Iterative Bass Model forecasts for unconventional oil production in the Eagle Ford Shale

•Shale oil production forecasting is improved using a Bass Model.•Initial URR estimates for the Eagle Ford appear to have been too low.•Bass Model forecasts offer an alternative to bottom-up methods.•The Bass Model can provide a useful heuristic to policymakers with limited resources.

Developing long-term forecasts for unconventional oil, gas and condensate production has challenged both analysts and academics because the shale revolution is still in its relatively early stages. Initial estimates of URR (ultimately recoverable reserves) in the Eagle Ford appear to have been too low. In this paper, forecast accuracy using limited early data of oil and condensate production in the Eagle Ford has been improved over OLS (ordinary least squares) methods using a Bass Diffusion Model that could be applicable to other shale field developments in the U.S., and eventually in other countries as well. Using only preliminary production data from 2006 through 2010, a Bass Model yields more accurate early predictions than conventional, bottom-up, data/labor-intensive OLS regression approaches. Further, the Bass Model suggests that in the absence of the recent oil price decline, the Eagle Ford oil and condensate production could have reached as high as 2.6 million barrels per day by 2020, significantly above a recent energy industry analyst prediction of 2 million barrels, and far higher than OLS regression forecasts that ranged from between 450,000 and 1.4 million barrels per day.