Insight Into Integrated Reservoir Management Using a Top-Down Intelligent-Reservoir-Modeling Technique: Application to a Giant and Complex Oil Field

Abstract

This article, written by Senior Technology Editor Dennis Denney, contains highlights of paper SPE 132621, ’New Insight Into Integrated Reservoir Management Using a Top- Down, Intelligent-Reservoir-Modeling Technique: Application to a Giant and Complex Oil Field in the Middle East,’ by Amirmasoud Kalantari- Dahaghi, Shahab D. Mohaghegh, and Yasaman Khazaeni, SPE, West Virginia University, prepared for the 2010 SPE Western Regional Meeting, Anaheim, California, 27-29 May. The paper has not been peer reviewed. Top-down intelligent reservoir modeling (TDIRM) integrates reservoir-engineering analytical techniques with artificial intelligence and data mining to arrive at an empirical, cohesive, and spatiotemporally calibrated full-field model. The model is used to predict reservoir performance to recommend field-development strategies. Introduction TDIRM attempts to provide insight into fluid flow in the reservoir by starting with field measurements such as well-production history and well logs. Other data such as core analysis, well tests, and seismic can be used to increase model accuracy. Although not intended as a substitute for conventional reservoir simulation of large complex fields, this approach to reservoir modeling and management can be used as an alternative to traditional reservoir simulation and modeling for cases in which performing conventional modeling is cost and manpower prohibitive. If a conventional model already exists, TDIRM may complement the conventional technique with an independent assessment of the reservoir/well data to optimize development-strategy and recovery enhancement. In this study, TDIRM was performed on a giant complex oil field in the Middle East that has produced for a half-century. The reservoir has produced under natural depletion, with a strong waterdrive maintaining reservoir energy. The formation is a complex mix of clastics and carbonates of variable reservoir quality. This prolific reservoir has highly complex flow behavior. Methodology TDIRM uses accepted reservoir-engineering techniques such as decline-curve analysis, type-curve matching, (single-well) production-history matching, volumetric-reserves estimation, and calculation of recovery factors. These analyses are performed on individual wells. By use of statistical techniques, multiple production indicators [i.e., first 3-, 6-, and 9-month cumulative production and the 1-, 3-, 5-, and 10-year cumulative oil, gas, and water production and gas/oil ratio (GOR) and water cut] are calculated. To model the behavior of nondeclining-production characteristics such as GOR and water cut, intelligent decline-curve analysis is performed on GOR for which a negative decline is normal behavior. Intelligent decline-curve analysis is specific to TDIRM, and it is performed by use of an adaptive time-series-modeling technique that trains itself with existing data and predicts system behavior for a specific length of time into the future.