Characterization and Derivative-Free Algorithms for Faster Field Development Optimization of Liquids-Rich Shale Reservoirs.

Abstract

Petroleum exploration and development comes with high risks and capital spending. It is an essential goal to balance capital and operational requirements for selecting an optimal field development plan. Understanding the origin of oil, condensate and gas, and the rocks where these fluids are stored helps to achieve that essential goal. This research addresses those issues as current optimization methods applied to field development problems are computationally expensive. Thus, the objectives of this research are to develop methodologies for (1) understanding petroleum generation through millions of years and its link with current reservoir rocks, and (2) devising algorithms and procedures for optimizing fast and at low cost the production efficiency of shale oil reservoirs. The first objective is met with the use of a modified Pickett plot that is extended from a snapshot in time (the time in which well logs are run) to millions of years of burial and maturation trajectory. The approach is explained with data of the Niobrara shale. The plot is further extended for the evaluation of Biot coefficient, which is important to solve drilling and completion problems. Porosity and permeability from drill cuttings are included in the analysis. The second objective is met with the development of an original algorithm, termed in this thesis climbing swarms (CS) algorithm, which is used for well control and design optimization problems. The CS is coupled first with a numerical simulator and next with a material balance. The CS converges faster to a higher quality solution and provides advantages over existing field development optimization methods. An application using Eagle Ford shale data is presented for optimizing oil recovery during Huff and Puff gas injection and re-fracturing operations. It is concluded that the methods developed in this thesis allow faster learning and at lower cost regarding possible field development plans for shale petroleum reservoirs, a task that would be time consuming, tedious and not as accurate, if carried out manually.