Pore pressure reduction due to hydrocarbon extraction from gas reservoirs causes compaction of the reservoir rock accompanied by a redistribution of stresses in the surrounding rock mass, leading to sub-surface deformations and surface subsidence. There is evidence to suggest that the induced sub-surface movements in some of the Netherlands’ gas fields (the motivation behind the study) are time-varying. The sandstone reservoirs in question are overlain by relatively thick layers of rocksalt, which flows viscoelastically when shear stress are applied to it. For this reason, rocksalt flow has been proposed as a mechanism to explain the observed time-dependence. The work reported here aims to understand the mechanisms by which viscoelastic rocksalt flow can cause surface displacements and to assess whether the magnitude of the resulting displacements can be significant over timescales similar to those observed in the field. Finite Element analyses of idealized reservoir geometries and simplified geological models will be employed for this purpose, with the aim of avoiding complexities that can mask the underlying processes and phenomena. Results here indicate that salt-driven flow has the potential of inducing significant timedependent subsidence, in excess to the ones calculated through solely elastic analyses of the same problem. They therefore indicate that in cases where accurate subsidence predictions are essential, salt has to be modeled accurately.
Extraction of natural gas from a reservoir rock formation leads to pore-fluid pressure decrease and so an effective stress increase in the reservoir rock. This in turn leads to reservoir compaction which is registered at the surface as subsidence. The link between production-induced pore pressure changes at the reservoir level and surface subsidence is well-established, e.g. [1, 2, 3, 4].
In many cases the ground is modelled as elastic and so an immediate subsidence response to the pore pressure reduction is expected at the surface (after [5, 6]). This thinking therefore assumes that subsidence stops immediately after the stop of production from a reservoir. However there is growing evidence to suggest that for some reservoirs there is a delay in the subsidence response, i.e. a time-dependence of ground deformations, see [7, 8].