With the increasing number of dams and the age of older dams, the issues of dam safety are becoming more prominent. The potential dam safety problem is not only a security issue but also a an issue of risk for areas downstream of dams. Therefore, risk management is the core of dam safety administration. Dam risk classification is an important part of risk management that should consider the probability of dam failure and the risk of loss (potential damage). This paper summarizes research on dam risk classification and finds that ‘three degrees’ and ‘four degrees’ of dam risk classification are used widely in developed countries or districts. It proposes that four degrees of risk could be adopted by the risk matrix approach for the current situation and relevant regulations of China. It might make the dam safety management more efficient and have a certain guiding significance in establishing the standard of dam risk classification in China.

INTRODUCTION

Construction of reservoirs and large dams is included in the aim to meet water resources management requirements including flood control, irrigation, water supply, hydropower, navigation, recreation, fisheries, etc. According to the Bulletin of First National Census for Water Conservancy of China (2013), the number of reservoirs totaled 98,002 with a combined storage capacity of 932.312 billion m3. Statistics show that most reservoirs in China were built in 1950–1973, and the number of reservoirs increased linearly from 23 to 72,131 in that period (Liu & Zhang 1999). More than 95% of all dams in China are of the embankment type (Ru et al. 2001), and most of the reservoirs are small in scale (capacity of reservoir <107 m3). For instance, at the end of 2012, there were 93,308 small-scale reservoirs (95.2% of the total). In general, the average lifespan of a small-scale embankment dam is approximately 50 years (Wieland 2010). Therefore, dam safety should be paid much attention on account of the increasing number of dams, coupled with the older dams in disrepair. There were 3,523 incidents of dam failure from 1954 to 2013 (He et al. 2008; Zhao 2014) that caused significant loss of life and economic losses, such as the failure of Banqiao and Shimantan dams which resulted in more than 200,000 fatalities in 1975 (Goldstein 2011), and the failure of Gouhou dam in 1993 (Li 1994). However, the potential dam breach is a ‘risk’ problem rather than a ‘safety’ issue.

People usually deem that the risk associated with a high dam or a large reservoir is high, but that is not the case, because higher safety standards will have been used in the design of a high dam and a large reservoir. The probability of them failing is lower than for small dams and the risk is also correspondingly lower. Dam risk encompasses the probability and consequence of failure, so engineers are more concerned with the highest risk dam rather than the highest dam or the largest reservoir. In order to manage dams effectively and ensure dam safety, the dam risk can be classified, based on an analysis of the possible ways in which dams can fail and the scope of the resulting flood, considering the dam failure modes, failure probability and the consequence. This paper introduces the status and progress of dam risk classification, in light of the actual situation in China, and puts forward the proposed risk level from a qualitative point of view.

RESEARCH STATUS ON DAM RISK CLASSIFICATION

Research status around the world

Many countries and states have classified the dam hazard by laws, regulations and policies. According to their actual situations, different countries or districts adopt different classifications, including ‘two degrees’, ‘three degrees’, ‘four degrees’ and ‘five degrees’. The most widely used are three degrees (‘high’, ‘significant’ and ‘low’) and four degrees (‘very high’, ‘high’, ‘significant’ and ‘low’). In the USA, more than 60% of states have adopted three degrees classification in light of the Federal Emergency Management Agency National Dam Safety Program (FEMA 1996, 2004). However, Montana and Georgia have adopted two degrees classification; California, Colorado, New Hampshire, Ohio, Virginia and West Virginia have adopted four degrees; and Connecticut has adopted five degrees (FEMA 2004). In Europe, classifications based on the height of dams and the capacities of reservoirs are adopted for the legislation of most countries (Dalliou 1998; Icold European Club 2001). Norway, Portugal and Spain have adopted three degrees (NVE 2000; PRSD 1990; Jesus Penas 2000). Sweden has adopted four degrees (Norstedt et al. 1998). Some countries stipulate that the elaboration of emergency action plans should be made by dam-break analysis, such as Finland, Germany, Italy, etc. (Loukola et al. 1998; Giesecke et al. 2001; Rogen 2003).

