Application of Big Loop Uncertainty Analysis to Assist History Matching and Optimize Development of a Waterflood Field

Abstract

Abstract Objectives/Scope: This paper showcases how an integrated static and dynamic modelling workflow was applied to cover a wide range of subsurface uncertainties in order to generate a more robust range of forecasts. These forecasts are in the process of being used to optimize the ongoing development plan for a marginal offshore field in South East Asia. Due to an anticipated lack of aquifer support, the initial phase of development was based on flank and crestal producers with peripheral water injection to deplete Miocene-aged reservoirs within a fluvial-dominated coastal plain to delta environment where the predominant lithologies are clean sandstones and sandy to muddy heterolithics. The results from the early development wells indicated a more complicated reservoir than has been previously interpreted, as evidenced from well results, formation pressure data, flowing /shut-in pressure trends and early production data across the field. Methods, Procedures, Process: For a field with an increasingly complex geological and dynamic background, use of an integrated workflow covering both the static and dynamic parameters allows for seamless updating and ensures continuity between geological and dynamic models when assessing the field uncertainty. This is achieved by predefining value sets and their uncertainty ranges and employing an experimental design technique. The information for each realization is used to generate a proxy model, which is then used to interpolate an objective function between runs and thus manage uncertainty. The proxy model provides a quantified distribution for each combination of input parameters in the reservoir simulation model with approximate results for any set of modifier values then generated. The methodology provides a better understanding of subsurface risk assessment in conjunction with assisted history matching, thereby optimizing the development plan. Using an integrated approach allows the geological model to be current at all times during the drilling campaign and potentially allows for optimization during the campaign. Results, Observations, Conclusions: This case study presents the results of the Integrated Uncertainty Analysis Workflow which resulted in sets of realizations of multiple scenario-based history matched results, whilst maintaining geological consistency between the static and dynamic models for reservoir behaviour predictions. The oil production and recovery ranges were then determined from the probability and cumulative density function resulting from these multiple realizations. The effectiveness and challenges encountered when applying the workflow in a real field example will also be discussed in this paper. Additional Information: Risk can be quantified and uncertainty managed within both static and dynamic models through an integrated workflow that covers the entire reservoir modelling process from seismic interpretation through to simulation and reservoir predictions. This implies significant cost savings, as this workflow provides results quicker and generates a comprehensive quantification and management of uncertainty within the reservoir modelling workflow. The geologically-consistent models enabled the delivery of an optimized reservoir model for waterflood management; a probabilistic approach to production forecasting; and crucial input into maximizing recovery and optimizing the field's economics. Such a workflow has the potential to be updated in near real time, and executed in a time space in line of well construction – thus optimization of well placement based on changes in the underlying geology, whilst maximize recovery are potentially possible.