Prioritisation of faults in district heating substations: Towards predictive maintenance and optimised operation

Effectively detecting and handling faults in district heating substations is vital to ensure the security of heatsupply and improve system efficiency. This is a challenging task due to the growing number of substations andlimited monitoring and service personnel. Digitalisation on the demand side offers an opportunity to developdata-driven methods for automatic fault detection, enabling utilities to optimise maintenance interventionsacross multiple customers. A variety of different faults can occur in substations, which can reflect differentlyon operational data. It is then necessary to prioritise faults to address the most relevant ones in developingadequate detection methods and supporting operators in their Operation and Maintenance (O&M) processes.Failure Modes and Effects Analysis (FMEA) is a widely used methodology to prioritise potential failures,but it misses aspects relevant to O&M. In this study, we propose an adaptation of the original FMEA forthe prioritisation of faults with focus on O&M optimisation. The methodology uses a Maintenance PriorityNumber (MPN) for the ranking of faults based on severity, occurrence, monitoring potential and maintenancecapability of the fault. Severe and frequent faults, which have a potential to be monitored and maintainedyield the highest MPNs and should be in focus from an O&M perspective. Using the proposed methodologythe most relevant faults for predictive maintenance in substations in Germany have been identified. These arethe contamination of strainers, pump failures and fouling of heat exchangers. These faults should be in focuswhen developing automatic fault detection and diagnosis methods.