Research status in China

Risk analysis, evaluation and management in China started later than in the developed countries. The 20th International Conference of Large Dams was held in Beijing in 2000, and this has greatly accelerated the research on dam risk in China. Ma et al. (2008) looked at specific features of earth and rock fill dams, analyzed the dam failure modes, established the risk early-warning index system and applied it in actual projects. Li & Wang (2006) based their work on foreign risk analysis theories and methods in combination with the Shahe Dam (in Chuzhou, Anhui province), and conducted a systematic risk analysis and safety evaluation of dams in China. The work of Fang (2006) was based on the calculation of the overall risk of a dam and proposed that China adopt four degrees: low risk (IV), medium risk (III), high risk (II) and ultrahigh risk (I). The Dam Safety Assessment Guidelines (SL258-2000) (The Industry Standard of People's Republic of China 2001) consider the dam characteristics for each special security classification and classify the dams using three degrees: Category I (safe and reliable), Category II (basically secure), Category III (unsafe), which is applied to dilapidated reservoirs which require reinforcing. According to the severity of dam risk, the Dam Risk Assessment Guidelines (The Industry Standard of People's Republic of China 2012) adopt four degrees: low risk (IV), medium risk (III), high risk (II) and ultrahigh risk (I).

DETERMINATION OF DAM RISK CLASSIFICATION

Failure modes and failure probability

Failure modes

According to the dam-break statistical data of China from 1954 to 2006 (Xie et al. 2009) and The National Register of Reservoir Collapse (1991), the modes of dam failure can be summarized in Table 1.

Table 1

The main dam failure modes

Failure modeNumberPercentage (%)Annual probability of failure (10−5)
Overtopping Excessive flood 440 12.58 8.47 
Inadequate spillway 1,352 38.65 26.03 
Dam quality Seepage 903 25.81 17.39 
Quality of dam body 50 1.43 0.96 
Landslide 113 3.23 2.18 
Spillway, culverts 232 6.63 4.47 
Sliding of dam foundation or collapse 0.17 0.12 
Burrows 0.11 0.08 
Mismanagement 168 4.80 3.23 
Others 179 5.12 3.45 
Unknown 51 1.46 0.98 
Total 3,498 100 67.35 
Failure modeNumberPercentage (%)Annual probability of failure (10−5)
Overtopping Excessive flood 440 12.58 8.47 
Inadequate spillway 1,352 38.65 26.03 
Dam quality Seepage 903 25.81 17.39 
Quality of dam body 50 1.43 0.96 
Landslide 113 3.23 2.18 
Spillway, culverts 232 6.63 4.47 
Sliding of dam foundation or collapse 0.17 0.12 
Burrows 0.11 0.08 
Mismanagement 168 4.80 3.23 
Others 179 5.12 3.45 
Unknown 51 1.46 0.98 
Total 3,498 100 67.35 

From Table 1, it can be seen that the main failure modes are flood overtopping and seepage. Penetration is one of the main types of failure of embankment dams, resulting in piping, movement of soil and erosion.

Failure probability

Failure probability can be calculated using statistics and the empirical estimation method. The empirical estimation method is where specialists analyze the stability of various parts of the dam hub in term of their experience and convert the possibility of an event occurring to quantitative probability. Vick (1992), Barneich et al. (1996) and USBR (1999) have proposed relationship tables between qualitative description and quantitative probability. The method is simple, but there is a lot of subjectivity. Different specialists may give different probabilities of the same event occurring.

From a statistical point of view, failure probability (FP) can be used instead of annual probability of failure (APF) and can be calculated by Equation (1) (Chen 2005): 
formula
1
where Nf is the number of failed constructions, Na is the total number of constructions and T is the statistics time interval (year).

Failure path analysis

This method is based on the potential loads, and analyzes the potential failure modes of spillways, sluices, discharging tunnels, etc. According to the failure order, there can be a failure path such as: loads, construction, partial failure, overall instability. The failure path analysis is widely used by the event tree (Foster et al. 2002), the fault tree (Li et al. 2001) and the Bayesian network analysis (Li et al. 2007), etc.

Risk of loss and consequences

Loss of life

Developed countries have carried out extensive research in the field of loss of life, but they are based on experience and statistics. The method of estimating loss of life can be summarized as follows: 
formula
2
where LOL is loss of life; PAR is population at risk; Wt is warning time; and F is flood intensity (Dekay & Mcclelland 1993). 
formula
3
where LOLu is loss of life in an area; PARu is population at risk in an area; and Pr is probability of survival (Assaf et al. 1997). 
formula
4
where LOL is loss of life; PAR is population at risk; and f is life loss rate (Graham 1999). 
formula
5

where LOL is loss of life; PAR is population at risk; i is the impact factor of flood severity; and c is the correction factor (Reiter 2001).

