Optimization of Hydraulic Fracturing Design using ANN- A Case Study

Meeting targeted post-hydraulic fracturing production in the industry remains challenging despite optimization efforts. The optimization process is often based on trial-and-error, past practices, and analogous approaches, which are costly, time-consuming, and high-risk. Therefore, it is of practical interest for engineers to develop a fast, reliable tool that extracts knowledge from historical data and use it to analyze and predict well performance and support future process design decisions. Artificial intelligence (AI) is becoming increasingly important in the industry for providing such tools, and this study utilizes artificial neural networks (ANNs) to develop a novel approach for an Omani condensate tight gas field to achieve these objectives. This study adopted a top-down modeling approach, where ANNs were developed, trained, validated, and tested for a tight gas field. Three ANNs were developed, each serving a different purpose. The First ANN is used to predict some controllable hydraulic fracturing (HF) design parameters using different well and reservoir properties in addition to frac data. The Second ANN is used to predict several frac data outputs through inverse-looking ANN using well and reservoir properties and actual frac design data to help optimize primary HF treatment design and to characterize the subject reservoir. As an add-on to the second ANN, a third ANN was developed to predict the expected productivity index resulting from an HF treatment. All developed ANNs were also used to identify the most influencing parameters affecting the hydraulic fracture job using ANN connection weight analysis. The results show that the developed ANNs estimate the problem's unknowns within an error margin of less than 10%. The ANNs complement the industry's standard methods for HF jobs, which involve design (First ANN), execution (Second ANN), and job evaluation (Third ANN). The developed networks were converted into a graphical user interface (GUI) to facilitate practical usage. The GUI can serve as a screening system to provide preliminary HF design parameters before the simulation process. Additionally, it can also be utilized as a standardized tool to validate any proposed design resulting from current practices used by the industry. This study is novel because it is the first AI data-driven model designed for a local gas field to predict several HF design parameters using a top-down modeling approach. The selection of input and output data differed from that of other studies, and the developed model enables individual stage performance analysis rather than the average performance of all stages combined. This study demonstrates the potential of AI and top-down modeling approaches for addressing challenges specific to the Omani fields to improve future design decisions, leading to significant benefits in terms of cost reduction, greenhouse gas (GHG) emissions reduction, production increase, and time efficiency in hydraulic fracturing operations.