ABSTRACT: We conduct a quantitative microstructural investigation of Slochteren sandstones recovered from the Groningen gas reservoir in the NE Netherlands, aiming to identify the micromechanical processes controlling inelastic strain accommodation. Core samples were provided by the Nederlandse Aardolie Maatschappij (NAM), from the Stedum (SDM)-1 and Zeerijp (ZRP)-3a wells, respectively drilled before and after gas depletion in 1965 and 2015. Sectioned samples were imaged using a scanning electron microscope for constructing section-scale, backscattered electron micrograph mosaics, which were interpreted by manual grain tracing and image processing to obtain digital phase and crack maps. Digitization of the sandstone microstructure in this way yields data on grain mineralogy, shape, (contact) size and orientation, as well as on coordination number and crack density and orientation. We investigated ways to improve the microstructural digitization method, including preparation of thin sections using Au vapor deposition, injection of copper sulphate (CuSO4) at elevated pressure, and etching using hydrogen fluoride (HF). Finally, an image analysis method was developed to enable semi-automatic phase/ crack mapping, which we first applied to a layered sample from the ZRP-3a core. Comparison of phase/ crack maps of layered sandstone samples may provide the key to help identify the processes controlling pore pressure depletion-induced compaction.
Hydrocarbon production causes changes in the effective stress state of the reservoir rock, frequently leading to compaction (Geertsma, 1957; Segall and Fitzgerald, 1998; Hettema et al., 2000) and surface subsidence (Geertsma, 1973; Nagel, 2001; Doornhof et al., 2006), causing major economic damage and social upheaval (e.g. Chan and Zoback, 2007; Muntendam-Bos and De Waal, 2015). Models predicting surface subsidence rely on constitutive relations describing compaction. Linear poroelastic compaction models are frequently insufficient (Hettema et al., 2002; van Thienen-Visser et al., 2015), suggesting that inelastic processes also play a role. An improved understanding of the strain-accommodating processes operating on the grain-scale, and how this scales to reservoir compaction, is necessary for adequate forecasting of surface subsidence.