Geology : How dangerous is Africa’s explosive lake??
On 22. May 2021, Nyiragongo, one of Africa’s most active volcanoes, began spewing lava toward the densely populated city of Goma in the Democratic Republic of Congo (DRC). The eruption destroyed several villages, claimed dozens of lives and forced an estimated 450,000 people to flee their homes.
Since then, the volcano has calmed down again and the immediate humanitarian crisis has subsided. But authorities and scientists are worried about something else, and it could be more dangerous than the Nyiragongo.
Goma is located on the shores of Lake Kivu, which is a geological feature: It contains 300 cubic kilometers of dissolved carbon dioxide, 60 cubic kilometers of methane, and toxic hydrogen sulfide to boot. The picturesque lake, located between the DR Congo and Rwanda, could release these gases explosively in what is known as a limnic eruption, a rare phenomenon. This would abruptly release a huge amount of greenhouse gases into the atmosphere: The lake contains the equivalent of 2.6 gigatons of CO2, or about five percent of annual global greenhouse gas emissions. But an even greater disaster is looming: if the surrounding valley is flooded with the choking, toxic gases, millions of people could die as a result. "This could trigger one of the worst, if not the worst, humanitarian disasters in history," says engineer Philip Morkel. He is founder of Vancouver-based Hydragas Energy and is trying to raise funds to remove the gas from the lake and make it usable.
The 2021 volcanic eruption did not trigger mass degassing of the lake, and on 1. In June, the Rwanda Environment Management Authority (REMA) said there was no immediate danger. However, lava is believed to have flowed through underground fissures beneath the city of Goma and Lake Kivu. A day after the eruption, an earthquake apparently caused a sandbank at the lake to partially collapse, possibly releasing a small amount of gas there. According to some reports, it appeared that the water was boiling near the shore of a prominent hotel.
The lake appears stable for now. It does contain a lot of gas, but in the area with the highest concentration, that amount would have to double again to reach the saturation point. However, a strong earthquake or volcanic eruption could destroy the layered structure of the lake or increase gas concentrations, leading to outgassing. Some researchers also fear that, in addition, human activity could bring about a disaster.
Methane is also used
Methane is already being pumped up from the depths of the lake and burned to generate much-needed electricity. Most people agree that this is a wise use of local natural resources while making the lake safer by removing the gas. A lot of money is at stake: researchers estimate the methane at the bottom of Lake Kivu could yield 36 billion euros over a 50-year period.
However, they disagree on the best method to extract the gas and whether such activities could disturb the lake in a way that increases, rather than decreases, the risk. Despite the prevailing debate, methane production is expanding. There are already plans to increase power generation fivefold in the coming years or decades.
"Many scientists disagree," says biochemist Eric Ruhanamirindi Mudakikwa. He was formerly head of Rwanda’s Lake Kivu Monitoring Program (LKMP), which has been taken over by the Rwanda Environment Management Authority’s Environment Analytics and Lake Kivu Monitoring Division. "What we are doing at the lake is still very new. We have no idea how it might react."
Lake Kivu is the largest of a small group of lakes worldwide where the possibility of a limnic eruption is thought to exist. Two much smaller lakes lie a few thousand kilometers to the west in Cameroon, and another, Lake Albano, is in Italy.
These lakes are all located above tectonically active regions, where volcanic gases such as CO2 rise from the earth’s interior. They are deep and their water layers are not mixed by seasonal temperature fluctuations from top to bottom. Instead, the dissolved gases collect in the denser lower layers and are held in place by the water pressure from above, as if by a cork. When so much gas accumulates that it forms bubbles, these lakes can literally blow up like a shaken champagne bottle. An external event can also "pop the cork". A falling water table due to drought could reduce pressure on the deeper gas-filled water layers. A landslide, earthquake or lava erupting at the bottom of the lake could cause the layers to shift or absorb enough heat to cause gas bubbles to rise.
The terrible potential of such lakes was demonstrated in August 1986, when Lake Nyos in Cameroon erupted with such a thud that some residents suspected a nuclear weapons test. The surrounding regions were flooded with a cubic kilometer of carbon dioxide. Since it is heavier than air, over 1700 people and 3500 farm animals suffocated from it.
After this eruption, a project was launched to prevent such events at Lake Nyos in the future: In 2001, physicist and engineer Michel Halbwachs, who was then working at the Universite Savoie in Chambery, France, and his team pumped gas-filled water from the depths of the lake. To do this, they lowered a long tube from a floating platform into the deep water layers. This created a self-sustaining fountain through which the gas could escape in a small, controlled version of a limnic eruption. In 2011, the team added two more pipes, and by 2019, Halbwachs and his colleagues considered Lake Nyos "all but completely free of risky levels of dissolved carbon dioxide".
After the initial installation, Halbwachs attended to the smaller counterpart to Lake Nyos, Lake Manoun, where a much smaller eruption had occurred in 1984. After installing vent pipes, this lake has been considered gas-free since 2009.
