Post-Appraisal – the Key to Confidence in Forecasting

Abstract

Any and every forecast, ranging from tomorrow's weather to the climate of the future, and from prospect EUR to a basin and global hydrocarbon endowments, are attempts to quantify uncertainty. Weather forecasters and geoscientists are well versed in both probability and volume estimates (chance of rain and the range of tomorrow's rainfall / chance of geological success and the range of the volume of hydrocarbons for a prospect). Leaders in these professions take pride in using post-appraisal data to calibrate their future estimates. They also understand the importance of both calibration – wells with a Pg of 40% are successful about 40% of the time; and resolution – high risk wells nearly always fail and low risk wells nearly always succeed. High performers in these fields have also explicitly recognized their inherent biases and taken steps to correct or ameliorate such biases. Successful organizations are capable of maintaining high levels of performance over successive generations of practitioners – for example a leading E&P company has a 20-year track record of calibrated chance and volume estimation. A common factor that supports these calibrated processes is that there have been a sufficiently large number of independent forecasts to enable these dedicated teams to study their estimates closely, make adjustments, and keep their calibrations accurate. In contrast, there are forecasts that, at least on the surface, lack the high frequency feedback essential to obtain and maintain calibration and resolution. For example – in 10 or in 20 years time, what sort of hydrocarbon deposits will be being discovered and produced?; where will the active hydrocarbon provinces be?; what depths and pressures will need to managed?; what skills will be needed then?; what will the demographics of the profession be? This presentation post-appraises the forecasts for 2016-17 that were made in 2002 and 2004. The controls on the accuracy, or otherwise of these forecasts is not the result of high frequency results, but rather the effect of aggregation. Aggregation is the ability to incorporate a very wide variety of diverse inputs into each forecast. In turn, this analysis leads to simple yet effective methods to forecast the future of our E&P industry, and the multiple interlocking components that make it work.

AAPG Datapages/Search and Discovery Article #90291 ©2017 AAPG Annual Convention and Exhibition, Houston, Texas, April 2-5, 2017