Problem: If all the information we have to go on is the raw success versus failure statistics of an analogous set of wells, how do we translate the number (75%) representing the frequency of past success into a go-forward prediction of the chance of success of an undrilled prospect (Pg)?
(A) Application of the Laplace (1774) sunrise problem, estimating the mean chance of success from the observation of a succession of 1 to 11 successful results. (B) Application of the rule of succession to estimate the mean chance of success based on a series of 1 to 11 results, including success and failure outcomes.
Illustration of scenarios in which past statistics may be a valid indication of future performance (undrilled prospects in disputed territory) and where they should be viewed with caution (a farm in an extensively drilled block).
Location of the Overlapping Claims Area (OCA) in the northern Gulf of Thailand, which has remained essentially unexplored. The adjacent part of the Pattani Basin, which lies in Thai-controlled waters, has been extensively drilled. Map compiled from online sources including Committee for Coordination of Joint Prospecting for Mineral Resources in Asian Offshore Areas (CCOP) (2002) and KrisEnergy (2015).
Outline map of the frontal fold trend of the Western Atwater fold belt, Gulf of Mexico (gray area on main map), showing the major structural traps with their original prospect names, with outlines from Moore et al. (2001) and Dias et al. (2010). Geological success cases at lower Miocene level are shown in solid black (Neptune, Atlantis, Mad Dog, and Shenzi oil fields); failure cases at lower Miocene level are shown in white. Note that some of these failure cases encountered oil at different stratigraphic levels or encountered dry gas in place of prognosed oil. Small squares are US Bureau of Ocean Energy Management Outer Continental Shelf lease blocks.
Sequence of drilling results for the first well testing a structure in lower Miocene play of the Western Atwater fold belt, showing a comparison of a prediction of the chance of success (Pg) of the next well, based on unmodified past frequency (open diamonds and gray line) compared with the probabilistic prediction using the rule of succession (black diamonds and black line). Black circles indicate geological success, and dry hole symbols (open circles with crosses) indicate that predrill geological model was not a success; note that some of these failure cases encountered hydrocarbons at other stratigraphic levels (e.g., Puma) or encountered dry gas in place of the prognosed oil (e.g., Frampton).
A simple nomogram approach to moderate a statistical prediction based on limited data. An estimate is made based on global data alone (a), such as the global frequency of stratigraphic trap success, adapted from Binns (2006), and a second estimate is made using only local data (b). The user makes a judgment as to the relative quality of the local data set (c), and the intersection gives an appropriate modified estimate (d).