Chapter 7: Probabilistic Resources Estimation

The prediction of reservoir properties and quantities (such as petroleum initially in place and Technically Recoverable Resources) is subject to uncertainty due to the limited reservoir sampling provided by data control points (e.g., wells), yet these predictions are major determinants in reservoir performance forecasting. These production forecasts are then used for decision making, driving investment in future projects where, in many cases, large sums of capital are required.

Industry uses various methods to understand and quantify the impact of reservoir uncertainty on production forecasts and estimates of resources. These vary from purely deterministic methods, where single values of each parameter are considered, to multi-scenario methods (hybrid methods considered an extension of deterministic methods), to fully probabilistic methodologies, where input parameters (e.g., porosity) are defined by probability distributions that are then combined to obtain cumulative distributions of the outcomes (e.g., in-place estimates).