Generalized Extreme Value Statistics, Physical Scaling, and Forecasts of Gas Production in the Barnett Shale

Abstract

We develop a new method of predicting fieldwide gas or oil production from unconventional reservoirs, using the Barnett shale as an illustration. First, we divide the qualifying 13 141 horizontal gas wells in the Barnett into six static samples in which reservoir quality and completion technologies are similar. These samples contain wells of all ages. The Barnett samples coincide with the main gas producing counties, Tarrant, Johnson, Denton, Wise, Parker, and Hood. Second, for each sample, we use a purely data-driven nonparametric approach to arrive at appropriate generalized extreme value (GEV) distributions of gas production from the sample’s dynamic well cohorts with at least 1, 2, 3, ..., up to 14 years on production. We now have up to 14 cumulative probability distribution functions (cdfs) of annual well productivity per sample. From these cdfs, we stitch together six P50, P10, and P90 statistical well prototypes, one per sample or county. Our statistical well prototypes are conditioned by well attr...