FRACTURE PROPAGATION ALONG A FAULT – STRESS HYSTERESIS EFFECT

According to the 2018 Intergovernmental Panel on Climate Change (IPCC) report, to keep temperature rise under 1.5 °C compared to pre-industrial level, no credible emission scenario can be considered without Carbon Dioxide Removal (CDR), CCS and afforestation being the only two solutions available. CCS on the scale of hundreds of gigatons of CO₂ will be needed to reach these goals, with tens of thousands of suitable reservoirs to be found and exploited to store CO₂. Two main types of geological formations are considered for storage: deep saline aquifers with thick sealing caprock and depleted oil or gas fields, already proven natural reservoirs. Several demonstration facilities have already proved feasibility of CO₂ storage. However, storing carbon dioxide underground may not be as safe as producing oil and gas in the long term, as it is mainly about increasing pressure in a reservoir rather than decreasing it. Therefore, when purely considering pore pressure change scenarios, depleted oil and gas reservoirs are the preferred storage choice, since one can limit the pressure increase to reach back the original pressure in place when the reservoir was discovered. This pressure had been maintained over geological time without any leakage of the pore fluid and should therefore be ideal to store CO₂. Another argument in favour of depleted reservoirs is the presence of existing infrastructure in place, which could probably be used with minor modifications, compared to virgin aquifer targets. On the other hand, the very presence of wells penetrating the underground dramatically increases the risks of leakage, principally along these wells, where many of them could be old and in poor condition.
These considerations about depleted oil and gas fields do not however take into account non-elastic associated effects in the proposed pore pressure rebound scenario. These effects are translated into stress hysteresis and stress concentration, which could exceed the strength of the formations in and around the proposed reservoir. Many oil and gas reservoirs are compartmentalised in pressure cells bound by sealing caprock and lateral faults. Stress changes accompanying depletion when the reservoir was in production mode may not be evenly distributed but concentrate on the bounding faults. Hysteresis related to plastic yielding (non-recoverable deformation) may also aggravate the solicitation of the faults upon CO₂ injection into the reservoir and lead to reactivation of the faults and hence leakage out of the intended reservoir.
In this paper, we implement a simple scenario looking at a fault on the side of a sandstone reservoir and simulate a depletion and recharge of the reservoir while monitoring stress evolution on the fault. A process zone around the fault core is explicitly implemented in the simulations, by assigning the zone weaker mechanical properties as compared to intact rock and placing in the zone pre-existing fractures.