The Chinese government has approved a number of large dam cascades in the south-west of China. But not everyone agrees that the plans are safe.
Earlier this year, Canadian NGO Probe International published a report called “Earthquake Hazards and Large Dams in Western China”, by John Jackson, a pseudonym. Probe International describes the author as a “geologist with a detailed knowledge of western China who must remain anonymous”. He reportedly has four decades experience studying earthquakes and seismic faults.
For the report, Jackson looked at 130 sites in western China where dams have been built, are under construction or proposed, cross-referencing dam locations with maps of seismic hazard. The rivers affected include the Yarlung Zangbo, Parlung Zangbo, Nu River, Lancang, Yalong, Dadu, Min and Yellow rivers.
He found that 48.2% of the dams were in zones of high to very high seismic hazard, while 50.4% were in zones of moderate seismic hazard. Only 1.4% were located in zones of low seismic hazard. The report argues that the rapid rate of dam construction in areas of high seismic hazard in western China could have disastrous consequences for both the economy and the people.
Since its publication, the report has drawn widespread academic and industry interest in China. In August, Zhang Boting, deputy secretary of the China Society for Hydropower Engineering, issued a written response in which he stressed that safety precautions are taken when dams are built in areas at risk of earthquakes. He added that reservoirs can trigger small earthquakes, releasing energy that might otherwise build up and cause large quakes. In other words, dams actually reduce earthquake risks.
Jackson and Zhang’s positions couldn’t be further apart.
Did a reservoir cause the Wenchuan Earthquake?
What the critics worry about is reservoir-induced seismicity (RIS): the idea that the pressure created by the weight of the reservoir water, and the infiltration of water into fissures can cause rockfalls, landslides and tremors, and even set off more destructive earthquakes. This view has been fiercely debated among scientists both inside and outside China over the last four years, since a link was drawn between the magnitude-eight Wenchuan earthquake in Sichuan in 2008 and the Zipingpu reservoir.
The large Zipingpu reservoir lies on China’s Min River. Between September 2005 and the moment the earthquake struck in 2008, it was filled three times and emptied twice. “A Human Trigger for the Great Quake of Sichuan” was how one article published in US magazine Science in 2009 described the filling of the Zipingpu Reservoir.
Some experts have rejected this analysis, pointing out that water from the reservoir could not have penetrated to the source of the Wenchuan Earthquake, which was 14 kilometres deep, and that RIS would not be able to trigger an earthquake of such magnitude. The most powerful RIS quake ever recorded measured 6.3 on the Richter scale.
In his dismissal of Jackson’s research, Zhang quoted Ji Shaocheng, a professor at Polytechnique Montreal’s Department of Civil, Geological and Mining Engineering, on the link between rock strata and earthquakes: “Strong rock means strong quakes, weak rock means weak quakes, and soft rock means no quakes”. Zhang added that the intrusion of water under high pressure can weaken strong rock and make weak rock soft, so filling a reservoir is very likely to result in weaker earthquakes.
But those who blame Zipingpu for the Wenchuan disaster have their own evidence. In December 2008, the journal Seismology and Geology published a paper by researchers from the China Earthquake Administration’s Institute of Geology, Japan’s National Institute of Advanced Industrial Science and Technology and the Sichuan Earthquake Bureau. That study drew the preliminary conclusion that the filling of the Zipingpu Reservoir did have a clear effect on the Longmenshan central and range-front faults. The paper described the possibility that this had caused the 8.0 magnitude earthquake to occur sooner than it would have otherwise as a “scientific question which should not be avoided and is worth further research.”
In April 2009, Hu Xianming, chief engineer at the Sichuan Earthquake Bureau added to this evidence. He found, based on a study of 262 minor tremors in the Zipingpu Reservoir area between August 2004 and December 2007, that the Wenchuan-Shuimo tremor cluster caused by the reservoir and the subsequent 8.0 magnitude earthquake had the same origin.
Then there is the risk of natural earthquakes – as opposed to those triggered by the reservoirs – hitting the proposed dam cascades on the Jinsha, Lancang, Yalong and Nu rivers.
China’s south-west lies in the Himalayan earthquake region, the world’s most active tectonic zone. According to John Jackson, large dams built in such high-risk areas will be at threat from naturally occurring earthquakes, and going ahead with construction is an extraordinarily risky experiment.
Zhang Boting, meanwhile, says that as long as dams are kept away from the faults themselves, the risks are negligible: “There’s nothing you can do about fault lines, and that’s why dams are always built elsewhere. To date, no dam has failed due to an earthquake anywhere in the world.” He believes that as long as the distribution of earthquakes is known and dams are not built directly over faults, planners can guarantee safety through earthquake-resistant construction.
But seismologist Ma Wentao points out that the maps of earthquake risk currently used by engineers are based on historical information about earthquakes and seismic activity. Due to inaccuracies in earthquake monitoring and a lack of monitoring of active faults, along with a scarcity of historical records in many regions, it is very hard to accurately map faults and estimate earthquake risk, Ma says.
Seismic research "only just getting started"
Research into RIS in China only started in 1963. In 2004, the China Earthquake Administration set up China’s first network of digital seismic monitoring stations designed to monitor RIS across a number of reservoirs in the Wu River basin. The Earthquake Monitoring Regulations, implemented the same year, specifically required that all dams capable of inducing earthquakes of 5.0 magnitude or over – in other words dams more than 100 metres high, or reservoirs of 500 million cubic metres in volume – should have dedicated seismic monitoring networks. In 2007, methodology for RIS risk evaluation was issued by the China Earthquake Standards Technology Commission, the first of its kind.
An expert at the Institute of Geology, who did not wish to be named, says that research on RIS in China is “only just getting started” and mainly focuses on areas of low seismic activity; there have been no in-depth studies in medium or high risk areas. Another geologist familiar with the seismology field in China says that RIS in western China only started to be taken seriously after the Wenchuan Earthquake.
Lei Xinglin, deputy director of the Tectonic Physics Laboratory at the Institute of Geology has written that, in the right circumstances, a reservoir may directly trigger a major quake, and “the risk is greater with big dams of over 100 metres in height.”
Jackson urges the Chinese government to commission third-party research into the risks of building dams in earthquake-prone regions and to consider possible changes to the current plans for cascade development. Otherwise, he says, “in the up to 150-year life span of these large dams, China will be plagued by the earthquake risks brought by this highly dangerous method of developing energy.
“And then, the only solution may be to stop using the dams, with China paying a huge economic price.”
This article was first published in Caixin’s New Century Weekly. Liu Hongqiao won the prize for "Best Young Journalist" for this article at the 2013 China Environmental Press Awards.