In petroleum industry, scarce information is available at the time of adopt the exploitation strategy, and the expected production profiles comprise a high level of uncertainty. Depending on the range of the uncertainty variables, the best decision may be to collect more information by drilling more wells, performing a new seismic acquisition or making long term tests. But, sometimes, the best way to protect the project is to assume an uncertainty-proof development strategy. These ideas were applied to help decision in two real cases, in Marlim Sul field, Campos Basin, Brazil. In the first case, the main uncertainty source is the quality of the flow transmissibility between producers and injectors wells. If that communication is restricted, the production decline is more accentuated and the best strategy would be to place the injectors nearer the producers, although the displaced oil is reduced. In order to elect the best strategy, we have compared two different plans containing different positions of injector wells, applied to three different scenarios of flow transmissibility. In the second case, we have a channeled reservoir, with an extensive fault in the middle, the transmissibility along which is unknown. In a scenario of good communication, the best strategy is to position the injectors on one side and the producers on the other side of the fault. But in scenarios of restricted communication, the best scheme could be to place injectors and producers on both sides of the fault. To make the best decision, we have used different plans, applied to different scenarios. The results demonstrate that, in both cases, the alternate plan has not the higher net present value (NPV) in the moderate scenario, but presents the higher expected monetary value (EMV), having also a lesser sensitivity to the reservoir uncertainties, being more protected to the risk. In both cases the original plan has been abandoned and the alternative plan started being implemented.

You can access this article if you purchase or spend a download.