Developing Predictive Oil Well Diagnostics Based on Intelligent Algorithms

Abstract

The current competitive market condition in the oil industry is so concentrated on companies' budgets that they require methods to extract oil from wells at the lowest possible cost. All pumping units, new or old, require regular preventive maintenance and constant inspection, and with 30 percent of total oil production coming from sucker rod pumps, the cost of diagnostics must be reduced accordingly.This article focuses on the intelligent diagnostics of the rod and borehole pump for preventive maintenance and monitoring during the life cycle of the well. The use of artificial intelligence methods for predictive diagnostics of equipment condition solves problems without production stoppages and without additional interventions from outside.