction potential

Essential details of the procedure for assessing liquefaction potential and the results of this assessment are presented in the following subsections.

The sliding block method

The sliding block method was used in this study to estimate the deformation potentials for Chang, Shivlakha, Tapar, Fatehgadh, Kaswati, and Suvi Dams. Procedural details and results from these analyses are as follows.

Analyses and results

Computer program XSTABL version 5.2 (Interactive Software Designs, Inc., 1994) and the Modified Bishop method were used in the pseudo-static slope stability analyses. The input parameters used in the analyses are listed in Table 3. For the semi-pervious shell within dam body, the assumed soil properties of Table 3 reflect typical shear strengths of materials used in Dam construction in the study area (Nadpura and Ramchand, 2005). The strength parameters of the liquefied and non-liquefied

Conclusions

A simple method of analysis has been used to estimate the permanent deformations within six earth dams due to Bhuj Earthquake. These dams partially fulfill the irrigation and drinking water needs of a semi-arid area that was affected by the M_w 7.6 earthquake. Although these facilities were within 150 km from the epicenter of the earthquake, only one of the three dams collapsed because of the earthquake. The performance could, however, have been worse had the reservoirs been full when the

References (11)

E.L. Krinitzsky et al.
The Bhuj, India, earthquake: lessons learned for earthquake safety of dams on alluvium
Engineering Geology
(2002)
Earthquake Engineering Research Institute (EERI), 2001. Bhuj, India Republic Day January 26, 2001. Earthquake...
Hynes-Griffin, M.E., Franklin, A.G., 1984. Rationalizing the Seismic Coefficient Method. Miscellaneous Paper GL-84-13,...
I.M. Idriss
Response of soft soil sites during earthquakes
Interactive Software Designs, Inc., 1994. XSTABL: An integrated slope stability analysis program for personal...

There are more references available in the full text version of this article.

Cited by (53)

