KEM-01 Safe operational bandwidth for gas storage in faulted reservoirs

The objective of the proposed study is to re-assess the basic geomechanical causes of seismicity, specific risk factors, and safe operational bandwidths for reservoir storage conditions, that have not been addressed in previous studies. In particular, the study should focus on those factors that fall beyond the well-known causes for fault slip and seismicity at peak depletion /minimum reservoir pressure (and consequently maximum effective stress conditions). For instance, one underground gas storage (UGS) shows micro-seismicity during the injection phase near the maximum allowed reservoir storage pressure, on a fault that was not active during depletion.

Another Dutch UGS showed an observed seismic event after re-injection of the cushion gas and initial work volume at a near-virgin reservoir pressure. In both cases, these were conditions under which fault criticality was least expected. Seismicity during injection of the cushion gas was also observed for the UGS Castor in Spain, with events as large as M=4.0. Hence, from a traditional fault stability point of view, we may be missing important causal mechanisms. Is the micro-seismicity observed in some of the Dutch UGS sites, at stress conditions and locations where it was not expected, a warning that higher magnitude events can occur at gas pressures where they are not expected?

RESEARCH REQUEST FULL TEXT

Empty gas fields can be used as storage by injecting natural gas into them. This natural gas is then extracted when demand is high, for example during the winter. Earthquakes are caused by movements along faults in the deep subsurface. These faults move because stresses are built up on the fault surfaces when the gas pressure in the rock decreases, as is the case when gas is extracted. The research showed that the seismic movements along faults in gas storage reservoirs can be modelled with the proper simulation tools using local parameters.

The computations executed show that in the case of gas injection, pressure rises and the slippage-inducing stress on faults initially decreases, but can increase as the gas pressure in the rock increases towards the initial reservoir pressure. Gas storage reservoirs that have been seismically active during initial gas production stage have a higher chance of inducing earthquakes along faults during filling than the gas storage reservoirs that did not show seismic activity in that stage.
The research also shows that the fault surface that is critically stressed during injection is smaller than the fault surface that is critically stressed during gas production. In other words, the risk of earthquakes in underground storage facilities is mainly determined by extracting natural gas, and to a lesser extent by injection.

Based on these new insights, guidelines have been drafted how to compute a safe operational bandwidth for specific gas storage reservoirs.

The project consisted of two phases and resulted in total six deliverables:

• KEM01 WP1 literature inventory

• KEM01 WP2 mechanism identification

• KEM01 WP3 phase 1 geomechanical model computations

• KEM01 WP4 phase 1 analysis intermediate results

• KEM01 WP5-6 phase 2 geomechanical model computations

• KEM01 WP7 Final report phases 1 and 2

The project results have been reviewed by KEM scientific experts. The meaning for practice in the Netherlands and consequences of applying the proposed guidelines are underlined by the KEM scientific expert panel in their final evaluation.
On June 24, 2021 Prof. Pietro Teatini presented the results of the KEM-01 project in the second KEM-DeepNL colloquium.