Application of machine learning models in predicting initial gas production rate from tight gas reservoirs

Abstract

Driven by advancements in technology, tight-gas fi eld development has become a signifi cant source of hydrocarbon to the energy industry. The amount of data generated in the process is immense as most platforms are now being digitized. Machine learning tools can be used to analyse this data in order to build patterns between several dependent and independent variables. Forecasting initial gas production rates has important implications in the planning production/processing facilities for new wells, aff ects investment decisions and is an important component of reporting to regulatory agencies. This study is based on the analysis of reservoir rock/fl uid properties and selected well parameters to build decision-based models that can predict initial gas production rates for tight gas formations. In this study, two machine learning predictive models; Artifi cial Neural Network (ANN) and Generalized Linear Model (GLM), were used to determine the expected recovery rate of planned new wells. Production data was retrieved from 224 wells and used in developing the model. The results obtained from these models were then compared to the actual recorded initial gas production rate from the wells. Results from the analysis carried out revealed a Mean Square Error (MSE) of 1.57 on a GLM model whereas the ANN model gave an MSE of 1.24. Key Performance Index for the ANN model revealed that reservoir thickness had the highest (36.5%) contribution to the initial gas production rate followed by the fl owback rate (29%). The reservoir/fl uid properties contribution to the initial gas production rate was 53% while the hydraulic fracture parameters contribution to the initial gas production rate was 47%.

Modelling oil and gas flow rate through chokes: A critical review of extant models

“…For instance, AI techniques such as ANN used by most researchers in the literature for developing flow rate correlations have been inadequate because the necessary details of the model namely the weights and biases of the network that can be used for reproducing the results of the models were not presented by the researchers. Instances are found in the works of ; Ahmadi et al (2013); Zangl et al (2014); Okon and Appah (2016); Buhulaigah et al (2017); ; Amaechi et al (2019); Marfo and Kporxah (2020); Khamis et al (2020); Bikmukhametov and Jäschke (2020b). However, only a few included these details in their work.…”

“…In forecasting the productivity of the fishbone wells, the neural network model outperformed the fuzzy logic and the radial basis function models Amaechi et al (2019) ANN and Generalized Linear Model (GLM) From the analysis, it was observed that the ANN model performed better than the GLM model with both models having a mean square error of 1.24 and 1.57 respectively. Rashid et al (2019) Artificial Neural Network The proposed model is valid given the how close the errors of the training, testing and validation datasets are to each other.…”

Well Performance Classification and Prediction: Deep Learning and Machine Learning Long Term Regression Experiments on Oil, Gas, and Water Production

“…Furthermore, the performance index of the ANN model revealed that reservoir thickness was responsible for 36.5 percent of the initial gas output, followed by flowback rate (29%). As a result, when it came to estimating gas production and looking back, the ANN model outperformed the GLM model [ 4 ].…”

Comparative Analysis of Historic Predictions and Actual Global Natural Gas Market Trends for the Period From 1997 to 2020

“…Due to European emission restrictions, mainly from coal-fired power plants, many member states have started to replace coal-fired power plants, with cleaner gas-fired power plants (Wynn and Coghe, 2017;Bayer and Aklin, 2020). Projections and predictions are a very important segment of the economy in optimizing production (Ikpeka et al, 2019; al Irsyad et al, 2019), a well-timed response to market demand (Strpić et al, 2017), optimal inventory management and price regulation (Lu et al, 2021). The accurate prediction of energy trends and prices is critical for orientation in the energy market, and it could give direction for policymakers and market participants.…”