Machine learning application in the oil and gas industry is rapidly becoming popular and in recent years has been applied in the optimization of production for various reservoirs. The objective of this paper is to evaluate the efficacy of advanced machine learning algorithms in reservoir production optimization. A 3-D geological model was constructed based on permeability calculated using a machine learning technique which involved different architectures of algorithms tested using a 5-fold cross-validation to decide the best machine learning algorithm. Sensitivity analysis and a subsequent history matching were conducted using a machine learning workflow. The aquifer properties, permeability heterogeneity in different directions and relative permeability were the control variables assessed. Field development scenarios were exploited with the objective to optimize cumulative oil recovery. The impact of using a normal depletion plan to a secondary recovery plan using waterflooding was investigated. Different injection well placement locations, well patterns as well as the possibility of converting existing oil producing wells to water injection wells were exploited. Considering the outcome of an economic analysis, the optimum development strategy was realized as an outcome for the optimization process. Prior to forecasting cumulative oil production using artificial neural network (ANN) for the optimization process on the generated surrogate model, a sensitivity analysis was performed where the well location, injection rates and bottomhole pressure of both the producer and injector wells were specified as control variables. The water cut as part of the optimization process was utilized as a secondary constraint. Forecasting was performed for a 15-year period. The history-matching results from the constructed geological model showed that the oil rate, water rate, bottom hole pressure, and average reservoir pressure were matched within a 10% deviation from the observed data. In this study, the ANN optimizer was found to provide the best results for the field cumulative oil production. Using a secondary recovery development plan was observed to significantly increase the cumulative oil production. A machine learning based proxy model was built for the prediction of cumulative oil production to reduce computational time. In this study, we propose an approach applied to reservoir production optimization utilizing a machine learning workflow. This was accomplished by utilizing a surrogate model which was calibrated with a number of training simulations and then optimized using advanced machine learning algorithms. A detailed economic analysis was also conducted showing the impact of a variety of field development strategies.

The objective of this paper is to present a framework that applies machine learning to reservoir characterization. Machine learning applications in the oil and gas industry is rapidly becoming popular and in recent years has been utilized for the characterization of various reservoirs. Conventional reservoir characterization employs core data measurements and local correlations between porosity and permeability as input data for reservoir property modeling. However, a strong correlation between porosity and permeability as well as reliable core measurements are not always available. The proposed approach uses both well logs and core data to construct different models to predict permeability using three distinct methods including a parametric, non-parametric, and machine learning technique. The parametric method employed the known relationship between porosity and the natural log of permeability. The non-parametric regression method utilized the alternating conditional expectation (ACE) algorithm. The third approach involved machine learning workflow implemented within a commercial software. The reservoir was first classified into distinct hydraulic flow units using the flow zone indicator (FZI) approach and k-means clustering. Permeability was then predicted using a supervised machine-learning framework. A field case study was then utilized to ascertain the effectiveness of these approaches by validating the model with data from one of the wells. The results of these three approaches were compared using the mean absolute error (MAE) and mean squared error (MSE) values in the validation process. An examination of the error calculated found the support vector machine (SVM) and linear regression algorithms in characterizing the upper reservoir region and the SVM for the lower reservoir characterization yielding the best results when using the machine learning approach thus, yielding the least error as compared to the other two approaches. Additional validation was performed by comparing different models based on permeability fields through numerical model calibration to historical data. It was found that machine learning-based permeability had the least error compared to calibration data prior to the history matching process. The investigated reservoir consists of two distinct productive oil zones separated by an impermeable shale. There are 15 existing wells that have been producing from both the upper and lower zones since 1997. Using machine learning permeability-based model, the history matching process was conducted successfully to match both observed production data and pressure data of 15 wells with less than 10% global deviation. This study presents the feasibility of applying several different approaches in predicting permeability based on gamma ray, bulk density, and deep resistivity logs. The machine learning approach proves its high potential and readiness in supporting reservoir characterization and history matching compared to the other approaches.