Exploration Chance of Success Predictions — Statistical Concepts and Realities

Abstract

There is much confusion in the conceptualisation and application of Chance of Success (COS) Predictions in oil and gas exploration. Although the basic statistical underpinnings of COS predictions are not mathematically complicated, in practice, there appear to be significant difficulties. The consequences of this in many cases include misplaced expectations and hence morale problems from results of exploration which fall outside expectations. In reality, commercial exploration success rates worldwide range from 30–40%. So, there is more pain than not in our industry with the unfolding of expectations. As a result of this, companies have many times reacted in a knee jerk fashion to ‘correct’ their course which sometimes results in restructuring exploration teams and also changing the course of exploration. Much of the misunderstandings appear to arise from the fact that most small companies are involved in limited trials campaigns where budgets allow the drilling of only a handful of wells over 1–5 years. Realistic COS’ can only be based on expectations related to drilling a statistically significant large number of wells. In this paper the vagaries of the actual unfolding of exploration results are simulated using MS Excel software’s Random Number Generator function. Despite the intrinsic difficulty of not being able to guarantee any specific success, it will be shown how companies can choose the COS range they should be involved in to enable sustainable growth over the longer term.