Development of an Optimal Maintenance Policy for Perkins Electrical Generating Sets Using Markov Chain Predictive Model

Abstract

The study aimed at developing an optimal maintenance policy for 800kVA Perkins electrical generating sets (gen-sets) using Markovian predictive model. The 800kVA Perkins gen-set used by the Michael Okpara University of Agriculture, Umudike (MOUAU) was employed as a case study. To derive the solution for the analysis, the Markovian predictive model utilized sought answer to such questions as, “At what rate should the generator be maintained; and at what condition would it be after maintenance?” Theoretical data derived from the system were compiled, tested and thereafter simulated. Results of the analysis gave a value of 10 weeks and 2 days as the optimal operation policy period for 800kVA gen-sets; meaning that the system should be regularly maintained at the end of this period. In conclusion, appropriate models for enhancing the performance of the MOUAU‟s main gen-set have been developed in this study. In particular, the models are for determining the effective maintenance policy for an electrical gen-set so as to ensure that the system lasts long in service. Hopefully, results from the study will assist maintenance engineers and plant operators in improving the performance of their gen-sets through preventive rather than reactive maintenance. Consequently, it is recommended that the engineers and plant operators of the 800kVA Perkins gen-set at MUOAU and any other institution should adopt the developed models in the plant maintenance practices. It is also recommended that the maintenance policy developed from the study be integrated with the automated generator condition assessment models.