Seismic correlated Mallik 3D gas hydrate distribution: Effect of geomechanics in non-homogeneous hydrate dissociation by depressurization

摘要

The delineation of the Mallik gas hydrate field has utilized extensive well logging and substantial 3D seismic testing and interpretation. This study explores the use of seismic data to quantify the areally heterogeneous gas hydrate distribution. The available Mallik 3D seismic data was compiled and compared/contrasted with available well log data from two adjacent wells. Based on the seismic information, two areally variable (i.e. non-homogeneous) scenarios for gas hydrate distributions are considered: Scenario I having the same initial total hydrate amount as our earlier model areally uniform (homogeneous) distribution, and Scenario II with significantly less overall total hydrate, but honouring the same relative distribution. The scenarios of variable gas hydrate distributions are used in dynamic simulations of the lower Mallik zone. Simulations of each were conducted with and without the role of geomechanics. In Scenario I, we observed multiple gas production peaks (which quite similar to 6 days production behaviour) with higher localized pressure pulses occurred due to strong gas hydrate heterogeneity. In Scenario II, this drastic change in gas production rate was not observed (due to faster pressure evolution in the reservoir). In both Scenarios, the overall reservoir gas production peak is delayed compared to the homogeneous case. This is further delayed by the role of geomechanics. More interestingly, all simulation cases show a very similar overall production trend. This is probably a unique for the Mallik gas hydrate production using single vertical well, including a gas production peak but terminating in a stabilized period of lower but significant gas production. With geomechanics, gas production in general and the gas production peak is shifted and delayed. The geomechanics effect is not purely compaction drive (as in conventional reservoirs, gas production increases with geomechanics). The simulations utilized two set of geomechanical parameters obtained from logs (dynamic parameters) and rocks testing (static parameters). Geomechanical responses based on dynamic parameters were essentially equivalent to simulations ignoring geomechanical effects. The geomechanics simulations indicate an essentially elastic reservoir response (i.e. no plastic failure) assuming a cased vertical well. The Mallik upper zone A and middle zone B are closer to the permafrost and nearer to plasticity limits should be explored.