Timothy McNealy and Mohammadreza Ghasemi
Halliburton Energy Services
One of the challenges in unconventional gas/oil reservoirs is characterization and modeling. There are different models for describing the production in an unconventional reservoir, i.e. dual porosity, triple porosity, etc. In this paper we generalize the solution for a Multi-Porosity Model (MPM). Different MPMs are considered and an appropriate measure will be discussed to select the best model that describes the production data using any known reservoir properties. After selecting the best model, the sensitivity analysis will be done to determine the main parameters that effect production. These parameters can be estimated by matching the MPM to the well’s production history. All reservoir models have the inherent flaw of generating non-unique solutions, more than one correct answer. We offer a solution to this challenge by randomizing the input variables and generating a statically probably of the unique solution. Two wells have been analyzed and validated through history matching and used for forecasting production.