A feasibility study of storing hydrogen in depleted gas reservoirs

If a full scale hydrogen economy is to become a reality, we are going to be producing a lot of hard-to-store fuel. One of the advantages of hydrogen as a resource is that it can be produced during times of surplus renewable energy and stored for a rainy day. However, for this cycle to prove reliable on a national scale, we will need to find ways to store excess hydrogen – possibly for many months.

A. Amid, D. Mignard and M. Wilkinson from the University of Edinburgh have just published a study in the International Journal of Hydrogen Energy, looking into the possibility of using depleted natural gas reservoirs for pressurised hydrogen storage. As of January 2013, a total of 688 natural gas storage facilities were operated worldwide with a combined working gas capacity of 377 billion m3, so if it proved technically feasible there would certainly be scope to take advantage of these spaces.

The researchers’ analysis foresaw three technical problems. First, that the remnant methane in reservoirs would contaminate the hydrogen. Second, that micro-organisms could feed on the hydrogen while it is in storage; and third, that the hydrogen will leak from its prison.

The team’s modelling was based on a partially depleted natural gas reservoir in the Southern North Sea, off the coast of Yorkshire. Their findings were,

  • Contamination: That initial injection cycles could see some methane contamination, but this was ‘not a serious concern’, and the contaminants would be cleansed over multiple cycles. Hydrogen sulphide could prove an issue, and reservoirs should be chosen to minimise the amount of sulphate reducing bacteria.
  • Consumption: There would be some loss of hydrogen as it is converted to methane and biomass, but the worst case scenario would put this at no more than 3.7% of the hydrogen.
  • Leakage: Hydrogen is more diffuse than methane, and so we could expect some leakage, but it is unlikely to be more than 0.035% of the stored hydrogen after 12 months.

In conclusion, the assessment found that there is no insurmountable technical barrier to this plan, given current technology. However it would be a major undertaking, with an average power in the order of 4–5 GW required during a six month injection cycle to fill the reservoir to capacity, provided that cushion gas is already present.

To read the study in full, click through to the International Journal of Hydrogen Energy, Volume 41, Issue 12, 6 April 2016, Pages 5549–5558.