Abstract
Corrosion is a major safety issue that can lead to unexpected failures in refinery plants. Ensuring the safety of refinery processes from corrosion requires regular maintenance procedures such as the inspection and replacement of pipes, which are essential tasks. Performing maintenance tasks more frequently reduces the failure cost of the process by enhancing the process safety and reliability, but this leads to increased costs of inspection and replacement. Therefore, there is an optimal point that minimizes the total maintenance cost (including inspection, replacement, and failure) while satisfying the minimum safety level limit.
The objective of this study is to determine the optimal planning by changing and adjusting maintenance variables such as the initial pipe wall thickness, the number of inspections, and the inspection time. The first step in achieving this goal is the development of a probabilistic model that calculates the total maintenance cost as a function of the remaining pipe wall thickness with the first order reliability method. The remaining wall thickness is decreased by corrosion over time, which affects the total cost. The next step is to minimize the total maintenance cost of the model. A mixed integer nonlinear programming (MINLP) algorithm was employed to perform the optimization. The optimal inspection time, number of inspections, and the initial wall thickness were determined for various corrosion rates. A periodic inspection case that had the same time interval between inspections was also studied.
