Impact of Integrated Geological and Reservoir Modeling Best Practices on Production Forecast Accuracy

Abstract

Major capital project subsurface workflows generally include a significant geological (static) modeling phase followed by a reservoir (dynamic) modeling phase from which reservoir forecasts of fluid/gas production over time are used to justify development decisions. A recent survey of industry major capital projects by Nandurdikar and Wallace (2011) showed that actual production is less than 75% of forecast production at project sanction and for projects with significant “reservoir issues” (as opposed to well or facility “issues”), the actual production was only 55% of that forecast. A variety of data from clastic and carbonate conventional reservoirs, unconventional reservoirs, and synthetic reservoirs been used to better understand the significant sources of forecast optimism and its mitigation. Analysis suggests that about 10–30% of the reported forecast optimism is due to sparse data and/or analog bias, about 15–40% is due to static and geological and reservoir modeling workflow decisions including the use of well location optimization, and the remainder due to human and/or management induced bias, perhaps compounded by unnecessary model complexity. Reservoir forecasts based on static geological and dynamic reservoir models can be improved by recognizing and addressing sources of bias through the use of sound, statistically rigorous uncertainty assessment and proper understanding of the limits imposed on model derived forecasts due to model workflow decisions. The use of external peer review teams, look-back studies, and best practice guides integrated across geological and engineering disciplines may also significantly reduce unintentional technical or management-driven bias.

AAPG Datapages/Search and Discovery Article #90216 ©2015 AAPG Annual Convention and Exhibition, Denver, CO., May 31 - June 3, 2015