LiFESim is a modular, spatially-distributed, dynamic simulation system for estimating potential loss of life from dam failure floods and is designed to serve multiple functions (Aboelata & Bowles 2005).

Economic loss

The economic loss includes the direct and the indirect. There is much more research on economic loss in China. The calculation method is shown in Table 2.

Table 2

The calculation method of economic loss

TypesBasisFormulaeScope of application
Direct loss Loss of projects Direct access from the project budget Projects loss 
Inundation losses Loss rate 
  • S = economic loss

  • βijk = the loss rate of ij–type property in the kth risk area

  • Wijk = the value of ij–type property in the kth risk area

  • n = the total number of categories of property

  • m = the total number of categories of the ith property

  • l = the total risk area

 
Fixed assets, current assets 
Size and other indicators 
  • Aijk = the size of jk–type in the ith extent of destruction

  • fijk = the repair cost of jk–type in the ith extent of destruction

  • n = the total number of facilities category

  • m = the total classification of the ith facility

  • l = the total degree of destruction

 
Railways, highways, pipelines, high-voltage power grid, water channels, dams, housing and other facilities 
Interruption time 
  • Tijk = the time of i–j–type in the kth economic activity

  • Sijk = the loss value of i–j–type in the kth economic activity

  • n =the number of departments

  • m =the number of types in the ith industry

  • l = the number of types in the ith department and the jth industry activity

 
The losses of industry, commerce, railways, highways, shipping, electricity, water and other sectors of economic activity caused by the interruption 
Loss type of agricultural income 
  • S0 = the loss in the year/quarter

  • Rc = the costs of restoration

  • Il = the loss in the time of recovery

 
The agricultural losses caused by flood damage 
Loss of engineering facilities destroyed 
  • V0 = the value before the disaster

  • VR = the increase cost of replacement

 
The engineering of construction, water conservancy, the municipal facilities, etc. 
Indirect losses Direct estimation Analysis of flood submerged area and estimation of various indirect economic losses The losses of enterprises, contingency fees, etc. 
Coefficient 
  • Sli = the indirect loss of the ith department/activity

  • Sdi = the direct loss of the ith department/activity

  • ki, bi = coefficient

 
Sampling from different regions 
TypesBasisFormulaeScope of application
Direct loss Loss of projects Direct access from the project budget Projects loss 
Inundation losses Loss rate 
  • S = economic loss

  • βijk = the loss rate of ij–type property in the kth risk area

  • Wijk = the value of ij–type property in the kth risk area

  • n = the total number of categories of property

  • m = the total number of categories of the ith property

  • l = the total risk area

 
Fixed assets, current assets 
Size and other indicators 
  • Aijk = the size of jk–type in the ith extent of destruction

  • fijk = the repair cost of jk–type in the ith extent of destruction

  • n = the total number of facilities category

  • m = the total classification of the ith facility

  • l = the total degree of destruction

 
Railways, highways, pipelines, high-voltage power grid, water channels, dams, housing and other facilities 
Interruption time 
  • Tijk = the time of i–j–type in the kth economic activity

  • Sijk = the loss value of i–j–type in the kth economic activity

  • n =the number of departments

  • m =the number of types in the ith industry

  • l = the number of types in the ith department and the jth industry activity

 
The losses of industry, commerce, railways, highways, shipping, electricity, water and other sectors of economic activity caused by the interruption 
Loss type of agricultural income 
  • S0 = the loss in the year/quarter

  • Rc = the costs of restoration

  • Il = the loss in the time of recovery

 
The agricultural losses caused by flood damage 
Loss of engineering facilities destroyed 
  • V0 = the value before the disaster

  • VR = the increase cost of replacement

 
The engineering of construction, water conservancy, the municipal facilities, etc. 
Indirect losses Direct estimation Analysis of flood submerged area and estimation of various indirect economic losses The losses of enterprises, contingency fees, etc. 
Coefficient 
  • Sli = the indirect loss of the ith department/activity

  • Sdi = the direct loss of the ith department/activity

  • ki, bi = coefficient

 
Sampling from different regions 

Social and environmental impacts

The impacts of dam failure on society and the environment are wide and complex, but there have been few studies on this subject. According to the actual situation, Li et al. (2006) proposed that the index of social and environmental impacts can be calculated by 
formula
6
where f is the index of social and environmental impacts; N is the index of population at risk; C is the important factor of the city; I is the important factor of the facility; h is the index of relics; R is the river's form factor; l is the biological environment factor; L is the cultural landscape coefficient; and P is the industrial pollution factor.

