Abstract
Models of total estimated reserves, of future selling price per barrel of oil, and of the chances that undertaking a sidetrack will not kill a producing well, and that the sidetrack will also be successful, all provide parameter values that are used to estimate the worth of undertaking the sidetrack. However, such parameter values are themselves uncertain and therefore contribute uncertainty to the sidetrack worth. To decide to undertake a sidetrack requires that some idea is to hand of the sensitivity of the assessed worth to parameter variations, so that one knows where to place effort to minimize the uncertainty of the estimated worth. In this paper we show how models of such effects can be incorporated into the evaluation of sidetrack worth and its sensitivity to parameter uncertainties. In addition we show how the relative contributions of ranges and distribution types of the uncertainties can be evaluated. In this way one knows precisely where to focus efforts to improve matters prior to undertaking the sidetrack, and one also determines the range of likely outcomes with some statistical sharpness should one decide for or against undertaking the sidetrack. Several simple numerical illustrations are given to demonstrate the influence of different choices of parameter models on the assessment of the best worth and its uncertainty for a sidetrack situation.