Application of Machine Learning for Oil Production Forecasting
One of the central questions in science is forecasting: based on the past history, how well can we predict the future? In many domains with complex multivariate correlation structures and nonlinear dynamics, forecasting is highly challenging. In the oil and gas industry, conventional approaches such as the modified hyperbolic method, have been utilized to analyze the production decline curve. Forecasting decline curves is an important component for E&P companies in business planning, asset evaluation, and decision making. Here we introduce a machine learning approach to tackle the problem, and to be more specific, an LSTM approach (LSTM stands for Long Short Term Memory, which is one kind of recurrent neural network). Compared with the hyperbolic approach, where the problem has been reduced to an over-simplified curve and essentially determined by a global curvature structure, the LSTM model is more dynamic and has a better chance of capturing non-linear events. In time series prediction, one main difficulty is how to stabilize the solution, as the error can easily accumulate over time. One way to make the algorithm more robust is through feature engineering, and here we leverage historical data from other wells, which improves our prediction significantly. We also build the prediction model from the accumulated curve domain, and eventually ensemble multiple models to reduce the variance. Given the fact that the model is only trained on the first 3 months of data (around 10% of the data), the oil rate prediction for the first 2 years shows great promise.
Cheng Zhan is a Senior Data Scientist at Anadarko Petroleum, where he works on field development optimization and long-term production forecasting. He focuses on building machine learning algorithms to create strategic and financial impact for the company. Prior to his current role, he worked as a Geophysicist at TGS and CGG, utilizing seismic data and inversion methods to help operators make better decisions in exploration. He holds a PhD in mathematics from University of Houston, and a B.S. in Mathematics from Sun Yat-sen University.