Risk classification

The research on dam risk classification is limited. It is mentioned that dam risk can be classified into four degrees in the Dam Risk Assessment Guidelines (The Industry Standard of People's Republic of China 2012) in China, but it is unclear how the categories should be divided there is no clear criteria and easily operable method. This paper grades the failure probability and the risk of loss, and adopts the risk matrix approach to grade dam risk. Different risk classifications of dams should adopt corresponding actions to improve the efficiency of dam management. The flowchart of dam risk classification and management actions is shown in Figure 1.

Figure 1

The process of dam risk classification and management actions.

Figure 1

The process of dam risk classification and management actions.

Failure probability classification

Dam failure probability classification is based on the probability theory that is widely used in the aerospace, military and other fields. This paper proposes that the dam failure probability is divided into five degrees in the light of China's actual situation and studying other industry standards, as specified in Table 3.

Table 3

The proposed failure probability classification

PossibilityRareUnlikelyPossibleLikelyVery likely
Classification 
Probability of failure ≤ 10−5 10−5 to 10−4 10−4 to 10−3 10−3 to 10−2 ≥ 10−2 
PossibilityRareUnlikelyPossibleLikelyVery likely
Classification 
Probability of failure ≤ 10−5 10−5 to 10−4 10−4 to 10−3 10−3 to 10−2 ≥ 10−2 

Risk of loss (consequence) classification

It should be simple, intuitive and an easy operation to estimate the risk of dam failure. The consequences include loss of life, economic losses and environmental impacts. According to the Production Safety Accident Reporting and Investigation Regulations, Decree No. 493 (State Administration of Production Safety Supervision and People's Republic of China 2007), the risk of loss can be classified into five degrees, as shown in Table 4.

Dam risk classification and management actions

This paper adopts the risk matrix approach and refers to the Standard for Classification of Risk Grade of Landslide Lake (SL450-2009) (Ministry of Water Resources, People's Republic of China 2009), which, suggests the dam risk classification employ four degrees that are: ultrahigh risk (I), high risk (II), significant risk (III) and low risk (IV), as shown in Table 5. Risk management should be regarded as an ongoing and iterative process that needs to adapt to new information. Different risk classifications of dams should take corresponding management actions that are shown in Figure 1.4

Table 4

The proposed dam risk of loss (consequence) classification

Describe the consequences
Consequence classificationLoss of lifeEconomic lossEnvironment impact
Insignificant No fatalities or less than three seriously injured people Less than 1 million RMB in direct losses No impact or, if impact occurs, then not to an extent that would draw concern from a reasonable person 
Minor Less than three fatalities or 3–19 seriously injured people 1–5 million RMB in direct losses Impact occurs but not to the extent that it would impair the overall environmental values, population or community 
Moderate 3–20 fatalities or more than 20 seriously injured people 5–50 million RMB in direct losses Significant impact on environmental values extending locally or to a species or community, but not on a population scale. Recovery periods of 10–20 years anticipated 
Major 20–50 fatalities 50–500 million RMB in direct losses Significant impact on environmental values within the Project area or to a species or community on a population scale. May lead to a local extinction or recovery periods of 20–50 years are likely 
Catastrophic More than 50 fatalities More than 500 million RMB in direct losses Impact on environmental values over a wide area or impact resulting in the extinction of a population or community or recovery periods of greater than 50 years likely 
Describe the consequences
Consequence classificationLoss of lifeEconomic lossEnvironment impact
Insignificant No fatalities or less than three seriously injured people Less than 1 million RMB in direct losses No impact or, if impact occurs, then not to an extent that would draw concern from a reasonable person 
Minor Less than three fatalities or 3–19 seriously injured people 1–5 million RMB in direct losses Impact occurs but not to the extent that it would impair the overall environmental values, population or community 
Moderate 3–20 fatalities or more than 20 seriously injured people 5–50 million RMB in direct losses Significant impact on environmental values extending locally or to a species or community, but not on a population scale. Recovery periods of 10–20 years anticipated 
Major 20–50 fatalities 50–500 million RMB in direct losses Significant impact on environmental values within the Project area or to a species or community on a population scale. May lead to a local extinction or recovery periods of 20–50 years are likely 
Catastrophic More than 50 fatalities More than 500 million RMB in direct losses Impact on environmental values over a wide area or impact resulting in the extinction of a population or community or recovery periods of greater than 50 years likely 

