Debunking the Debunkers: Dams Really Can Trigger Quakes

Amid the recent hubbub surrounding whether Zipingpu Dam may have triggered the disastrous Sichuan earthquake in May 2008, it's important to clear up a few misconceptions about Reservoir-Induced Seismicity (RIS):

1.No convincing correlation has yet been demonstrated between earthquakes and reservoirs.

False. While scientists lack comprehensive data to conclusively link earthquakes to dams in some potential RIS cases, in a number of other cases with careful RIS studies, a definite cause-and-effect relationship has been found. Best known cases are Koyna, India (1967) and Xinfengjiang Dam, China (1962).

2. If the natural seismicity is low, the danger of RIS is also low.

Only partly true. While hotspot areas with historically unstable geology and fault lines are more likely to experience RIS, areas previously unknown to have seismic activity have also experienced RIS or earthquakes induced by other human activities, such as fluid injections. One example is the Killari reservoir in India, which may have triggered the M6.1 Latur earthquake that leveled 20 villages. Isn't it therefore more important to pay attention to these areas, where there is less earthquake preparedness in both dam designs and in local communities?

3. Only a few large dams out of 11,000 large dams worldwide have experienced significant RIS.

Again false. More than a few large dams and their associated earthquakes are convincingly RIS cases (to name "a few": Koyna, India; Lake Oroville, US; Kariba, Zambia; Aswan, Egypt; Srinagarind, Thailand; Xinfengjiang, China, Hoover, US).

4. No dam has yet failed disastrously because of RIS, so the danger is grossly exaggerated.

True, but… Both Koyna and Xinfengjiang came incredibly close to failure during each RIS event, and both would have unleashed huge killer flood waves. Given our current state of understanding about RIS (which, while limited, isn't as limited as the dam-industry wants us to think), no new reservoirs anywhere in the world should be declared free from danger.

Even with our limited knowledge on RIS (which can be partly attributed to a falling off of funding since the 1960s), scientists recognize the key factors controlling the probability of RIS, including water volume, height of water column, local geology, and water level fluctuations.

The important thing now is to practice "the precautionary principle." One good example of this is the Auburn Dam in California. A 1975 quake triggered by Oroville Dam, which was built on a fault previously thought inactive, led to the suspension of Auburn Dam, then being constructed in a nearby watershed. If Auburn Dam had been built, and failed, it would have overwhelmed Folsom Dam downstream and completely flooded Sacramento. Even with the suspension of Auburn, the Oroville region isn't safe from future RIS and subsequent dam failure. If Oroville were to fail, more than 90,000 people in the area would be at risk. California, I hope you're ready.