THE APPLICATION OF PRINCIPAL COMPONENT ANALYSIS IN PRODUCTION FORECASTING
MetadataShow full item record
Current methods of production forecasting, such as Decline Curve Analysis and Rate Transient Analysis, require years of production data, and their accuracy is affected by the artificial choice of model parameters. Unconventional resources, which usually lack long-term production history and have hard-to-determine model parameters, challenge traditional methods. This paper proposes a new method using principal components Analysis to estimate production with reasonable certainty. PCA is a statistical tool which unveils the hidden patterns of production by reducing high-dimension rate-time data into a linear combination of only a few principal components. This paper establishes a PCA-based predictive model which makes predictions by using information from the first few months’ production data from a well. Its efficacy has been examined with both simulation data and field data. Also, this study shows that the K-means clustering technique can enhance predictive model performance and give a reasonably certain future production range estimate based on historical data.
principal component analysis
Zhou, Yuyang (2017). THE APPLICATION OF PRINCIPAL COMPONENT ANALYSIS IN PRODUCTION FORECASTING. Master's thesis, Texas A & M University. Available electronically from