PERFORMANCE ANALYSIS AND MAINTENANCE PLANNING FOR THE REAL OPERATING SYSTEMS

Abstract

India is in the race of global competition, is adopting rapid change in technology and increasing complexity of the operating industrial systems. Due to complex nature of modern and complex machines it become challenging to maintenance staff to maintain failure free operating system for long duration specifically for process industry (which runs 24×7 hrs). Therefore, to achieve high operating efficiency it is necessary that the system should be highly reliable or minimum downtime. Thus a cost effective maintenance strategy will be useful to analyze system behaviour. In recent time, with the increase in complexities of the real operating systems the job of the system analyst/maintenance manager has become more challenging as they have to study, characterize, measure and analyze the performance/behaviour of the systems. The main aim of this research work is to analyze the performance of real operating systems and plan a suitable maintenance policy for them. A comprehensive review of literature was conducted to identify the gaps and relevant research issues in this area. In the past, several classic methods are reported in literature to perform the performance analysis and maintenance planning for the various operating systems of different process industries. But, most of them are not able to consider the uncertainty and imprecision/vagueness in the collected data and the results so obtained using this data may also be uncertain. Hence, to overcome this problem a fuzzy methodology (FM) based integrated framework (based on qualitative and quantitative approaches) has been developed and is applied for the performance analysis and maintenance planning of various real operating systems of the medium size coal fired thermal power plant located in northern part of India. In the qualitative analysis, firstly, the Root Cause Analysis (RCA) of the considered systems have been done where the various failure causes (causing systems unavailability) are listed under each unit of a particular subsystem. Then to list all the potential failure modes, their causes and effect on the system performance, Failure Mode Effect Analysis (FMEA) is carried out where Risk Priority Numbers (RPN) for the various failure causes are computed based upon probability of occurrence of failure (Of), severity of failure (S) and probability of non-detection (Od) values. Further, the limitation of FMEA in risk ranking is overcomed by using Fuzzy Decision Making System (FDMS) and Grey Relation Analysis (GRA) and the ranking results are compared for accurate decision making of critical components. v In the quantitative analysis, first, the PN model of the real operating system is developed (to represents its series/parallel combination) from its equivalent fault tree model and the crisp failure and repair time data is collected from the various sources such as plant maintenance log book, expert's opinion and reliability data base. In order to overcome the uncertainty and imprecision/vagueness in the collected data, it is converted into Triangular Fuzzy Number (TFN). Using these TFN as input for the different components of the PN model various reliability parameters namely failure rate, repair time, availability, unavailability, reliability, unreliability, Expected Number Of Failure (ENOF) and Mean Time Between Failure (MTBF) of the considered systems are computed at different spreads and the failure dynamics of the considered systems are studied in crisp, fuzzified and defuzzified terms. After analyzing the behaviour of operating systems in qualitative and quantitative terms, the management is more concerned about the reliable operation of the critical units of the considered plant. Therefore, a combined fuzzy-AHP and fuzzy TOPSIS approach based integrated framework has been developed and is implemented for deciding the optimal maintenance strategy for the Steam Generating and Power Generating system of the considered thermal power plant. The maintenance decision making for the considered systems has been done considering the qualitatative and quantitative aspects of the system. In the proposed framework, the weights computed under fuzzy AHP approach are included in the fuzzy TOPSIS approach for the selection of the optimal maintenance strategy for the considered systems. Further, for measuring the robustness of the selected maintenance strategy sensitivity analysis has been done corresponding to sub-criteria weights. Lastly, to take care of reliability, availability and maintenance aspects of the system collectively, the resource allocation for the considered units of the thermal power plant has been done using Dynamic Programming approach. The problem is treated as Multi Stage Decision Making (MSDM) problem where the manpower and maintenance cost decision are optimized in order to allocate the resources to various interrelated subsystems of the considered thermal power plant.