Challenges in Building Consistent Multi-Scale Geological Models
Geological models are necessary for several different applications span many different scales. Large scale basin models are typically built for play fairway analysis; regional scale models are often designed to evaluate hydrocarbon resources, identify sweet spots, and optimize well placements. Finer scale wellbore models can also be designed for understanding the impact of lithological variation along horizontal wells and for better completion design. Often, these models are built in separate efforts by different teams of subject matter experts (SMEs), each with a focus on the data that are available at the scale relevant to their problem. For example, seismic data would typically be deemed too low in resolution to be relevant for completion design, but would be one of the primary sources of conditioning data for regional models. Because each team uses different sources of information, parameters and assumptions they often develop inconsistent models that lead to significantly varied conclusions. This paper presents the key challenges involved in building a single model that is consistent across multiple scales while honoring all of the available information. These include (1) computational challenges, (2) methodological challenges, and (3) data challenges. The computational challenges relate to run time and memory, both of which can become prohibitive when trying to simulate very large regions with very fine grids. Methodological challenges relate to the need to ensure consistency when properties are being correlated, not only spatially but also across multiple scales, each of which can have unique interpolation parameters, including variograms, histograms, facies proportions, and associations. Data challenges relate to the availability of very dense information at a specific scale and sparse information at other scales. Data from West Texas are used in this presentation to demonstrate and discuss how these challenges complicate building for the entire Permian Basin a single, consistent, multi-scale model that can be examined at any resolution. Such issues and proposed solutions arising from these Permian Basin examples are relevant to other geographic regions and depositional settings.
AAPG Datapages/Search and Discovery Article #90341 ©2019 AAPG Geoscience Technology Workshop, Exploration and Development of Siliciclastic and Carbonate Reservoirs in the Eastern Mediterranean, Tel Aviv, Israel, February 26-27, 2019