Automated Decline Curve Analysis - DCA from a Data Science Perspective

Summary Decline Curve Analysis (DCA) is arguably the simplest method for forecasting production. For wells on steady decline, and particularly for fast and efficient forecasting at scale, its simplicity is also its strength: DCA is robust, fast, fully automated and easy to interpret. Still, simple does not mean easy - to get good performance the least-squares curve fitting routine should be modified. Care must be taken in data preprocessing, model construction and hyperparameter tuning. If done right, DCA offers great promise in automating and streamlining forecasting for many wells in the NCS.