Abstract

District heating (DH) networks form a central part of the energy infrastructure and ageing
components increasing the need for effective and forward-looking maintenance strategies. The aim of this study is to investigate current inspection and maintenance practices within DH chambers, and to analyse how predictive maintenance can be developed to improve
maintenance planning and extend the asset lifetime.

The study was conducted as a qualitative case study of Vattenfall´s DH network in Uppsala,
combining a literature review with empirical data collected through semi-structured interviews and document analysis. The analysis focuses on inspection routines, data collection and usage, decision-making processes, and which predictive maintenance methods are feasible given current organisational and technical conditions.

The results show that inspections are currently based on fixed intervals and that digital systems are primarily used for documentation rather than analysis. Although relevant data is collected, the potential for predictive maintenance is limited by insufficient system integration, variations in data quality and fragmented historical data. Furthermore, maintenance decisions are largely experience-based and reactive.

A modified Failure Mode and Effects Analysis based method is identified as a suitable first step towards implementing predictive maintenance, as it builds on existing working methods whilst enabling more structured and traceable prioritisation. In the longer term, more advanced methods, such as Discrete Hidden Markov Model, may become relevant. The study indicates a need for a step-by-step development, where both technical solutions and working methods evolve in line with the organisation’s maturity.