A Novel Workflow for Optimizing Well Placement Under Geomechanical Constraints: A Case Study

Optimum well spacing is an essential element for the economic development of shale gas reservoirs. We present an integrated assisted history matching (AHM) and embedded discrete fracture model (EDFM) workflow for well spacing optimization by considering multiple uncertainty realizations and economic analysis. This workflow is applied in shale gas reservoirs of the Sichuan Basin in China. Firstly, we applied the AHM to calibrate ten matrix and fracture uncertain parameters using a real shale-gas well, including matrix permeability, matrix porosity, three relative permeability parameters, fracture height, fracture half-length, fracture width, fracture conductivity, and fracture water saturation. There are 71 history matching solutions obtained to quantify their posterior distributions. Integrating these uncertainty realizations with five well spacing scenarios, which are 517 ft, 620 ft, 775 ft, 1030 ft, and 1550 ft, we generated 355 cases to perform production simulations using the EDFM method coupled with a reservoir simulator. Then, P10, P50, and P90 values of gas estimated ultimate recovery (EUR) for different well spacing scenarios were determined. Additionally, the degradation of EUR with and without well interference was analyzed. Next, we calculated the NPVs of all simulation cases and trained the K-Nearest Neighbors (KNN) proxy, which describes the relationship between the NPV and all uncertain matrix and fracture parameters and varying well spacing. After that, the KNN proxy was used to maximize the NPV under the current operation cost and natural gas price. Finally, the maximum NPV of 3 million USD with well spacing of 766 ft was determined.

An Integrated Assisted History Matching and Embedded Discrete Fracture Model Workflow for Well Spacing Optimization in Shale Gas Reservoirs

Rui

et al. 2021

Impact of Integration of the Production Systems and Reservoir of a Benchmark Based on Carbonate Fields

Victorino

Filho

Castro

et al. 2022

Integrating production system and reservoir is used in several studies of offshore oilfield development and management for improved production forecasts through more realistic boundary conditions. This study evaluates the influence of the parameters in a production strategy of a reference model (carbonate oil reservoir) on financial and production performance. We first considered nonintegrated system (NI) with the reservoir and fixed boundary conditions for well and gathering system. We then considered integrated system (I) with the variable boundary conditions for the wells and gathering system. Finally, we compared both systems. Our analysis involved several steps to define the best production strategy for both systems based on net present value (NPV) and how integrated modeling helps define production strategy. For NI, three stages were considered: number of wells, placement of wells, and platform capacity. For I, five stages were evaluated (all NI parameters) with diameters and gas lift evaluation, and platform placement. The results are similar, but the simplifications (NI) may affect financial performance. The cross analysis indicated (in the hypothesis that the integrated system is closer to reality) that integrating the NI case resulted in considerable financial and production differences and may be useful in simplified systems. Since the main aspects of the object-function financial return were related to reservoir model behavior, one can first apply the nonintegrated optimization cycles, then add the integrated cycles, obtaining an intermediate time of the integrated model optimization in similar cases.

A Data-Driven Proxy Modeling Approach Adapted to Well Placement Optimization Problem

Kolajoobi

Niri

Amini

et al. 2022

Well placement optimization (WPO) plays an essential role in field management and economy. However, it entails massive computational time and demand since hundreds, even thousands, simulation runs are needed. Different types of proxy models have been utilized to address this issue. Among different Proxy models, Data-Driven proxies are preferred as they can determine the combined effect of several parameters without suffering from the type and the number of modeling parameters. This paper aims to develop a Data-Driven proxy model in an artificial intelligence framework adapted to WPO problem. This proxy estimates and compares the oil recovery for different well configurations. Our contribution is building a dynamic proxy by training a sequence of static proxies in a time-dependent manner to make more benefit from modeling capability of artificial neural networks (ANNs). The workflow comprises preparing a learning database using experimental design techniques, finding the significant parameters by searching the parameter space, training and validating a series of ANNs to obtain the desired field response, and conducting a blind test to ensure the model performance and generality. This proxy is then coupled with genetic algorithm to find an optimal well configuration in a test case. Verifying the results obtained by our proxy with those of a commercial simulator shows that the objectives of constructing this proxy for WPO are successfully achieved.