New Statistical Approach to Study How an Electric Submersible Pump Population Improves Mean Time Between Failure by 42% in an Ecuador Field

Abstract In any electric submersible pump (ESP) population, key performance indicators (KPIs) such as mean time between failure (MTBF), index failure, or run life (RL)/target run life (TRL) are used to keep track the performance of an ESP population. However, what should the stakeholders do in case these metrics decay? In this paper, we introduce a new systematic approach to accurately use population statistics to identify the causes affecting the RL performance so actions to solve them can be taken. The overall approach is to perform a deep analysis of the metrics and detect the causes affecting the RL in a population. A step-by-step statistical analysis aims to segregate the studied population into quadrants. The quadrant range is dependent on the desired TRL of the project or key run life (KRL). Four quadrants are defined, and each quadrant is studied independently using box-and-whiskers plots to learn from them to set accurate actions. Equipment can fail due to the same general reasons; however, this method will study the different root causes and will be digested using different kinds of graphs. This methodology has been used for two consecutive years, showing good results in terms of MTBF on the studied field by increasing it 42%.