Search citation statements
Paper Sections
Citation Types
Year Published
Publication Types
Relationship
Authors
Journals
The application of Artificial Intelligence for planning has received increased attention in the energy industry in the past few years, particularly for the increased production efficiency requirements and environmental standards. The objective of this paper is to show the successful integration of production, completion, subsurface and spatial data using machine-learning algorithms to predict production performance for future development wells. The internal Marcellus Business Unit (MBU) well database, populated with data of 500+ historical wells, has been used in this study. Production data, treated as timeseries, has been processed using functional Principal Component Analysis (PCA) to allow removal of outliers and mode detection. Utilizing this data, a suite of machine-learning algorithms has been applied to reconstruct gas production from available and target well data. Uncertainty quantification has been provided for production curves to identify the quality of prediction. During the study, the sensitivity analysis on input variables has been performed iteratively to screen and rank the most important variables for prediction. The workflow, Unconventional Reservoir Assistant (URA), has been implemented in a proprietary cloud-based platform providing the necessary means for data upload, integration, pre-processing, and finally model training and deployment. This allows the user to focus on the evaluation of model output quality, data filter and workspace generation for continuous model testing and improvement. The full well dataset, split into trained and tested data, has been used for prediction as a preliminary guide to where the most prolific areas of development are located. Results were ranked based on production expected by pad and based on normalized performance. The information was then used to compare with type curves and original development order. In parallel, economic evaluation of break-even was performed to rank all future pads. Consequently, integration of the model prediction and breakeven ranking were used to generate the final development order for the MBU. The URA tool has shown preliminary success in predicting production performance for the pilot development area. Multiple case studies have been run achieving blind test results with high accuracy for historical prediction. Results show some dependency of predictor variable ranking on the field development area, providing insight on how subsurface may affect well dynamic behavior. This paper describes how the integration of URA can complement the development workflow for unconventional reservoirs and be used to predict performance based on complex data integration. The methodology used is superior to standard machine learning models providing only production indicators, as it gives the user the possibility to evaluate economics and completion design sensitivity for future well activities. The methodology can be further extended as a proxy model for well schedule optimization in planning or for better insight into well refrac selection.
Several studies have used machine learning-based techniques to improve the production behavior prediction in existing shale gas wells. However, few studies have investigated production prediction in new wells wherein no prior information is available. This is challenging because these predictions are generally based on the analysis of data available on existing wells. Therefore, in this study, data-driven analytics is utilized to analyze the production characteristics of existing wells and improve the predictive performance of the production for new wells. Field data on the Marcellus Shale wells with production histories exceeding 48 months were collected from Enverus’ Drillinginfo. We derived production-related attributes of these wells and identified the key factors using principal component analysis to establish the production dependency on them. Subsequently, the prediction reliability was improved by classifying different production characteristics into groups, and using their trend lines to estimate the cumulative production of the existing wells. For new wells, we developed a model to classify groups based on key factors, and utilized probabilistic values from the classified groups to predict stochastic ranges of cumulative production using an artificial neural network. The field data were normalized with respect to lateral length or the number of stages to enable comparison between multiple wells. Outliers of each input factor were excluded during pre-processing. An analysis of production characteristics was performed by classifying the existing wells into three groups. Results indicated that Group 2 was highly productive, with evident influence of normalized fluid volume during the middle and late phases of production. Further, initial variations in production tendency were observed in Groups 1 and 3 by spacing. Trend lines of classified groups were used to forecast the cumulative production per unit length (NP) of the existing wells. The observed error was less than 10 % in the prediction of NP 48 based on the analysis of NP 6 and NP 12. Additionally, high production variability in shale play is known to induce a rapid reduction in the production trend after the initial production. Therefore, a prediction model with NP of 6, 12, and 48 months was developed. To validate the model, probabilistic values of spacing and decline factors were used in the predictions of NP 6, 12, and 48, yielding an accuracy exceeding 80% and an error of approximately 10%. The proposed multi-well productivity analysis is a trial-and-error process based on data-driven analytics, which can be used to predict shale production in any shale play. In addition, the range of the predicted probabilistic production includes the actual values; therefore, the prediction errors are small compared to those of previous methods for new wells. Consequently, time and resources expended for data acquisition are reduced, and the reliability of productivity forecasts in shale development is improved.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.