Improved oil recovery estimation with data analytic methods: A workflow for open data analysis

Reserve evaluators use the Secondary to Primary recovery ratio (S:P ratio) to infer a first-order reserve estimate for Improved Oil Recovery (IOR) project screening. Determining S:P ratios via analog processes (e.g. data gathering, screening properties, reserve estimates, etc.) is tedious; implementing data analytic methods improves our understanding of the risks and uncertianies associated with reserve/resource estimates for IOR projects by allowing evaluators the ability to analyze significant amounts of data readily. The objective of this paper is to illustrate an application of data analytic methods for IOR performance estimates using geospatial datasets with historical waterflood records in the State of Texas. A spatial-temporal correlation workflow is applied to quantify the uncertainty of IOR estimates based on spatial proximity and S:P ratios. Multivariate regression was applied to predict IOR estimates and validated with exsisting projects. The results show spatial proximity by itself is not a reliable method for the IOR estimation; S:P ratios vary significantly with nearby analgous projects. In addition, the uncertainty analysis herein provides an insight to the reliability of IOR estimates. The example modeling results provided in this paper are spourous, and not to be taken as a final reserve/resource analysis; the analytic methods presented in this paper can serve as tuning and screening tools for performance prediction of IOR prospects.