Table 5

The proposed dam risk classification

Dam risk of loss
Dam risk classificationInsignificant AMinor BModerate CMajor DCatastrophic E
Dam failure probability Rare II II 
Unlikely II II III 
Possible II II III IV 
Likely II II III III IV 
Very likely II III III IV IV 
Dam risk of loss
Dam risk classificationInsignificant AMinor BModerate CMajor DCatastrophic E
Dam failure probability Rare II II 
Unlikely II II III 
Possible II II III IV 
Likely II II III III IV 
Very likely II III III IV IV 

CONCLUSION

Dam risk classification is intended to rank dams for the purpose of making the management of dams more effective. This paper summarizes the status of dam risk classification research, and finds that three degrees (high, significant, low) and four degrees (very high, high, significant, low) are widely adopted to classify dam risk or the potential hazard in developed countries, such as America, Canada, Austria, the UK, etc. According to the actual situation and relevant regulations of China, this paper advises that the probability of dam failure adopts five degrees and the risk of loss adopts four degrees. On this basis, the dam risk classification of four degrees is proposed (ultrahigh (I), high (II), significant (III) and low (IV)) by the risk matrix approach. It may assist in dam safety management and decision-making.

ACKNOWLEDGEMENT

The support of China National Basic Research Program under grant 2013CB036403 is gratefully acknowledged.

