Abstract:In the process of developing tight oil and gas reservoirs, multistage fractured horizontal wells (NFHWs) can greatly increase the production rate, and the optimal design of its fracturing parameters is also an important means to further increase the production rate. Accurate production prediction is essential for the formulation of effective development strategies and development plans before and during project execution. In this study, a novel workflow incorporating machine learning (ML) and particle swarm optimization algorithms (PSO) is proposed to predict the production rate of multi-stage fractured horizontal wells in tight reservoirs and optimize the fracturing parameters. The researchers conducted 10,000 numerical simulation experiments to build a complete training and validation dataset, based on which five machine learning production prediction models were developed. As input variables for yield prediction, eight key factors affecting yield were selected. The results of the study show that among the five models, the random forest (RF) model best establishes the mapping relationship between feature variables and yield. After verifying the validity of the Random Forest-based yield prediction model, the researchers combined it with the particle swarm optimization algorithm to determine the optimal combination of fracturing parameters under the condition of maximizing the net present value. A hybrid model, called ML-PSO, is proposed to overcome the limitations of current production forecasting studies, which are difficult to maximize economic returns and optimize the fracturing scheme based on operator preferences (e.g., target NPV). The designed workflow can not only accurately and efficiently predict the production of multi-stage fractured horizontal wells in real-time, but also be used as a parameter selection tool to optimize the fracture design. This study promotes data-driven decision-making for oil and gas development, and its tight reservoir production forecasts provide the basis for accurate forecasting models for the oil and gas industry.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to improve productivity by optimizing the fracturing parameters of multi - stage fractured horizontal wells (MFHWs) during the development of tight oil and gas reservoirs, and on this basis, maximize economic benefits. Specifically, the paper proposes a new method that combines machine learning (ML) and particle swarm optimization algorithm (PSO) to predict the production of multi - stage fractured horizontal wells in tight oil reservoirs and optimize fracturing parameters. This method aims to overcome the limitations in existing production prediction studies, such as the difficulty in maximizing economic returns and optimizing fracturing schemes according to operator preferences (such as target net present value). ### Background and Problems of the Paper - **Background**: In recent years, unconventional resources have received increasing attention due to the continuous growth of fossil fuel demand. The successful development of multi - stage fractured horizontal wells has unlocked a large amount of unconventional oil and gas resources worldwide. - **Problems**: - **Production Prediction**: Accurately predicting the production performance of multi - stage fractured horizontal wells is crucial for production optimization. - **Fracturing Parameter Optimization**: Before the development of unconventional reservoirs, optimizing fracturing parameters is extremely important. - **Limitations of Existing Methods**: Although traditional numerical simulation methods can achieve relatively accurate production prediction, they usually require a large amount of computing time and it is difficult to comprehensively consider complex reservoir mechanisms and influencing factors. ### Solutions - **Methods**: The paper proposes a workflow that combines machine learning and particle swarm optimization algorithm. The specific steps are as follows: 1. **Data Preparation**: Select characteristic parameters that have a significant impact on production. 2. **Numerical Simulation Construction**: Establish a representative numerical model of multi - stage fractured horizontal wells. 3. **Data Generation**: Conduct numerical simulations to generate samples for training production prediction models. 4. **Production Prediction Model Selection**: Develop multiple machine learning models, train them and evaluate the effectiveness and robustness of these models. 5. **Case Verification**: Determine the optimal machine learning model as a proxy tool for predicting tight oil production. 6. **Predict Production**: Verify the selected machine learning model through actual cases, and predict tight oil production and net present value (NPV). 7. **Fracturing Parameter Optimization**: Use the particle swarm optimization algorithm to optimize the selected fracturing parameters. ### Key Technologies - **Machine Learning Models**: Including linear regression (LR), support vector regression (SVR), gradient - boosted decision trees (GBDT), random forest (RF) and multi - layer perceptron (MLP). - **Particle Swarm Optimization Algorithm (PSO)**: Used to optimize fracturing parameters to maximize net present value (NPV). ### Results - **Model Performance**: Among the five machine learning models, the random forest (RF) model performs the best, with \( R^2 \) values of 0.994 and 0.963 on the training set and the test set respectively. - **Optimization Results**: By combining the random forest model and the particle swarm optimization algorithm, the optimal combination of fracturing parameters under the condition of maximizing net present value has been successfully determined. ### Conclusions - **Contributions**: This research proposes a complete machine - learning - based workflow that can efficiently and accurately predict the production of multi - stage fractured horizontal wells and optimize fracturing parameters, thereby improving the economic benefits of unconventional reservoir development. - **Future Work**: Further explore the interpretability and reliability of the model, as well as potential problems in practical engineering applications. Through this method, the paper not only improves the accuracy and efficiency of production prediction, but also provides data - driven decision - support for the development of unconventional oil and gas resources.

Machine learning-based fracturing parameter optimization for horizontal wells in Panke field shale oil

Shale oil production prediction and fracturing optimization based on machine learning

Fracturing Productivity Prediction Model and Optimization of the Operation Parameters of Shale Gas Well Based on Machine Learning

A data-driven workflow for prediction of fracturing parameters with machine learning

A Novel Shale Gas Production Prediction Model Based on Machine Learning and Its Application in Optimization of Multistage Fractured Horizontal Wells

Optimization of Fine-Fracture Distribution Patterns for Multi-Stage and Multi-Cluster Fractured Horizontal Wells in Tight Gas Reservoirs

Automatic Optimization of Multi-Well Multi-Stage Fracturing Treatments Combining Geomechanical Simulation, Reservoir Simulation and Intelligent Algorithm

Optimization of tight gas reservoir fracturing parameters via gradient boosting regression modeling

Parameters Optimization of Fractures with One for Injection and Two Others for Production of Horizontal Wells Fracturing on Offshore Oilfields

Fracture Spacing Optimization Method for Multi-Stage Fractured Horizontal Wells in Shale Oil Reservoir Based on Dynamic Production Data Analysis

Research on Production Prediction Method of Multi-stage Fractured Shale Gas Horizontal Well

Multi-Task Learning Network-Based Prediction of Hydraulic Fracturing Effects in Horizontal Wells Within the Ordos Yanchang Formation Tight Reservoir

Effect Evaluation of Staged Fracturing and Productivity Prediction of Horizontal Wells in Tight Reservoirs

Optimization Method of Fracture Parameters in Multistage Fracturing of Horizontal Wells

Application of Machine Learning to Predict the Capacity of Fractured Horizontal Wells in Shale Reservoirs

Ensemble-based optimization of hydraulically fractured horizontal well placement in shale gas reservoir through Hough transform parameterization

Efficient Optimization: Unveiling the Application of Ensemble Learning Combined with the CMA-ES Algorithm in Hydraulic Fracturing Design

A Productivity Prediction Model for Multistage Fractured Horizontal Wells in Shale Gas Reservoirs

Data-driven model for hydraulic fracturing design optimization: focus on building digital database and production forecast

Optimization of Staged-Fracturing in Heterogeneous Tight Gas Reservoirs of Western-Sichuan Gas Field

Using Machine Learning Method to Optimize Well Stimulation Design in Heterogeneous Naturally Fractured Tight Reservoirs