Probabilistic Approach to Reservoir Performance Predictions
Top-Down Modeling (TDM) versus Experimental Design/Design of Experiments (DoE) — Faster, Fully Probabilistic Reservoir Forecasts — Worldwide
Assign your company's probabilistic reservoir performance prediction challenges to A&A, and let the high-level, seasoned consultants of the A&A Consulting Team choose — and apply — whichever method actually fits your data and your deadline, Top-Down Modeling or Design of Experiments.
Avasthi & Associates, Inc. (A&A) is an independent petroleum engineering consulting firm. A&A does not sell reservoir-simulation software or machine-learning platforms — our only business is expert technical consulting, which means our recommendation of TDM, DoE, or a traditional deterministic approach is based solely on your project's data, objectives, and timeline, never on a stake in any particular software vendor or technology.
Use of a deterministic approach for assessing petroleum resources and predicting reservoir performance through the different stages of an oil & gas field's life — acquisition, exploration, appraisal, development, production, and management — is generally in decline, as the industry increasingly recognizes that subsurface uncertainties govern a range of outcomes, not just a few single-valued assessments.
While a probabilistic approach has been widely employed by both majors and independents, especially during the exploration and appraisal stages, several challenges remain ongoing concerns:
• Identification of key uncertainties and the types of distributions that describe them
• Determination of ranges and dependencies
• Successful integration of subsurface geological uncertainties with operational and economic uncertainties
None of these challenges is fully resolved, and each remains an ongoing concern, since they pose significant risk to conceptual field development planning (FDP), project selection, and, ultimately, capital investment decisions.
Why Probabilistic Assessment Matters
Predicting reservoir performance is a critical task at every stage of an oil & gas field's life. Because of the complex physics of the subsurface, reservoir response is typically computed numerically, using a reservoir simulator. But since the input parameters used to build reservoir models carry inherent uncertainty, so do the resulting output forecasts. Correctly characterizing these uncertainties, and probabilistically evaluating the model's response, is essential to helping decision-makers make well-informed business decisions and achieve their business goals.
The Design of Experiments (DoE) Method
The Experimental Design, or Design of Experiments (DoE), method has proven the most popular technique to date, providing a systematic, probabilistic-principles-based approach that yields a suitable range of outcomes for decision-makers to consider. Its disadvantages as an uncertainty-analysis tool include:
• Its reliance on a proxy model as a substitute for full reservoir simulation
• The time and effort required to construct a relatively large set of discrete reservoir models to set up the proxy equations
• Its selection, at the final stage, of multiple reservoir models for the forecast — at which point the method tends to resemble a multi-scenario or multi-deterministic methodology more than a true probabilistic one
Top-Down Modeling (TDM): A Faster, Fully Probabilistic Alternative
Top-Down Modeling (TDM) is a data-driven approach that builds a full-field reservoir model directly from measured field data — production history, well logs, seismic data, and operational parameters — using machine-learning and pattern-recognition techniques, rather than solving the governing flow equations through a large number of full-physics simulation runs. Because TDM does not require constructing a proxy model, as the DoE method does, it delivers a fully probabilistic assessment of reservoir performance at considerable time and cost savings compared with DoE.
A&A's Expertise
The high-level, seasoned consultants of the A&A Consulting Team have the expertise and extensive experience to utilize both the TDM tool and the DoE method — applying whichever is best suited to a given project's data and objectives — resulting in considerable time and cost savings for our clients. This expertise applies to unconventional (shale) and conventional oil & gas fields, onshore or offshore, around the world.
What A&A Delivers
A probabilistic reservoir performance prediction engagement with the A&A Consulting Team is built around the practical decisions your company needs to make, including, but not limited to:
• A recommended methodology — TDM, DoE, or a traditional deterministic approach — suited to your project's available data, objectives, and timeline
• A fully probabilistic reservoir performance forecast, with key uncertainties identified and characterized
• Integration of subsurface geological uncertainty with operational and economic uncertainty, to support field development planning (FDP) and capital-investment decisions
• Presentation-ready results for your company's management, partners, and investors
Training & Knowledge Transfer
In addition, we will be delighted to conduct customized training workshops on probabilistic reservoir performance predictions, for the benefit of your company's professionals and managers, at your company's offices, around the world.
We can put teams of our high-level, seasoned consultants — with extensive experience conducting probabilistic reservoir performance predictions, using the TDM tool or the DoE method — to work on your company's unconventional (shale) and conventional oil & gas field projects, onshore or offshore, whenever requested. To find out how we can help, please contact us.
