A BAYESIAN STATISTICAL FRAMEWORK FOR RELIABILITY, AND MAINTENANCE COST MODELING OF NATURAL GAS COMPRESSOR

Predicting failure occurrences is vital for ensuring operational efficiency and minimizing downtime in production systems. This study characterizes the reliability parameters of a High-Pressure Compressor (HPC-2) using a Weibull Bayesian framework and estimates its annual maintenance cost. Operational data including shutdown and start times were obtained from a major crude oil and gas company in the Niger Delta, from which relevant secondary data were extracted. The developed Bayesian model under the Weibull distribution estimated an expected failure rate of 0.008749 failures per hour (standard error, SE = 3.74×10⁻⁵) with a 95% two-tail Bayesian prediction interval of [0.00866, 0.00893]. At 36 hours of operation, reliability was 72.98% (SE = 9.91×10⁻⁴; 95% prediction interval, PI [0.725, 0.732]). The non-informative prior produced a mean time to failure (MTTF) of 105.89 hours (95% PI [68.42, 161.20]), while the Gamma prior estimated 150.95 hours. The annual maintenance cost was estimated at 41,954.65 USD.