III - TAILING DAMS
Tailing dams differ from classical dams in many respects:
 Their design (if any) is different.
Construction, filling and operation are simultaneous.
Overtopping by large floods is unlikely.
The corresponding risks may be high: the probability of sliding, piping or earthquake liquefaction is higher than for classical dams, failure is often instantaneous, warning systems are not very effective, and mud may be more dangerous than water. In the U.S., Europe and Japan, 5 tailings-dam failures have caused a total of 1,000 victims in the last 30 years, whereas 30 failures of classical dams more than 15 m high caused a total of 100 victims (and some small dam failures a few hundred victims).
Risk assessment of tailing dams may be more difficult than for classical dams, and costly structural measures will be more often justified.
The same construction methods as for tailings dams have been used for many small dams in the Yellow River basin in China for centuries, as the loess content of the water is extremely high; experience in this technique is now very extensive, especially as a result of dam building in the last 50 years.
IV - ORGANIZATION OF RISK ANALYSIS
Dam owners, consultants, or regulatory authorities have been carrying out some kind of risk analysis for many years. More sophisticated methods have been developed recently.
For very large dams, financial resources may be available for extensive risk analysis to be carried out by the owner, an engineering company, or a specialized consultant. The cost may be in the area of USD 100,000 and sometimes much more. It may be wholly justified for very large dams. Such analysis may also be carried out by the regulatory authorities in some countries.
But there are tens of thousands of medium or small dams throughout the world; a substantial percentage of them may represent serious risks and most owners have low revenues and no knowledge of technical problems or the possible risks. The regulatory bodies which leave the entire responsibility to the owners should offer services and advice, and possibly impose a minimum level of risk analysis. Cost is not the key problem: if organized by a single team for a number of dams and limited to just a few main risks, the time spent may be limited to few man- days per dam to identify the dams at risk, and possibly to one man-month/dam to have a reasonable analysis of the risk. A few years ago such a method was used in France for hundreds of dams in the 15-30 m height range, and proved to be very cost-effective.
In the U.K., the safety of each dam is the responsibility of a single person chosen from amongst experienced engineers: that person is in charge of carrying out the risk analysis which may be considered necessary.
Consequently, it seems that the problem of risk-analysis effectiveness is more a problem of organization and adaptation to dam size and to the circumstances of the main risks than a problem of cost.
CONCLUSION
In the next century, one billion people will be living downstream of dams. On the basis of past data, millions may be at risk within the next fifty years due to dam failures or to large floods more or less well controlled by dams. And tens of thousands of people will be killed or permanently disabled when working in dam construction; risk analysis is therefore absolutely necessary. But what sort of analysis, by whom, for which dams, at what cost, and with what effectiveness?
Analysis may cover all possible risks, and be carried out by specialized consultants; the corresponding cost-which is justified for very large dams-may be high. But for global analysis of many small or medium dams, limited to the small number of main risks, the cost per dam may be limited to several man-days to identify dams at risk and several man-weeks to quantify the risk and suggest improvements. As most owners of medium and small dams have no knowledge of the risks involved, such analysis should be organized or possibly imposed by the regulatory authorities.
The cost-effectiveness of risk analysis may be very good for flood failures which represent half of the total failure risk, but the analysis should refer to the Imminent Failure Flood and not to the Design Flood; cost-effectiveness may also be very good for the impact of dams on all large floods. On the other hand, risk analysis may be not so effective for earthquakes or for some dams subject to sudden failures such as tailings dams, hydraulic fill dams, or old masonry dams.
The usual guidelines pay a great deal of attention to physical problems and mathematical and statistical work, whereas in fact human behaviour is the key element to be taken into account when assessing the causes of failure under normal circumstances (first filling, ageingmaintenance and monitoring-, work accidents). Consideration of human behaviour is also essential when evaluating the consequences of all failures: well organized emergency planning and early-waming systems could divide the number of victims by ten. The study of hum an behaviour is the basis of risk analysis of many other human activities-air transport, the petroleum or nuclear industries, etc.-, but is not yet applied to dams at the scale it may deserve.
Previous < 1 2 3 4 > |