PREVISION AND OPTIMISATION OF REPAIRABLE SPARE PARTS: A CASE STUDY IN THE PETROLEUM INDUSTRY

Danilo Gonçalves Calvacante,Denis Borenstein,Luciano Ferreira

doi:10.7166/31-2-2221

Abstract

This study was motivated by the high costs incurred by an energy company for repairable spare parts for faulty mission-critical items, particularly with those that operate until failure. The objective of this paper is to develop and apply a method for repairable spare part inventory management of run-to-failure equipment. To achieve a robust method that incorporates the data collected from previous failures, such as environmental factors and operating conditions, we propose an optimisation approach based on an accelerated failure time model. Accelerated failure time is used as a reliability regression model with covariates to describe different operational conditions. An algorithm is also developed to consider the repairable nature of the equipment, predicting the number of spare parts based on the expected number of failures in the period and the equipment repair cycle. The proposed method is applied using data from three different power units of electrical submersible pumps, a mission-critical item in oil production. The results show an average reduction of 60.6 per cent in the required number of spare parts, considering an average fill rate of 95.33 per cent. This reduction implies an estimated annual savings of around US$664,720 in inventory costs, considering the analysed units.

Highlights

Equipment maintenance accounts for a considerable portion of the operating costs in a number of industrial sectors, and may represent 15–70 per cent of the total production costs [1]
To achieve a robust method that incorporates the data collected from previous failures, such as environmental factors and operating conditions, we propose an optimisation approach based on an accelerated failure time model
The electrical submersible pumps (ESP) system is used as an artificial lift method in oil production, where the centrifugal pump functions as an operating element of the fluid mixture produced by the reservoir

Summary

INTRODUCTION

Equipment maintenance accounts for a considerable portion of the operating costs in a number of industrial sectors, and may represent 15–70 per cent of the total production costs [1]. The papers using covariates address the problem of non-repairable equipment/components; and the analysed items can be maintained using preventive or condition-based maintenance As a consequence, it is unclear whether they can be directly applied to our problem, since the ESPs could demand different techniques and methods for analysing failure data. To the best of our knowledge, this is the first study to apply accelerated failure time (AFT) as a reliability regression model with covariates (RRMC), instead of using the proportional hazards model (PHM) [20], to optimise the number of repairable spare parts in a real-world application. The Kolmogorov-Smirnov (KS) adherence test is used to select the statistical distribution With both the RRMC and a standard probability distribution to represent the failure data defined, the effects of the covariates on the reliability of each component can be determined and the systemic reliability of the equipment measured.

Stage 3

Stage 2

Findings

CONCLUSION