An artificial intelligence based model for implementation in the petroleum storage industry to optimize maintenance

Sporadic equipment breakdowns and unplanned downtime due to the predominant use of Reactive Maintenance and Preventive Maintenance at Company X necessitate the enhancement of the maintenance management system. This paper presents an Artificial Intelligence based model for optimizing the conventional maintenance strategies currently employed. Critical equipment at the fuel depot was identified through the Nowlan and Heap risk analysis matrix procedure. The critical equipment identified was pumps, storage tanks, valves and the standby power supply system. Ishikawa diagrams and FMECA analysis were then used in optimizing the Preventive Maintenance strategy and developing the Intelligent Maintenance model for each critical equipment. The focus of the AI Maintenance model was on pumps, as pumps were identified to be the most critical equipment. An Expert System was developed, tested and run for the pumps. The pump diagnosis application developed was programmed using Jess, a rule based system that accepts input from the operators.