I Would Rather be Vaguely Right than Precisely Wrong

Bratvold, Reidar, Steve Begg, Adelaide University, Adelaide, Australia

Over the past two decades many oil companies have consistently under-performed in returning the economic metrics that were the basis of investment decisions. One must con­clude that evaluation procedures fail to properly account for uncertainty, resulting in a sys­tematic over-estimate of returns and/or under-estimate of the risks of loss. Also, in today’s business climate many companies are finding it imperative to do more with less. This requires making faster, smarter decisions that use an appropriate level of technical analysis (detail, rigor) with the acquisition of appropriate data (type, quantity and quality). But what is “appropriate” and how is it impacted by the presence of uncertainty?

It is proposed that answering this question requires a holistic approach that incorporates all of the key components (G&G, Drilling, Production, Facilities, Economics, etc.) that might influence a decision - a Stochastic Integrated Asset Model (SIAM). In this approach we deliberately trade-off precise technical analysis (classical modeling) within any one compo­nent for the ability to model the complex dynamics of the interactions between components, whilst gaining an accurate, global assessment of the impacts of uncertainties. The role of classical modeling changes to one of supporting the SIAM approach by providing uncertain­ty estimates on the input parameters of simplified component models, calibrating them or generating simple surrogates.

Such an approach can identify which “state-of nature” uncertainties (e.g. porosity, satu­ration, oil price) can be ignored, mitigated or need to be resolved, thus focusing classical analysis and/or additional data collection on only those areas that have the biggest impact on the decision criteria. It can also identify which “choice” variables (e.g. numbers of wells, pipeline diameter, …) the investment decision criteria are most sensitive to, thus indicating the value levers that can be used to optimize the investment.