Dams in the wake-up call of the 2023 Türkiye earthquake sequence: Insights from observed damages, risk assessment, and monitoring
2024, International Journal of Disaster Risk Reduction
On February 6, 2023, Türkiye experienced a significant earthquake with a magnitude of $M_{w}$ 7.8¹, followed by four additional earthquakes throughout the day, measuring $M_{w}$ 5.8 and above, including one with a $M_{w}$ 7.5. As a result, over 100 dams in the region experienced low to high seismic loads. The impact of these seismic events resulted in a range of damages to dams, varying from light to severe. The authors participated in several post-earthquake reconnaissance missions to collect perishable data regarding the damage. This paper is three-fold: First, it explores the seismic performance and observed damage in 17 dams (majority embankments) based on data collected from the field. During the seismic activity, embankment dams underwent failure modes such as strength deterioration, yielding, settlement, tension crack formation, as well as sliding along planar or circular surfaces. Second, the paper offers an updated seismic risk analysis for ten selective damaged dams in the regions. Total risk factors are quantified using two different methods. The results show that the majority of the studied dams are classified as high and very high-risk dams. Third, the paper evaluates 11 other dams in the region that were equipped with sensors for monitoring the ground motions. The recorded acceleration at the crest and the abutment are further used to investigate the amount of amplification based on transfer functions. These recordings will be crucial for subsequent model calibration and system identification. Overall, the paper’s objective is to contribute to sharing post-earthquake reconnaissance mission information and to establish the groundwork for in-depth numerical analyses and quantitative risk assessments of dams located in seismically vulnerable areas.
Probabilistic seismic analysis for liquefiable embankment through multi-fidelity codes approach
2021, Soil Dynamics and Earthquake Engineering
Citation Excerpt :
Soil liquefaction continues to be considered one of the most devastating and complex behaviours that affects the soil due to earthquake loading. Due to the combined effects of excess pore pressure generation, permanent soil deformation induced by the ground shaking, a large number of geotechnical structures, such as river dikes, highway embankments, and earth dams, founded on saturated loose sandy ground have been frequently damaged during past major earthquakes in the last few decades [1–6], among others]. Numerical and experimental studies (e.g. centrifuge tests) have shown that the induced damage to such embankments is principally due to the liquefaction of foundation soil, resulting in excessive settlements, lateral spreading and slope instability.
The study focuses on both the effects of soil foundation liquefaction on the induced damage of an embankment and on the generation of a probabilistic seismic demand model that relates few specific characteristics of seismic hazard at a site with the liquefaction-induced settlement. To this aim, several non-linear dynamic Finite Element (FE) analyses were performed. However, given the costly evaluation of the numerical model simulations for a large number of earthquake records, a surrogate model based on multi-fidelity approach is used to represent the output of the expensive FE model. In this approach, the training database is composed by computational low-fidelity data together with limited high-fidelity one. Thus, the proposed methodology is used to generate the curves describing the annual rate of exceeding different values of damage levels. A comparison with the FE reference results suggests that the accuracy of the prediction of the proposed surrogate model is comparable to those from direct numerical FE analysis. Findings also illustrate clearly the importance and the advantages of an adequate definition of the input parameters to build the surrogate model.
Cyclic and postcyclic behaviour of silts and silty sands from the Indo Gangetic Plain
2019, Soil Dynamics and Earthquake Engineering
Citation Excerpt :
In such scenarios, shear strength and stiffness reduction are likely occurrences. Bearing capacity failure due to soil softening was observed during earthquakes such as the 1999 Izmit earthquake [9], the 1989 Loma Prieta earthquake [10], the 2001 Bhuj earthquake [11,12]) and the 2004 Great Sumatran earthquake [13] among others. In a laboratory environment, the strength loss and deformation behavior due to undrained cyclic loading is often studied by conducting post cyclic monotonic load tests on soil samples, with or without a reconsolidation stage [62–65].
The Indo Gangetic Plain (IGP) that lies on the verge of subduction zone of Indian and Eurasian plate is a seismically active, densely populated region with deep soil deposits. A number of critical structures such as rail and highway bridges and nuclear power plants are being built in the region to cater to the large population of the region, and hence the characterization of deep soil deposits in this region are important. In the present study undisturbed soil samples (UDS) from a site in the IGP plane are examined for static, cyclic, and post-cyclic behaviour. The soils tested have fines content varying from about 20% to 80%, and the fines are non-plastic in nature. The modulus reduction and damping ratio of soil at small, medium and large strain range is examined with the UDS by carrying out bender element, resonant column, and cyclic triaxial tests, respectively. Post cyclic behaviour of non-plastic silty soils is also examined with single stage and multistage cyclic triaxial testing. It is observed that the post-cyclic shear strength increases with increase in the non-plastic fines in the soil. The history of cyclic loading preceding the post cyclic tests is found to influence the tangent modulus observed in post cyclic tests. A site specific calibration parameter is proposed to predict pore pressure response using the energy based Green Mitchel Polito (GMP) pore pressure model.
Natural frequency of earthen dams at different induced strain levels
2019, Engineering Geology
Citation Excerpt :
