ABSTRACT: A multi-scenario geomechanical modeling approach is presented to investigate the potential mechanisms that lead to a series of observed well deformation incidents in the deep overburden overlying a depleting and compacting chalk reservoir in the North Sea. Complementary to the commonly pursued approach of building a full-field geomechanical model, a suite of detailed ideal geomechanical models is developed based on data and evidence driven scenarios that take into considerations fault population and juxtaposition, contrasting mechanical stratigraphy, intra-bedding contact conditions, and lateral persistence of thin stringer beds. The simulation results of extensive runs suggest that as the majority of faults in the overburden mudstones would remain stable under the pre-production stress state, however, the contrasting mechanical stratigraphy consisting of thick compliant mudstone and thin stiff carbonate stringers facilitates “domino” or “bookshelf” style intra-bedded shear slippage in response to the developing dominant shear zone around the periphery of the active compacting region. This multi-scenario approach provides new and accelerated insights into the diagnosis of compaction-induced well failure and allows mapping of competing mechanisms potentially responsible for the problem, which helps to steer further modeling efforts, and guides future data acquisition.
Despite previous industry studies [1-7] and continuing investigation and modelling efforts, there remain challenges with regards to the understanding of key mechanisms relevant to well casing damage and deformation due to reservoir compaction, subsidence and corresponding deformation of the overburden. These challenges are particularly associated with insufficient data collected outside the reservoir which inhibit diagnosis and appropriate analytical and numerical techniques that are able to replicate and predict the occurrence of well incidents. As a commonly pursued approach, full-field geomechanical models practically loaded with all available information are perceived the way to go, as reported in previous studies [1, 3, 5] and a companion publication . In this study we consider some key technical challenges in modeling the overburden deformation:
(i) Recognizing that geological faults are necessary in the understanding of well deformation in the overburden, full-field models are constrained in terms of representing the complete fault system present in the volume of interest.
(ii) The full-field model suggests that faults under the current in-situ stress states are far from being critically stressed, as opposed to initial expectation during model framing.
(iii) The effect of reefs and key mechanical layers, though introduced and assigned mechanical properties relevant to field data, do not appear to show an impact.
(iv) The coarse mesh resolution is suboptimal for geological faults and detailed mechanical stratigraphy consisting of thin layers of strong property contrast.
(v) The effect of other forms of geological discontinuities, such as lithological contacts, are completely unaccounted.