Creating Geologically Realistic Models Used for Reservoir Management
ExxonMobil Upstream Research Company
To make sensible reservoir management decisions, it is necessary to predict future reservoir performance. This allows testing and optimization of reservoir management strategies before making large investments. When displacement mechanisms change or geologic description is different from current well locations, this prediction is usually done with reservoir simulation models. Because geologic features determine the connectivity and productivity of the reservoir, it is important to ensure that models realistically represent the reservoir description in order to provide plausible predictions. Challenges associated with constructing these models include:
- Uncertainty in the geologic description – measurements are sparse, and do not always resolve the relevant features.It isn’t always known which features are relevant to reservoir performance.
- Geometry and stacking of geologicobjects like channels and lobes are difficult to represent in cellular models
- Multiple descriptions may exist that are consistent with available data
This presentation describes how reservoir models are used in making reservoir management decisions, and outlines a strategy for creating realistic reservoir models. Examples are provided of applying some elements of this strategy.
Dave Stern is a career researcher at ExxonMobil’s Upstream Research Company (URC) He joined URC in 1984 with a PhD in Chemical Engineering from University of California at Berkeley and a BS in Chemical Engineering from MIT. Research areas include experimental measurement of gas injection performance, development and use of simplified models for reservoir management, gridding and scale-up, and history matching. He led a team that developed tools for construction of simulation models from detailed geologic models, worked with software developers to implement that technology, and trained the rest of the corporation in its use. Dave is the author of an SPE distinguished author paper on practical aspects of gridding and scale-up, describing learnings from that experience. Dave also led a team that developed tools and methods for history matching, with emphasis on preserving geologic realism during the history match process. The team worked with software developers to implement the tools, and trained the rest of the corporation in their use. He is currently a Reservoir Engineering Advisor to a large project that develops and maintains software for reservoir modeling and simulation.
Dave is a career-long member of SPE, and has served as session chair or discussion leader in SPE forums on gridding and scale-up, reservoir modeling for asset teams, and data analytics.