Kivu lakeMethane
Gas Extraction
Putting methane
to work
Natural hazards


Our project proposes the immersion of the entire system (separator and washing column) at a shallow depth.
the fact of working entirely under water has numerous advantages :

  • the volume of gas freed by the separator and needing to be washed decreases
  • the pressure of the gas rises, so the transport by gas pipeline to the shore is made easier
  • washing and transport operations can be achieved with a minimum of energy from the exterior
  • the richness in methane of the gas mixture is directly optimised at the separator
  • the efficacity of putting the CO2 into solution in the washing column is increased.

But this option has also some negative aspects :

  • the driving force of the autosiphon is diminished
  • part of the methane stays in solution in the discharged water and is thus unexploited.

These remarks are of major importance since they illustrate the compromise to be made between total gas flow and loss of methane.



Degassing below the surface of the lake

Methane being 25 times less soluble in water than carbon dioxide, its ex-solution should occur before that of CO2. In fact, if the two gases behave independently of each other, the first bubbles of methane should appear at a depth of about 120m whereas those of carbon dioxide should form only at about a depth of 15 metres.

Apparently, by positioning the separator at this depth, one should obtain pure methane, the carbon dioxide staying dissolved in the water. This theoretical shortcut is obviously wrong, since it does not take into account the establishment of a balance between chemical types : when a bubble forms, all of the dissolved gas spreads in the bubble until a balance of partial pressures with the water is obtained. The moment a bubble of methane appears, the instability thus created transforms the other dissolved elements into gas until the balance of partial pressures is achieved both inside the bubble and in the liquid.

This important observation is nevertheless difficult to quantify precisely : the calculation of the behaviour of diphasic fluid moving in a degassing column is in fact difficult to make, given the great number of parameters which come into play. It is, however, recognised that in proceeding with the separation of gas and liquid under pressure, that is to say submerging the separator at a certain depth, the richness of the methane in the mixture will be clearly greater than that obtained with a separator on the surface. The conclusions of the majority of projects during the past thirty years point in that direction.

On the other hand, from the fact that a large percentage of the carbon dioxide stays in solution when you operate at depth, it follows that the volume of gas needing treatment is smaller. Successive washing operations can thus be recalculated in order to obtain a gas even richer in methane.

Unfortunately there are two negative points to consider.

  • Firstly, as the volume of gas resulting from the ex-solution is less, the driving force of the auto-pumping is equally diminished.
  • Secondly, the separation of liquid and gas at depth increases the quantity of methane remaining in solution in the water, giving rise to a loss in the total quantity of potentially recuperable methane.


Methane loss, gas flow and criteria for chosing energy

The diminution of gas flow by volume according to the increase in the depth of the separator is of major importance in the overall concept of the system. In fact the efficacity of the washing and its cost in energy depends directly on the volume of gas to be treated. By decreasing this quantity of gas, the quantity of washing water needed is decreased similarly, this leading to a gain in the dimensions of the washing column.

To minimize the quantities of gas to be washed it seems that a minimum depth of 10 m is necessary. But the deeper the separation of gas and liquid takes place, the more methane, which remains in solution, is lost. The depth will thus largely be limited by this phenomenon of the non-ex-solution of methane. For the loss to be minimised, it seems that a maximum of 30 m would represent the limit of exploitation.

It is interesting to evaluate the variations in energy production of the system according to the depth of installation of the separator. This criteria clearly shows the need not to try to work too deeply. In fact a calculation based on gas and liquid flow allows an estimation of 13% potential energy loss for an increase in depth of 5 m.

If the extraction and washing of the gas under the lake surface seems the best solution, it would nevertheless be wise not to exagerate the benefits of such a method. Losses in extractible methane are such that energy production is seriously penalised.