Look-backs on major capital projects have suggested that technical teams did not fully assess the range of subsurface interpretations. Specifically, many projects did not provide the production rates and reserves that were promised …the result of either biasing results to a more favorable outcome or anchoring on deterministic interpretations and development plans that did not take into account the full range of possible uncertainty values. This has ushered in the era of probabilistic subsurface assessments.
A key breakthrough in probabilistic assessments has been the adaptation of Design of Experiments to subsurface modeling on both the geologic and reservoir engineering disciplines. Since the list of subsurface uncertainties can be quite long, and the ranges of uncertainties themselves can be wide, a full assessment of reservoir performance outcomes would require hundreds or thousands of geologic and reservoir simulation models. Design of Experiments has provided a structure to this assessment process; but more importantly, DOE can significantly reduce the number of interpretations and runs needed to fully assessment reservoir performance outcomes and the main effects of each of the uncertainties.
This talk will present a structured method employed by Chevron to create unbiased P10-50-90 models of reservoir performance, which are then used to assess field development concepts. Mark will also present some lessons learned and best practices related to using DOE and incorporating that into field development planning.