Halbwachs’ company Limnological Engineering has now secured a contract worth almost 4.5 million euros. It involves degassing the Gulf of Kabuno, a small northern offshoot of Lake Kivu, which has a high concentration of CO2 even at shallow depths. The company has already been working on a pilot project since 2017.
Giant among gas lakes
However, Lake Kivu, which is disproportionately larger, presents a very different problem. Lake Kivu is geologically older than Lake Nyos, and the surrounding soil contains more organic matter. As a result, unlike Lake Nyos, it harbors significant amounts of methane, explains biogeochemist George Kling, who studies limnic eruptions at the University of Michigan in Ann Arbor, U.S. Microorganisms that break down organic matter produce methane, and methane or hydrogen of volcanic origin can enter the lake directly from the lower rock layers. The solubility of methane is far below that of carbon dioxide, so rising bubbles occur much more quickly. "The methane is the real problem. That’s different from Lake Nyos," says lake physicist Alfred John Wuest of the Swiss Federal Institute of Aquatic Science and Technology, or Eawag for short, in Kastanienbaum.
The lake could hold significant amounts of CO2 without safety risk if the methane didn’t contribute to the gas pressure. Scientists believe carbon dioxide will no longer be a problem if methane is diverted as an energy source.
Despite the potential danger of Lake Kivu, there is considerable disagreement about the basics, such as where the gas is coming from, whether it is becoming more, and even whether Lake Kivu has erupted before. Robert Hecky, retired lake ecologist from the U.S. University of Minnesota Duluth, has studied Lake Kivu. He says that while the sediments show nine brown layers suggesting mixing over the past 2,000 years, he has found no evidence of an event as massive as a limnetic eruption for the past 12,000 years. Another interpretation of the facts is that an eruption occurred at least 4000 years ago.
Some facts are clear. The upper water layers of the lake are relatively cool and rich in fish. About 260 meters down, a drastic change occurs. The water there is much warmer and more saline due to hydrothermal vents. Here are the deep "resource-rich water layers" that are rich in dissolved gases.
In a 2005 study, Eawag environmental scientist Martin Schmid and his colleagues, including Halbwachs, compared gas levels in this deep layer with measurements made in 1975. They concluded that methane concentrations had increased by 15 percent. If the increase remained the same, the deeper layers would be saturated around 2090 and would trigger an eruption. However, another 2020 publication, for which Schmid was a co-author, said that gas concentrations had not increased at all.
This reassured many researchers, but the results remain contradictory. For example, different gas measurement methods were used to collect the data. Whether or not gas concentrations have risen, however, does not change the fact that the future is uncertain; they could rise dramatically at any time without warning. "The subsurface branching in the volcanic system of the rift zone surrounding Lake Kivu is largely unknown," Kling says. "It is quite possible that the gas supply will increase dramatically due to stronger subsurface volcanic or geological activity."
The same volcanic eruptions and earthquakes could theoretically also trigger an eruption. "With a gas-rich lake near a volcano, there is a huge potential for triggers," Hecky said. The only question is how strong they would have to be. "The lake is exceptionally stable. Circulation would require a huge amount of energy," he says. Volcanologist Dario Tedesco of Italy’s Università degli Studi della Campania Luigi Vanvitelli, who works in Rwanda, says his data on the 2021 volcanic eruption show that no gases escaped from fissures around Goma or from the lake. According to him, either there was no underground magma or the flows were so small or so deep that they had no effect.
However, most of the dozen or so scientists contacted by Nature continue to express concern about the lake’s methane content, given the geological activity in the region. If 90 percent of the methane were discharged over about 50 years, Morkel argues, this could reduce the likelihood of a limnic eruption by 90 percent in the first ten years. "In the best-case scenario, this will never happen," says Morkel.
The methane is tapped
Small amounts of methane have already been pumped from Lake Kivu for decades to generate energy. These activities were significantly increased when London-based ContourGlobal’s KivuWatt power plant began operating in 2016. The 170-million-euro project currently provides 26 megawatts of electricity. Contractually agreed to increase to 100 megawatts. This will noticeably help cover the base load in Rwanda’s power grid, which is designed for 200 megawatts.
So far, KivuWatt has made relatively small reductions in the lake’s supply: at the current rate of extraction, the company will remove less than five percent of the methane in 25 years. "This rate is certainly not sufficient to significantly reduce the risk of a limnic eruption," says limnologist Francois Darchambeau, who works at KivuWatt. "So we have to increase the capacity." However, expansion plans will have to wait until demand for power matches supply, company says. At KivuWatt, they are also considering pumping carbon dioxide from the lake and selling it as a commercial product.
Meanwhile, Rwandan company Shema Power Lake Kivu has bought the small KP-1 pilot plant, which began pumping methane from the lake in 2006. They are currently building a plant there that will supply 56 megawatts. The company’s website says the construction phase is scheduled to end in early 2022. However, Shema Power’s project manager, Tony de la Motte, would not answer questions from Nature about the timing of the plant or its operation.
