DPR polymer gel treatment in oil reservoirs: A workflow for treatment optimization using static proxy models

Abstract

Excessive water production from oil producing reservoirs has become one of the operators’ major concerns. Various control methods have been used in recent years; however, they all have high degrees of uncertainty. Disproportionate permeability reduction (DPR) polymer gel treatment is classified as one of the chemical methods of water reduction and the risk which is associated with this type of treatment urges us to analyze it in a more detailed and exact way.There is a need to have an accurate analysis of gel treatment behavior before applying it in a reservoir. There are various parameters that affect the gel treatment operation and even these parameters have interactions between themselves. If they can be found, in a simulator their effect should be measured and a good combination of them should be represented to maximize the profit of the process. Most of the literature works have not represented a comprehensive work that covers all above issues. Rarely a detailed analysis for screening the effective parameters and their interaction and their effect on final profit has been represented so far, and of course using simulator for testing distinctive values on the model and Net Present Value (NPV) is an expensive and time-consuming method. In addition, it should be mentioned that most previous studies were focused just on simple models (not three-dimensionalthree-dimensional, three-phase) for simplification.Here for filling this gap, a comprehensive study about the gel treatment operation is represented. First, different possible candidate parameters that seem to have an effect on process were selected. Then, using Pareto charts the most effective ones were selected and their corresponding interaction based on their t-values was determined. Afterward, the need to a fast and accurate model was felt, and thus, an artificial neural network was created. Finally, the effect of these parameters on NPV is discussed and using the genetic algorithm, the best combination of parameters for maximizing the profit is represented. All these parts of study confirmed the good performance and applicability of the gel treatment operation.