Methods for Evaluation of Wellbore Collision Probability in Critical Close-Proximity Regions

Summary A previous paper (Bang, 2017) has described a general method for calculation of the probability of collision between two wellbores. The present paper focuses on the sections along both wells where the collision probability is significant. Our hypothesis for this study was that for such limited sections of the wellbores, approximations and alternative mathematical approaches might increase the computational efficiency without too large sacrifices of the result's accuracy. The potential benefits have been investigated through numerical calculations. The results indicate that using a constant uncertainty instead of the actual, varying uncertainty across the entire close-proximity region may increase the computation efficiency. However, we also found that this approximation should be applied with care, because it may generally be difficult to define the constant uncertainty with sufficient accuracy, based on a priori data. Furthermore, we have evaluated the current method of calculating probabilities in metric space against calculating in Mahalanobis space. These two approaches are mathematically equivalent. For our implementation and limited range of testing, the two methods are about equal with respect to computation efficiency. We have also identified some important topics for future studies. Among these are identification of adequate parameters and efficient algorithms for collision risk analysis at various well proximity ranges, and evaluation of the appropriateness of established model assumptions in light of the industry's current needs and practices.