Dams and levees are among the most crucially important engineered structures that serve the society by primarily providing flood protection, hydroelectric power, water for irrigation purposes and as recreation sites (FEMA, 2017). They are stable under normal operating conditions, but are susceptible to failure when subjected to seismic excitations (Krinitzsky and Hynes, 2002; Singh et al., 2005). Not all earthquake events cause significant damage to existing structures; however, the occurrence of near resonance condition, where the predominant frequency of an earthquake excitation is close to the natural frequency of the structure, may result in amplified vibrations that may cause extensive structural damage (Banerjee, 2017; Chakraborty et al., 2018a, 2018b; Chakraborty et al., 2017; Dakoulas and Gazetas, 1985; Gazetas, 1987; Jibson, 2007; Parish et al., 2009; Zhu and Zhou, 2010).
A method is proposed to estimate the degradation of the first natural frequency of vibration of earthen dams with increase in strain levels induced due to seismic events. A synthesised wave, referred to as the ‘sum of sines’, generated by the superposition of sinusoidal waves with frequencies ranging from 0.01 Hz to 25 Hz, is scaled to different peak accelerations and used to simulate seismic excitations at the base of the dam. The natural frequency is then determined by studying the response of the structure in the frequency domain. For this study, four dams with varying complexities of geometry, constituent material properties, and known natural frequencies were at first selected from previously published literature to establish the validity of the proposed method for determining natural frequencies of the dams at small strain levels. Plane strain models of these structures were constructed and analysed, using a commercially available finite element method-based software that is capable of performing time-history analyses. Results from the analyses indicate a good agreement between the natural frequencies predicted using the proposed method and the frequency values reported in the literature for the corresponding structures at small strain levels. The method was later used to determine the strain-dependent natural frequency of a hydraulic-fill dam in North Texas. Twenty-one different earthquake conditions, with different peak ground accelerations, frequency contents, and mean periods, were used to thoroughly validate the applicability of the developed methodology. Numerical analyses indicate that the strain-dependent variation of the first natural frequency follows a similar trend as that obtained using the ‘sum of sines’ excitation when the dam is subjected to widely different earthquake conditions.
Fuzzy evaluation on seismic behavior of reservoir dams during the 2008 Wenchuan earthquake, China
2015, Engineering Geology
Citation Excerpt :
The damage phenomena of dams were the most real and the most direct response to the great earthquake. Thus, the seismic damage investigations on the reservoir dams have always attracted great attentions from researchers all over the world (Elgamal et al., 1990; Tani, 2000; Raghvendra et al., 2005; Ozkan et al., 2006; Yoshikazu et al., 2012; Chen et al., 2014). The reservoir dams suffered from Wenchuan earthquake exhibited rich seismic damage phenomena.
The post-earthquake rapid risk evaluation on the reservoir dams plays an important role in the emergency danger disposal and reinforcement. The fuzzy Data Envelopment Analysis/Assurance Region (DEA/AR) approach was introduced evaluating the seismic behavior of reservoir dams, in which the seismic risk of dam site, the structure characteristics of reservoir dam, and pre-earthquake state of dams were taken as input variables; in the meantime the seismic crack, leakage and deformation of dams induced by the earthquake were taken as output variables. The proposed DEA/AR approach was adopted to evaluate the seismic behavior of 19 reservoir dams suffered from Wenchuan earthquake in Luojiang County, Sichuan Province. The input and output variable values all originated from the post-earthquake field investigation data. And the weight values of DEA/AR model were determined by the entropy weight method which was able to avoid the interference from subjective consciousness. The evaluation results on the seismic behavior obtained from DEA/AR approach were mainly consistent with those from the experts in the field, which verified that the approach was valid and rational. The proposed DEA/AR approach was based on the objective data and the cognitive ability of the model self to give the evaluation results, which was adaptive for the situation that the seismic behaviors of the reservoir dams were not easy to distinguish. A ranking method was recommended to judge the order of quality of seismic behavior of the evaluated reservoir dams. The ranking results could serve for the post-earthquake emergency disposal and reinforcement.
Seismic damage and behavior analysis of earth dams during the 2008 Wenchuan earthquake, China
2014, Engineering Geology
Citation Excerpt :
Besides, the damage of ancillary facility accounted for 60.6% of all the damaged dams. The dam cracks are the most common type among the earthquake damage of earth dams (Krinitzsky and Hynes, 2002; Raghvendra et al., 2005), which could be found in almost all the damaged dams at dam-break risk or at high risk. There were a large number of longitudinal cracks at the dam crest, and most are located in the middle segment of dam crest as well as both sides near the dam slope (see Figure 7).
The May 12, 2008, Ms 8.0 Wenchuan great earthquake, China caused earthquake damages of 2666 earth dams, these damaged dams are in medium and small scale, located in 8 provinces or municipalities, of which 69 dams were at dam-break risk, and 331 were at high-risk. There were 1996 damaged dams located in Sichuan Province, including all the 69 at dam-break risk and 310 at high-risk. Through a detailed scientific investigation, earthquake damage and emergency data were collected, and the isoseismal map of Wenchuan earthquake, with damaged dams marked on it, was given. Also distribution feature of the above 69 dams along the fault zones were discussed. Typical damage types of earth dams were analyzed. And the action of rare great earthquake, poor quality of dam construction, adverse running state of earth dams before earthquake, and frequent aftershocks were concluded as four main reasons of dam damages. Taking 147 dams with complete geometric parameters as statistical samples, some empirical graphic formulas were proposed, to make rapid evaluation of damage degree. Finally, some useful practical measures were provided, to serve the post-earthquake emergency danger removal, also it may improve the seismic performance of dams. In view of the insufficiency of local geology, site effect and soil property data of damaged dams, application of the qualitative results and empirical relationships to other sites, should be very prudent.

View all citing articles on Scopus

View full text