Multi-level optimization of maintenance plan for natural gas system exposed to deterioration process

Area of Science:

• Asset Integrity Management
• Risk-Based Maintenance
• Reliability Engineering

Background:

• Effective maintenance scheduling is crucial for asset integrity in industrial facilities.
• Uncertainty from process variables and external perturbations complicates traditional maintenance planning.
• Existing methods often lack a comprehensive approach to integrate failure probability with maintenance costs.

Purpose of the Study:

• To develop a risk-based optimization methodology for maintenance scheduling.
• To integrate process variables (PVs) and failure probability into asset integrity assessment.
• To provide a framework for optimizing maintenance activities considering costs and system condition.

Main Methods:

• Integration of Dynamic Bayesian Networks (DBN), damage modeling, and sensitivity analysis.
• Development of decision configurations and utility nodes within the DBN framework.
• Application of regression analysis to model the impact of perturbations on PVs.

Main Results:

• The methodology successfully clarifies the behavior of failure probability under perturbations.
• Demonstrated application to a chemical plant's natural gas regulating and metering stations.
• Showcased ability to analyze failures using precursor PV data or optimize schedules based on system condition.

Conclusions:

• The proposed methodology offers a robust approach to risk-based maintenance optimization.
• It enables proactive failure analysis and condition-based maintenance scheduling.
• Applicable for enhancing asset integrity in chemical and other process industries.