Abstract
This paper presents a methodology for the probabilistic analysis of an infill or step-out opportunity using numerical simulation. Sensitivity and uncertainty analyses for all involved parameters were evaluated through different experimental design techniques. Subsequently, a proxy model was established to reproduce the numerical model performance. Finally, three appropriate solutions were selected from a large population of realizations corresponding to probabilistic percentiles (90%, 50%, and 10% certainty that the specified volume will be recovered).
This proposed methodology helped the asset team to evaluate the well candidates more precisely, confidently, and in less time than the current standard methodology. More knowledge about the variables and their effects on overall outcomes was also gained, which helped the team make more-informed decisions. The workflow used the same numerical modeling software, incorporating and facilitating the changes of both static and dynamic properties simultaneously.
A case study from Teak field, on the east coast of Trinidad, illustrates the applicability of the methodology and compares its results to those obtained using the standard workflow for the asset. The methodology is one of the latest developments in reservoir simulation, and it has not yet been incorporated into the operator's common practices and procedures for exploitation of the TSP fields.
Better state pictures facilitating state machine characteristic conjecture
