Reliability Assessment of Gas Pressure Reduction Stations Based on Monte Carlo Markov Chain and Continuous-Time Markov Chain Methods: A Case Study of a City Gas Pressure Reduction Station

Ali  Karimi; Esmaeil  Zarei; Rajabali  Hokmabadi

doi:10.18502/jhsw.v15i4.21453

Ali Karimi Department of Occupational Health, School of Public Health, Tehran University of Medical Sciences, Tehran, Iran
Esmaeil Zarei Department of Safety Science, College of Aviation, Embry-Riddle Aeronautical University, Prescott, AZ 86301, USA.
Rajabali Hokmabadi Department of Occupational Health Engineering, School of Public Health, North Khorasan University of Medical Sciences, Bojnurd, Iran

DOI: https://doi.org/10.18502/jhsw.v15i4.21453

Keywords: Reliability; Continuous Time Markov Chain (CTMC) Method; Markov Chain Monte Carlo (MCMC) Method; National Gas Pressure Reducing Station

Abstract

Introduction: A gas pressure reduction station is an important facility in gas transmission systems. Thesesystems consist of various sections, the reliability of each section affecting the station’s overall reliability.Therefore, this study aimed to assess the reliability of station sections using Markov chain Monte Carlo(MCMC) and the continuous-time Markov chain (CTMC) method.

Material and Methods: Equipment failure and repair rates were simulated using the MCMC method inWinBUGS14. Then, based on the failure and repair rates, the station reliability was evaluated using theCTMC. The results of the equipment failure rate simulation were validated using two criteria: MC Errorand the Goleman-Rubin test. Also, the results of station reliability evaluation were validated using RealityCheck and Partial Benchmark Exercise methods.

Results: Failures in the filtration and pressure reduction sections were more frequent than in othersections of the station. Therefore, these sections were considered the most critical sections in thereliability assessment. The posterior standard error was less than 0.01, indicating good convergence of thedata for the parameter posterior distribution. The results of the Goleman-Rabin test showed values lessthan 1.2, indicating proper convergence of the chains. For all sections and stations, a systematic approachwas determined using the Markov model. The results showed a strong correlation between the CTMC andthe block diagram method (R2=0.9499).

Conclusion: The proposed approach combines failures of system components and can display multiplefailures. It also accounts for time factors in its calculations and minimizes subjective expert evaluations.