From Bureau of Economic Geology, The
University of Texas at Austin (www.beg.utexas.edu).
For more information, please contact the author.
Bureau Seminar, October 11, 2013
Link to streaming video: available 10.25.2013 at 8:55am
Bureau of Economic Geology
In 1986, on a cool winter night in a remote area of the Cameroon a highly unusual event occurred. Now known as the Lake Nyos disaster, a large evulsion of gas occurred from an unusual deep permanently stratified lake killed 1,700 people (over half the population of the villages surrounding the lake). Geologists on the international science teams that investigated the incident concluded that the gas cloud that caused the disaster was essentially pure CO2. The Lake Nyos incident has become widely seen as reflecting the dangers associated with CO2 sequestration. The consensus of literature on the risks of CO2 sequestration is that the Lake Nyos disaster was an example of mass death from a cloud of CO2. Curiously the reports of the medical investigations of the disaster did not come to this conclusion. The US medical team wrote that "doubt must remain that all the findings can be attributed to [CO2] alone". Evaluation of the symptoms documented in medical studies lead several medical experts to suggest that the Lake Nyos victims died from carbon monoxide poisoning. Although the medical teams reported no definitive evidence of cause of death, they did note that many victims had prominent skin bullae (blister like features) or has succumbed to comas, and that neither of these prominent symptoms can be related to CO2.
Although the role of CO in the Lake Nyos incident has been dismissed by geologists studying the Lake Nyos event, their reasoning is flawed. A plausible model for the causative agents for the Nyos disaster was a combination of CO2 (quite possibly at non-lethal concentrations), reduced oxygen, and the critical deadly agent, CO. Supporting evidence for this suggestion comes from an unusual sources, the investigation of deaths associated with suicides involving vehicles fitted with catalytic convertors. Carbon monoxide poisoning utilizing automobile exhaust in an enclosed space for decades was a classic suicide strategy. Since the development of catalytic convertors the carbon monoxide has been in large part converted to carbon dioxide and a nominally non-lethal level of carbon monoxide. Unfortunately suicidal acts continued unabated due to the powerful effects of mixed CO2 and CO. Reconstruction of the composition of lethal gas mixtures in these suicides are consistent with experimental data on the mortality of rats in mixed CO2/CO atmospheres and show that levels of CO2 and CO that on their own would be non-lethal can be deadly when combined. It has been suggested that CO2 levels during the Lake Nyos incident were on the order of 10%. Similar levels have been measured in reconstruction of vehicular suicides. In the absence of carbon monoxide neither the gas in exhausts from catalytic convertor nor the gas at Lake Nyos, would be lethal. Perhaps the real lesson of the Lake Nyos incident for carbon capture and storage is to draw attention to the inherent danger from mixtures of toxic gases. For example mixtures of H2S with CO2 will increase the toxic effects of H2S. The risk profile for a sequestration project changed a decision is made to inject mixed streams of gas such as CO2 + H2S rather than essentially pure CO2.
A key aspect of the Lake Nyos incident was the exceptionally large quantity of CO2 that was abruptly released in the incident. The exact quantity is uncertain however there is a consensus view that it amounted to between 1.0 and 1.6 million metric tons of CO2. In the context of CO2 sequestration this minimum amount of CO2 corresponds with approximately four months emissions from a 275 MW FutureGen type IGCC or over a month of CO2 captured from a large, but not atypical, 1000 MW coal fired power plant. This volume of CO2 is also equivalent to weeks or months of gas that would be transported in the largest pipelines contemplated for a future sequestration project. Given the isolation valves that would be activated the CO2 released at Lake Nyos was 2 to 3 orders of magnitude larger than the largest release plausible from CO2 pipelines. A more useful analogy between natural CO2 seepage and the worst-case scenario for leakage from CO2 sequestered in deep brine reservoirs is provided by the volcanic/geothermal terrains in Italy, the Azores, and Africa. In this presentation the available information and analysis of the presumed CO2 releases at these sites and the related deaths are reviewed and the implications for evaluating the risk posed by either slow or rapid leakage from CO2 sequestration are considered.