Degradation modeling and condition-based maintenance of boiler heat exchangers using gamma processes. (March 2019)

• Record Type: Journal Article
• Authors: Cholette, Michael E.; Yu, Hongyang; Borghesani, Pietro; Ma, Lin; Kent, Geoff
• Is Part Of: Reliability engineering & system safety. Volume 183(2019)
• Page Start: 184
• Page End: 196
• Publication Date: 2019-03
• Subjects: Boiler tube erosion modelling -- Condition-based maintenance -- Gamma process; Reliability (Engineering) -- Periodicals; System safety -- Periodicals; Industrial safety -- Periodicals; Fiabilité -- Périodiques; Sécurité des systèmes -- Périodiques; Sécurité du travail -- Périodiques; 620.00452
• DOI: 10.1016/j.ress.2018.11.023
• Languages: English
• ISSNs: 0951-8320

Abstract

Highlights: A new degradation model for boiler heat exchanger erosion is developed. The new model is a combination of a physics-based erosion model and a Gamma Process. The new model is used to predict the total cost of maintenance for planning exchanger renewals. A condition-based maintenance method is developed and demonstrated on a real boiler. Abstract: Boiler heat exchanger tubes erode over time, leading to costly leaks and capacity loss as damaged tubes are taken out of operation. These losses can be recovered by replacing the heat exchanger, albeit at significant capital cost. Deciding when to replace the heat exchanger is greatly aided by prediction of the future capacity and downtime losses associated with tube thickness degradation. However, current thickness loss models study a single phenomenon and therefore provide oversimplified point predictions of thickness loss. Additionally, the available models do not account for uncertainty in the degradation process and therefore cannot provide information on future tube failure probabilities required for decision making. In this paper a novel degradation model is presented using a physical erosion model and a Gamma process to account for the uncertainties in the thickness degradation process. Subsequently, new algorithms are presented to predict the probability of key events in boiler operation: the bursting of a tube during operation and the plugging of a tube preventively during a scheduled shutdown. Finally, a condition-based maintenance methodology is presented based on the new degradation model and applied to a real-world case study of a boiler operating in a real Australian sugar production facility.