REAL TIME UPDATING OF RISK ASSESSMENTS DURING A DRILLING OPERATION — ALTERNATIVE BAYESIAN APPROACHES

摘要

Risk analyses of drilling operations performed in the planning phase of wells often fail to meet their objectives as decision support tools as they become outdated when more information is made available during the drilling operation. To ensure that drilling operation risk analyses at all times reflect the available knowledge and that decisions are made on a sufficiently strong basis, the risk analyses should be updated as valuable information reducing the uncertainty about important quantities is acquired during the drilling operation. Such updating requires procedures that automatically update uncertainty assessments as new information arrives. In this paper we present an updating procedure for assessments of the formation pore pressure in a reservoir, within a Bayesian context. We show that the problem can be solved by different Bayesian approaches, generated by introducing or not introducing an underlying probability model. Our recommended approach is based on assessment of uncertainties about quantities that can be given a physical interpretation. The drilling operation is the starting point for the paper, but the discussion is to large extent general and applies to other areas as well.