Kivu lakeMethane
Gas Extraction
Putting methane
to work
Natural hazards


The recent volcano-tectonic crisis at the Nyiragongo volcano which destroyed part of the town of Goma in January 2002 has highlighted a new risk factor due to the presence of dissolved gas in the depths of Lake Kivu (cf. Publication in Appendix).

Before this disaster the risks of gas emanations - and a fortiori of a cataclysmic explosion - caused by a natural destabilisation of the lake would have been thought infinitely small. This is no longer the case : the hypothesis of a magmatic intrusion - even a volcanic eruption in the depth itself of the lake - can no longer be totally excluded.

Pigs island , a old volcano located in the northern part of lake Kivu
(photo H. Tazieff).

The eruption of Nyiragongo flowed towards the lake through eruptive vents along a tectonic faultline up to 15 kms from the crater of the volcano (5 km from the lake shore). If this advance had continued 6 or 7 km further south, magmatic effusions would have occurred at a depth of 300 metres in the lake, inevitably provoking an upswell of gas-filled water. A cloud made up of a mixture of carbon dioxide and methane would appear whose impact is difficult to determine but which could turn out to be catastrophic : the entire populations of the towns of Goma (400,000 inhabitants) and Gisenyi could be directly threatened.

This 'limited gas emanation' scenario is only a 'minor' hypothesis : a major volcanic eruption could take place one day directly on the lake bed.

The volcanic eruption of January 2002 is unique in the history of vulcanology because it was caused by a phenomenon of tectonic fracture between plates. It is quite likely that the tectonic faultline which continues under the lake is presently weakened. Moreover, seismic activity seems at the moment to be concentrated to the south of Goma, that is to say directly under the lake. A volcanic eruption at depth, adding to the thermal element of an outflow of lava an element of mechanical energy linked to the pressure of magmatic gases, could have the cataclysmic result of pulling the deep waters of the lake up towards the surface with an 'avalanche' mechanism, as at Nyos.

Note that the quantity of dissolved gas in Lake Kivu is a thousand times greater than in Lake Nyos, which caused 1800 victims in 1986 in a sparcely populated region of Cameroon. If a real 'overturning' of the lake occurred it would be the whole area of Kivu which would be threatened : Goma, Gisenyi, Bukavu, Cyangugu, even Bujumbura. The gas cloud could cover the areas around the lake and flow even as far as Lake Tanganyika, 700 m below Bukavu. A poulation of around two million is living under permanent threat.

We have purchased raw data files of a bathymetric field trip of 1998 from the OSEA Company. This data has been processed so as to extract 3D images of the lake floor. These images show the existence of a group of ancient volcanic forms situated on the line of tectonic faults between Nyiragongo and Nyamulagira. These ancient volcanoes are at a depth of between 300 m and 400 m, that is to say right in the middle of those layers of the lake which are full of gas. Any new eruption would inevitably trigger off, by chain reaction, the liberation of an enormous quantity of dissolved gas, with consequences that would probably be catastrophic for the entire Kivu region. A population of several millions (Goma, Bukavu, even Bujumbura) are living under this threat.

The type of event which could cause a disturbance large enough to trigger off, in the lake, an 'avalanche' type gas explosion has to contain a considerable amount of energy. This could be the case with a volcanic eruption at a medium depth. One is not trying to spread panic but, to repeat, the possibility of such a disaster is not an imaginary scenario. Even if the risk is probably quite slight, one has to take into account the consequences, which would be nothing short of cataclysmic.

We there think that it is important to seriously study this new type of risk and that it would be wise to set up a basic alarm system at the lake which would indicate the appearance of any movement in the liquid mass of the lake waters.