Computer-Aided Methodology for the Analysis, Design and Optimization of Production from Unconventional Gas Reservoirs.



Journal Title

Journal ISSN

Volume Title



During the past decade, the production of unconventional natural gas, particularly gas from shale rocks, grew to reach more than 50 percent of the annual U.S. natural gas output. The term unconventional refers to gas contained in rock formations of very low permeability, which makes gas extraction difficult, because - in contrast to conventional resources - gas cannot easily flow through the reservoir rock into a drilled well and travel to the surface. There are two key technologies that have made production from unconventional resources practical: Massive hydraulic fracturing and horizontal drilling. However, the systematic development of unconventional gas resources entails substantial uncertainty and risk. Therefore, computational tools that mitigate such risk are valuable for economic development of such resources. This research focuses on developing such tools, and is divided into two parts.
The first part focuses on a methodology for the analysis of production data from existing wells that can be used for future well planning. The methodology relies on standard principal component analysis (PCA) and regression (PCR), and can help answer questions such as (a) Which wells behave similarly? (b) Which wells behave differently from each other or from standard expectations? (c) What factors contribute to these differences? (d) How can data from existing wells be used to anticipate the performance of new wells? The methodology is illustrated through the analysis of historical production data from twelve wells in the Holly Branch field. The proposed methodology would be even more valuable for larger data sets, for which manual analysis of production data is more cumbersome. The second part addresses the problem of fracture design and optimization, namely, decision making on the optimum number of horizontal wells, the optimum number of transverse fractures per well, fracture dimensions, and the quantity of proppant required per fracture. For this problem, the usual strategy of parametric sensitivity analysis is time consuming and possibly ineffective. The proposed approach accounts for the complex interactions among formation and fracture properties, fracture geometry, production behaviors, design constraints and the objective function. Explicit analytical expressions are developed that can be easily used to implement the proposed methodology on problems in the field.



Production data analysis, Principal component analysis, Production optimization, Unconventional reservoirs, Unified fracture design, Mathematical Optimization, Multiple transverse fractures, Horizontal well