Three-dimensional reservoir modeling has become an accepted tool in reservoir description and is used for various purposes, such as reservoir per-formance prediction or integration and visualization of data. In this case study, a small northern North Sea turbiditic reservoir was to be developed with a line-drive strategy utilizing a series of horizontal producer and injector pairs, oriented north-south. This development plan was to be evaluated and the expected outcome of the wells was to be assessed and risked.

Detailed analyses of core, well log, and analog data have led to the development of two geological end member scenarios, thus accounting for uncertainties associated with the geological model. Both scenarios have been modeled using the sequential indicator simulation method in a hybrid deterministic-stochastic approach. The resulting equiprobable realizations have been subjected to detailed statistical well placement optimization and analysis techniques. Based upon bivariate statistical evaluation of more than 1000 numerical well trajectories for each of the two scenarios, it was found that the inclinations and lengths of the wells had a greater impact on the wells' success, whereas the azimuth was found to have only a minor impact. After integration of these results, the actual well paths were redesigned to meet external drilling constraints, resulting in substantial reductions in drilling time and costs.

Although three development wells drilled subsequent to this study were very successful, their outcome raises questions about the validity of the stochastic model, which is based on geological assumptions which, in turn, were derived from much fewer well data. It is clear that a better quantitative sedimentological understanding of the reservoir, specifically in the lateral dimension, would have resulted in a more reliable reservoir model.