A challenge in the oil industry is the ability to forecast uncertainty reduction resulting from new information. This is particularly important in exploration, appraisal and pilot programs, where management weighs the amount of information expenditure needed to make drill or development decisions. Although a key decision criterion for this is often value, another important criterion is uncertainty reduction, or confidence in the interpretation of critical uncertainties.

So how can teams understand project uncertainties and the effects of different information or appraisal options on their uncertainty assessments? Through an efficient, multi-discipline reliability of information assessment interview process designed to help the team think clearly and openly about the amount of uncertainty reduction to be gained for different information/appraisal programs.

The product is a series of confidence graphs depicting uncertainty reduction versus information options. The plots are developed using Bayesian mathematics and the reliability of information assessment interviews. They address: (1) the probability of a correct interpretation; (2) the probability the actual outcome will be the low outcome after having forecast a P50 outcome; (3) the probability the actual outcome will be the low outcome after having forecast a high outcome; and (4) the probability of getting a given outcome after having correctly forecast that outcome from the information (e.g., the probability of getting a P50 outcome after forecasting a P50 outcome from a test).

The important aspects of this paper are the set of reliability of interpretation questions used to forecast uncertainty reduction potential, how to conduct those interviews and how to develop the resulting confidence plots. The questions address the drivers of uncertainty reduction potential: information tool accuracy; effect of the environment from which the information is being collected; the tendency of the interpretation to better forecast true low, true P50 or true high states of nature; and the representativeness of the information.

You can access this article if you purchase or spend a download.