Capturing reservoir production uncertainty for channelized reservoirs using channel amalgamation indexes

One of the main concerns in reservoir studies is to accurately define the internal architecture and the geological characteristics of the reservoir so as to estimate the amount of hydrocarbons that could be recovered for a given development strategy. This can represent a major challenge especially during the appraisal stage of a reservoir, because the information available is still very limited, or in the presence of geological heterogeneities, which increase the architectural complexity and the uncertainty associated to the internal description, such as in channelized depositional settings. At the appraisal stage of a reservoir study, all the uncertainties affecting the quantity and distribution of hydrocarbons in the reservoir should be captured and accounted for in the evaluation of the final hydrocarbon recovery to properly assess the viability of any development plan. A typical modeling workflow accounting for geological uncertainties consists in creating a large set of 3-D stochastic geological (static) models from a set of geological input parameters. Subsequently, a few representative reservoir realizations are selected out of this set based on the calculated hydrocarbons originally in place and simulated to estimate future production so as to propagate the uncertainty onto the final recovery factors. However, even in homogeneous reservoirs, the estimation of the hydrocarbon stored in the reservoir can be affected by uncertainties because it is calculated mostly from local petrophysical parameters, which might not be representative of the rock properties at the reservoir scale. This especially applies to channelized reservoirs characterized by depositional elements with high geological heterogeneity, both in the lateral and in the vertical directions. Thus for these depositional settings a more attractive criterion for the model selection is offered by the study of the connectivity layout of the reservoir elements. In the technical literature, connectivity is defined through numerical indexes that account for geological connectivity between reservoir elements, which might not be indicative of reservoir production performance. In fact, the latter is influenced by the degree of connectivity among sand bodies and only deep merging of the channels guarantees that the reservoir can be efficiently drained by just a few wells. Therefore, in the first place, the present study was aimed at thoroughly investigating the validity of the indexes previously proposed in the technical literature by evaluating the reservoir production uncertainty associated to sets of synthetic equi-probable models of channelized oil reservoirs. Secondly, the goal of the research was to develop new indexes to express the channel connectivity, capable of incorporating information on the quality of the connectivity through the evaluation of channel amalgamation. When applied to the same set of reservoir equi-probable realizations, these indexes proved that a more effective selection of the geological realizations can be made to capture the uncertainty affecting the forecasted reservoir production performance.