Abstract
A comprehensive overview and analysis of the productivity of 1216 recently abandoned multi-stage hydraulically fractured horizontal wells from five shale formations in the United States (US) is presented in this study. In this study, two decline curve analysis (DCA) methods were used to match actual production history data using least-squares fitting to find the best fit production parameters to reliably forecast production. The production history matching conducted resulted in very accurate matches (correlation coefficient of 0.99) between actual production data and the two DCA methods (Arps hyperbolic decline and stretched exponential production decline (SEPD) models). Using the outcomes from production history matching, universal averages of decline parameters for Arps hyperbolic decline and SEPD models were developed for each of the five formations. Furthermore, hindcasting was performed by matching a portion of the known production history and comparing the remaining portion of the known production history to the forecast. The Arps hyperbolic decline and SEPD methods were used to match production using only limited early production data (three months, six months, one year and two years). The main goals for fitting the DCA model to early production data was to estimate the optimum decline parameters that are then used to forecast production and estimate ultimate recovery. Production history matching using limited early production periods produced accurate production forecasts using as few as six months of production history (correlation coefficients between 0.85 and 0.94 using Arps hyperbolic decline). The main outcome of this study was a production analysis conducted on the production data of more than 1000 wells from five different shale formations to present the expected production behaviors of similar wells. Different production key performance indicators (KPIs) such as average well life, cumulative production volumes at different periods, average drop in production rate within the first year of production, average time to reach maximum flow rate, and the maximum flow rate were measured on all the wells from the five formations to provide an overview of the production performance of each formation.
Highlights
History matching is defined as the act of adjusting a model of a reservoir until it closely reproduces the past behavior of a reservoir [1]
Production data analyses performed on 1216 multi-stage hydraulically fractured horizontal wells from five shale formations (Niobrara, Barnett, Haynesville, Eagle Ford, and Bakken) that were recently abandoned
Production history matching with two decline curve analysis (DCA) models was performed using least-square fitting on all the wells from this study
Summary
History matching is defined as the act of adjusting a model of a reservoir until it closely reproduces the past behavior of a reservoir [1]. Ilk et al (2008) introduced the power law exponential decline (PLED) method for future performance forecasting and estimated ultimate recovery (EUR) prediction of tight gas and shale production They presented the “power-law loss-ratio” rate/time relation, which assumes that the decline rate follows a power law function on a log–log scale at early time and becomes constant at late time. Several of the common DCA model have been compared in the literature, for instance, Wilson (2015) compared different DCA methods (Arps, Power law exponential (PLE), Duong, and SEPD) with four unconventional gas and oil plays: the Barnett, the Woodford, the Bakken, and the Eagle Ford Their conclusions were that most DCA methods produced great history matches for up to 4–8 years (for all four plays), but in forecasting they differ a lot [15]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
