Detailed Modeling of Injection and Production Induced Rock Displacements

Reservoir injection and production leads to stress changes in the reservoir and the surrounding rock. For compacting reservoirs, these effects are particularly prominent. The associated rock displacements make infill drilling and optimization of performance and integrity challenging. Constant monitoring and prediction of the displacements of the rock is key to identify risks early. Once such risks are identified the production strategy can be adjusted to mitigate such risks and overall increase the recovery of hydrocarbons.

This paper describes how displacement information obtained from time-lapse seismic data and estimates of rock material properties from well data or databases can be used in detailed modeling of geomechanical effects. The output of the geomechanical simulation is the stress change and the vertical and horizontal rock displacements. The displacements at the boundary of the model are imposed based on displacement estimates from the time-lapse seismic data, and together with the computed displacements in the interior of the model a complete description of the stress exchange can be obtained.

An advantage of the method is, that no input from reservoir flow simulations coupled with potentially complex yield functions is required, and a detailed description truly consistent with the time-lapse seismic data can be obtained. As a starting point the displacement at top reservoir is estimated from the observed time-lapse time shift using an R-factor model for the time shift to depth shift conversion. Together with the locally dependent material parameters the boundary displacements are used as input to the geomechanical model. The displacements can then be simulated everywhere in the interior of the model. In a next step these displacements are compared to the displacements observed in the time-lapse seismic data. According to the mismatch, the accuracy of the current geomechanical model can be evaluated. To improve the match between simulated displacements and the displacements estimated from the time-lapse seismic data, the local material properties in the model and/or the depth shift estimates can be adjusted. The whole process improves the confidence in the proposed model parameters.

Once the parameters are adjusted and the simulated displacements agree well with the seismic observations the resulting local stress conditions can be trusted. Knowledge of the local stress conditions is important in infill drilling and optimizing well performance and integrity.