A huge volume of carbon dioxide was released from Lake Nyos in Cameroon on August 26, 1986. Seventeen-hundred people died of asphyxiation. It initially was thought that a volcanic eruption disturbed the carbon dioxide layer in the lake, causing its release.
That theory has largely been discounted and now it’s thought that a large landslide caused the stratification to break and the gas to be released. However, it’s also accepted that if the partial pressure of the carbon dioxide was near saturation before the release, almost any activity might have led to local oversaturation and some gas release.
While we may never know the cause, a team of Swiss scientists is working hard to make sure a similar incident doesn’t occur at Lake Kivu, between Rwanda and the Democratic Republic of the Congo.
Almost 1,500 meters above sea level and covering about 2,400 km2, Lake Kivu has a 500 meter maximum depth. At the bottom lies approximately 250 billion m3 of carbon dioxide and 55 billion m3 of methane. The Rwandan government wants to tap into the methane — to reduce the risk of a deadly gas eruption and to ensure power supplies. Upwards of 2 million people inhabit the shores.
Scientists from the Swiss Federal Institute of Aquatic Science and Technology (Eawag), Dubendorf, Switzerland, have watched developments in the lake in recent years. They have found that gas concentrations are increasing, with a methane rise of up to 20% since the 1970s. They attributed this to an increase in nutrient inputs associated with population growth around the lake, and the introduction of a sardine species that has had a major impact on nutrient cycles.
Presently, due to very high water pressure, the gas is dissolved in the bottom layers. So exchanges between the bottom and surface waters are very limited. What worries the scientists is that if gas concentrations continue to rise and the lake was disrupted, for example by an earthquake or a landslide, then there could be a massive methane and carbon dioxide eruption.
Meanwhile, the Lake Kivu Methane-to-Power project has won $87.8 million in funding from the World Bank. The Rwandan government has just awarded a contract to Murray & Roberts, Johannesburg, South Africa, to construct a pilot gas extraction facility.
The principle behind the project is simple: if a pipe extending into the depths of the lake is installed, water rises spontaneously as a result of the gas bubbles forming in the pipe. At the surface, the water effervesces — like carbonated water from a bottle that has been shaken before being opened. The methane then has to be separated from the carbon dioxide before it can be used.
“It makes sense to use the gas, especially if the risk of an eruption can thereby be reduced at the same time. But because nobody knows exactly how the lake will respond to this extraction, even small-scale pilot studies have to be performed and monitored extremely carefully,” notes Professor Alfred Wüest, head of Eawag’s Surface Waters Department.
There’s already a small operating plant bringing gas-laden water to the surface. The methane is separated and used to run boilers at a brewery. However, the level of monitoring remains unclear.
So Wüest and his team have been asked to oversee planning for methane recovery by the Rwandan government and the Netherlands Commission for Environmental Impact Assessment.
At the end of 2007 they carried out a series of workshops involving international experts to consider the best way to proceed. One question, for example, concerns the depth at which the degassed water should be returned to the lake to prevent stratification disruption. Also under discussion is whether at least some carbon dioxide can be piped back into the deep water, so that greenhouse gas emissions to the atmosphere from methane exploitation are minimized. Another key question is how methane recovery will affect algae growth.
Part of Eawag’s task also involves developing a continuous monitoring program and a computer model for simulating processes in the lake. “I’m confident we will get answers to these questions and much depends on the computer simulation program that is being developed. We expect this to now be complete within one-to-two months,” explains Dr. Martin Schmid, Eawag team member.
Start up of the pilot plant is due either late this year or early in 2009, followed by a couple of years more work before any scale up to full production is considered, he adds.