Monitoring and modeling land subsidence due to hydrocarbon production integrating geodesy and geophysics

As part of the DeepNL program, we study induced seismicity and subsidence in Groningen, the Netherlands, resulting from gas production and injection. Combined efforts in the fields of geodesy and geophysics provide a better estimation, modeling, and understanding of the deep and shallow subsurface processes that cause subsidence. This will lead to improved risk assessment and enable adaptive hydrocarbon production.

Interferometric Synthetic Aperture-Radar (InSAR) satellite observations are used to estimate cumulative surface displacements over the area of interest, using a functional model based on geophysical prediction. Contextual information is attributed to the geodetic observations to help determine their physical origin and to link them with the subsurface parameters. Additionally, shallow and deep subsurface processes are modelled to understand and possibly forecast ongoing and future subsidence. The shallow subsurface model quantifies individual processes such as oxidation, compaction and groundwater extraction. Numerical modelling and field measurements are necessary to upscale geological and hydrological data and to study the time variation of the shallow (de)compaction processes.

Finite Element Models are used to characterize the processes that play a role in the reservoir and the overburden, such as the fault system and pressure gradients leading to differential compaction and subsequently to differential surface deformation.

Data assimilation can help to disentangle drivers of subsidence by comparing estimates of subsurface models and geodetic data given the probabilistic distribution of model and data uncertainty. We quantify the effectiveness of ensemble-based methods for systems of increasing complexity.

Eventually, the integration of geophysical modelling and geodetic data processing by means of data assimilation done on a scalable platform, will highlight the spatio-temporal interdependence of the driving processes of subsidence over a longer observation period.