Atkins | 06 Jun 2008 | Comments

The earthquake that struck China on 12 May 2008 could have aftershocks that reverberate long after the seismic activity has abated. What can engineers learn about the design of ever more resistant dams?

At 7.8 on the Richter scale, the earthquake that hit China’s Sichuan province caused catastrophic damage to infrastructure and resulted in more than 69,000 fatalities. For days, residents of the town of Dujiangyan waited for a second catastrophe to strike, as damage to the 156m concrete-faced, rockfill Zipingpu dam triggered fears that it was on the verge of more extensive failure.

“Few large dams have been exposed to such severe shaking and people will pore over the performance of the dam for some time to come,” says chairman of the British Dam Society Jonathan Hinks.

Read the signs

“Every incident is learned from and there are already some early findings,” says Professor Andy Hughes, Atkins’ director of dam engineering. “The reservoir was only a third full, but there was cracking and obvious leakage,” he says.

The dam holds 1.1 billion cubic metres of water, but was only holding 320 million cubic metres at the time of impact, greatly reducing the dynamic loading on the structure. But, despite this, the shallow depth and severity of the quake caused damage serious enough to raise concerns that the structure could fail and unleash torrents of water, further devastating areas downstream.

Such concrete-faced rockfill dams are designed as a concrete slab, or series of concrete slabs, that sit on a rubble transition zone above the mass of rock at the heart of the dam. This creates a water-tight membrane. They are generally expected to perform well in earthquakes and considered to have less risk of failure than earth-filled structures, as the rock medium allows some water seepage. However, because of the enormous quantity of materials required to build a dam, the design and the type of fill is often dictated by what materials are available locally, rather than by performance criteria alone.

“The approach has changed a lot over the years,” says Hinks. “There’s now much greater awareness of the dangers of seismic liquefaction in dams and foundations, and consequently the avoidance of low density sands. Hydraulic fill has fallen out of favour.”

Seismic liquefaction occurs when the fill material in the dam loses strength and stiffness under earthquake loading. It happens in water-logged (saturated) soils as forces from the movement generated by the earthquake cause spaces between the grains of soil or sand to collapse, leading to the fill no longer being able to sustain the pressure from overlying soil or structures. This fill material then flows like a liquid and can cause extensive damage.

Hinks also sits on the seismic committee of the International Committee of Large Dams (ICOLD), a forum of 88 countries that work together to improve technical guidance and experience in dam engineering.

“We will next meet in Bulgaria in June. Improvements in practice often follow a close study of incidents and failures, both worldwide and in the UK, where post-incident reporting is likely to be made mandatory,” he says.

This goal is supported by Professor Hughes, who is working with the Environment Agency on recommendations for improving reservoir safety. Such mandatory information, he says, would be an extremely valuable resource for all designers. This research is about to be submitted to the Department for the Environment Food and Rural Affairs and identifies 49 areas that require further study.

New challenges

One area highlighted is that of climate change and its effects on dam safety. “Issues include more frequent, extreme floods, resulting in more frequent overtopping of dams and perhaps the requirement for increased spillway capacity to accommodate larger floods,” he says. Such changes could require dams to be raised and spillways to be reinforced as design criteria adapt to allow for the increasing severity of storm events.

One dam that could have benefited from such retrospective attention is the Ulley reservoir near Rotherham, UK, which suffered from erosion of the downstream face in June 2007 after storms. The turbulent action of the water over steps that were built in 1873 led to some stones being moved and undermined the dam itself.

The age of assets is a problem, particularly in the UK, where the average dam is 110 years old. This, combined with new legislative requirements to produce flood hazard maps along with risk management plans (EU Floods Directive 2007), and requirements to quantify individual dam risk and establish contingency measures (Civil Contingencies Act 2004), is leading to new responsibilities for engineers.

“The biggest challenge for the profession is not technical – we can solve those issues – it is about knowledge transfer and succession planning,” says Hughes. The lack of major new dam construction in the UK is considered to be one reason that few young engineers are specialising in dams and hydraulic engineering. In China and India, where a wealth of major dams, mainly for hydroelectric power generation, are under way, there is no such problem. But such schemes are also contentious, with the anti-dam lobby arguing that they should not go ahead.

“There is also a requirement for engineers to be much broader in their knowledge and views and not hide behind the technical design. You have to be an environmentalist, an accountant and a politician,” says Hughes. To this end, he is assisting the University of Bristol in teaching undergraduate courses in dam engineering and supporting research initiatives. “We are promoting dams in a world where there is a desperate need for water, power, food and irrigation,” he adds. “Engineers must talk the same language as environmentalists. Better training through a good centre for learning seems a good place to start.”

Download PDF