REFERENCES

REFERENCES
Aboelata
M.
Bowles
D. S.
2005
LIFESim: A Model for Estimating Dam Failure Loss. Report to Institute for Water Resources
.
US Army Corps of Engineers and Australian National Committee on Large Dams by Institute for Dam Safety Risk Management
,
Utah State University
,
Logan, Utah, USA
.
Assaf
H.
Hartford
D. N. D.
Cattanach
J. D.
1997
Estimating dam breach flood survival probabilities
.
Ancold Bulletin
107
,
23
42
.
Barneich
J.
Majors
D.
Moriwaki
Y.
Davidson
R.
,
1996
Application of Reliability Analysis in the Environment Impact Report and Design of a Major Dam Project
,
Uncertainty ‘96’, ASCE
.
Bulletin of First National Census for Water Conservancy of China
2013
Ministry of Water Resources of the People's Republic of China, the National Bureau of Statistics of the People's Republic of China. http://www.mwr.gov.cn/2013pcgb/merge1.pdf
.
Chen
Z.
2005
Water Conservancy Project Risk Analysis and Reliability Design Method
.
IWHR
,
Beijing, China
,
p. 114 (in Chinese)
.
DeKay
M. L.
McClelland
G. H.
1993
Predicting loss of life in cases of dam failure and flash flood
.
Risk Analysis
13
(
2
),
193
205
.
Fang
W.
2006
Dam safety evaluation and grading methods
. In:
International Seminar on Hydropower
,
23 October, 2006, Kunming, China
.
FEMA
1996
(US Federal Emergency Management Agency) National Dam Safety Program
.
FEMA Martins
2004
Federal Guidelines for Dam Safety: Hazard Potential Classification System for Dams, National Dam Safety Program. http://www.fema.gov/fima/damsafe/fema-333-toc.shtm
.
Foster
M. A.
Fell
R.
Davidson
R.
Wan
C. F.
2002 Estimation of the probability of failure of embankment dams by internal erosion and piping using event tree methods
.
Ancold Bulletin
121
,
75
88
.
Giesecke
J.
Glasebach
H. J.
Muller
U.
2001
German Standardization in Dam Construction
.
Dams in Germany
78
.
Goldstein
L.
2011
“Safe” Power an Oxymoron: Goldstein
.
Toronto Sun
,
Graham
W. J.
1999
A procedure for estimating loss of life caused by dam failure, US Department of the Interior, Bureau of Reclamation, Dam Safety Office
.
He
X. Y.
Wang
Z. Y.
Huang
J. C.
2008
Temporal and spatial distribution of dam failure events in China
.
International Journal of Sediment Research
23
(
4
),
398
405
.
Icold European Club
2001
Dam Legislations
,
Paris
.
Jesus
Penas
2000
Spanish Legislation for Safety of Dams. Technical Meeting “Lefislacao sobre Seguranca de Barragens”, National Laboratory of Civil Engineering, LNEC, 2000 Lisbon, Portugal
.
Le Dalliou
P.
1998
Dam legislations in some European Countries
. (
Bureau d’étude Technique et de Contrôle des Grands Barrages
.)
In: Dam Safety
(
Berga
L.
, ed.).
Balkema
,
Rotterdam
.
Li
J.-C.
1994
Gouhou Dam and Analysis for Causes of the Dam Failure
.
Chinese Journal of Geotechnical Engineering
6
.
Li
L.
Wang
R.
2006
Dam Risk Assessment and Risk Management
.
China Water & Power Press
,
Beijing
(in Chinese)
.
Li
D.
Yan
L.
Shao
D.
2007
Reliability evaluation of earth-rock dams using Bayesian network
.
Engineering Journal of Wuhan University
40
(
16
),
24
29
(in Chinese)
.
Li
X. B.
Jiang
W. D.
Zhao
F. J.
2001
Fault tree analysis of tailings dam break during flood season [J]
.
Journal of Safety and Environment
5
,
010
.
Liu
B.
Zhang
J.
1991
Chinese construction of reservoirs and flood control, Yangtze River, 1 (in Chinese)
.
Loukola
E.
Maijala
T.
1998
New dam safety guidelines in Finland
. In:
Dam Safety
(
Berga
L.
, ed.).
Balkema
,
Rotterdam
.
Ma
F.-H.
He
X.-W.
Wu
G.-Y.
2008
Risk early-warning index system for earth and rock fill dams [J]
.
Chinese Journal of Geotechnical Engineering
30
,
11
(in Chinese)
.
Ministry of Water Resources, People's Republic of China
2009
Standard for Classification of Risk Grade of Landslide Lake (SL450-2009) (in Chinese)
.
Norstedt
U.
Andersson
R.
Eriksson
H.
Lindh
A.
Markland
L.
1998
Swedish dam safety guidelines
. In:
Dam Safety
(
Berga
L.
, ed.).
Balkema
,
Rotterdam
.
NVE
2000
The Norwegian Regulations for Planning, Construction and Operation of Dams, 1981/86; Regulations Governing the Safety and Supervision of Watercourse Structures
,
Oslo
.
Portuguese Regulations for Safety of Dams PRSD
1990
Decree-law nr. 11/90
,
Lisbon
.
Reiter
P.
2001
Loss of Life Caused by Dam Failure: The RESCDAM LOL Method and its Application to Kyrkosjarvi Dam in Seinajoki
.
PR Water Consulting Ltd
,
Helsinki
.
Rogen
M.
2003
The Regulation of Dams in Italy
.
High Council of Public Works
,
Rome
.
Ru
N.
Niu
Y.
2001
Dam Accidents and Safety: Embankment
.
China Water Power Press
,
Beijing, China
,
10
12
.
State Administration of Production Safety Supervision and People's Republic of China
2007
The Production Safety Accident Reporting and Investigation Regulations (State Council Decree No. 493) (in Chinese)
.
The Industry Standard of People's Republic of China
2001
Dam Safety Assessment Guidelines
(SL258-2000).
China Water & Power Press
,
Beijing
(in Chinese)
.
The Industry Standard of People's Republic of China
2012
Dam Risk Assessment Guidelines (draft)
,
Ministry of Water Resources of the People's Republic of China (in Chinese)
.
The National Register of Reservoir Collapse
1991
Collapse, Construction and Management Division
,
People's Republic of China Ministry of Water Resources (PRCMWR)
.
USBR
1999
Dam Safety Risk Analysis Methodology
.
Versions 3.3, Technical Service Center
.
Vick
S. G.
1992
Risk in Geotechnical Practice in Geo-Technique and Natural Hazards
.
Vancouver Geotechnical Society and Canadian Geotechnical Society
,
Vancouver, Canada
, pp.
41
57
.
Wieland
M.
2010
Life span of storage dams
.
International Water Power and Dam Construction
62
(
2
),
32
35
.
Xie
Jiabi
Sun
Dongya
2009
National statistic of dam break and analysis
.
Water Resources and Hydropower Engineering
40
(
12
),
124
128
.
Zhao
X.
Wang
Z.
Shen
J.
2014
Statistics of small dam failure and result classified analysis [J]
.
China Water Conservancy
64
(
10
),
33
35
.