All of these projects generally follow the principle of bringing water up from the depths so that the methane can be released from the water and it can be cleaned and pumped to a power plant. The degassed water will then be returned to the lake. Questions revolve around how best to accomplish this, and plans vary by company and proposal.
The degassed water still contains high concentrations of nutrients and toxic hydrogen sulfide. If it is returned to the lake too close to the surface, some researchers say it could lead to fish kills and harmful algae blooms. It is also salty and rich in CO2, so it has a relatively high density. So if discharged at too shallow a depth, the methane-free water would sink, potentially disrupting the key density gradient at 260 meters that keeps the gas-rich water secure in the resource zone below. "That wouldn’t necessarily lead to an eruption," Morkel says, "but the risk of an eruption would increase."
Shifting this gradient upward could also be problematic because it would lower the pressure on the gas-rich water. And diluting the resource layer with degassed water could lower the gas concentration to the point where commercial recovery would no longer be possible. In that case, a large amount of the dangerous gas would remain in the lake, with no meaningful way to remove it. The only remaining option would be to discharge it to the surface, but this would release strong greenhouse gases and contaminate the surface water.
In 2009, a group of international scientists, including Morkel, Wuest and Schmid, published "management rules" (BV) summarizing proven methods for extracting methane from the lake. The majority of experts favor the so-called method of preserving the density zone. The density of the degassed water is controlled by the CO2 content so that it can be carefully returned to the lake without triggering mixing. This is a technical challenge, but would preserve the current structure of the lake.
KivuWatt has opted for a different strategy: Degassed water is returned just above the main gradient. It’s easier to achieve and shouldn’t dilute the resource zone, but will likely change the structure of the lake.
Sarchambeau explains that KivuWatt monitors surface water on a daily basis and conducts weekly analyses to provide reliable data on the lake’s stability. He says that after five years of operation, as expected, there is a weakening of stability, but only a slight. "If we continue the current gas extraction for another 50 years," says Sarchambeau, "it will reduce the stability of the lake by one percent."This value is well below the BV recommendations, which state that stability should not be reduced by more than 25 percent, measured against the energy that would be required to fully mix the lake.
However, there are voices that consider KivuWatt’s approach problematic. "This way leads to disaster," says engineer Finn Hirslund of consulting firm COWI in Lyngby, Denmark. He is one of the authors of the BV and has published peer-reviewed papers on Lake Kivu. Hirslund argues this project will "destroy the main gradient" and is concerned that continuation and expansion of gas extraction from the lake using similar methods may have long-term consequences that won’t become apparent for decades.
Morkel is also critical of KivuWatt’s approach. He objects that the company’s degassed water contains too much carbon dioxide and has too high a density. This, he thinks, will push it through the main gradient. Morkel is in favor of taking water and returning it to depths other than KivuWatt. He believes this will better preserve the lake’s stratification when gas is extracted for energy production. That is why it is trying to attract funding for this approach.
However, there are also scientists who are not concerned. "In terms of safety, I am absolutely confident," says Wuest, who is also part of KivuWatt’s independent panel of experts. "I see the whole thing as absolutely positive," emphasizes physicist Bertram Boehrer of the Helmholtz Center for Environmental Research in Magdeburg, who has worked on the lake. "In the case of unexpected developments, there is still plenty of time to act."
What the future holds
The only way out of the debate over possible consequences for the lake is probably close monitoring of whether and how the density layers change. The Lake Kivu Monitoring Division (LKMD) monitors the depths of the lake and audits gas-producing companies. Mudakikwa says weekly analysis shows the lake remains stable so far. "The main gradient is not changing," he says. "If the lake becomes unstable, we’ll be the first to worry about it."
KivuWatt clarifies that the guidelines of the LKMD are binding and that it adheres to them. In addition, the company’s independent council of experts, which includes Hecky and Wuest, has access to KivuWatt’s data and is reviewing its annual report to the Rwandan government. "We’re very open to the science," Sarchambeau says, although some information, such as the design of the customized sensors for gas concentration, remains classified. "Everyone wants KivuWatt’s data," says company director Priysham Nundah. "I can’t give anything to competitors, but we give [to LKMD] what we are contractually obligated to do and according to agreement."
Some researchers contacted by Nature complained that they have difficulty getting such data. "In our [BV] guidelines, we clearly specified that this data must be public," Wuest says. "To my knowledge, the Rwandan government has never acted on it." Mudakikwa states that data from gas-producing companies is confidential. The lake’s analysis data can be obtained by researchers writing to REMA’s general manager, stating what they need and why.
The combination of Lake Kivu’s economic value, its explosive potential and the many divergent opinions on how best to deal with it is heating tempers among scientists working there. "I’m just obsessed with understanding what’s going on in this lake," Hirslund says. "Working at Lake Kivu is a passion," agrees Augusta Umutoni, head of LKMP.
Pumping gas from lake should make it safer, says Mudakikwa. But there are things, such as volcanic eruptions, that neither scientists, companies nor regulators can avert or prevent. "When in doubt, we can’t beat Mother Nature"."