Javascript är avstängt eller blockerat i din webbläsare. Detta kan leda till att vissa delar av vår webbplats inte fungerar som de ska. Sätt på javascript för optimal funktionalitet och utseende.

Webbläsaren som du använder stöds inte av denna webbplats. Alla versioner av Internet Explorer stöds inte längre, av oss eller Microsoft (läs mer här: *

Var god och använd en modern webbläsare för att ta del av denna webbplats, som t.ex. nyaste versioner av Edge, Chrome, Firefox eller Safari osv.

Spot the difference! : On the way towards automated fault handling in district heating buildings


Summary, in English

This thesis focuses on how district heating utilities may work with fault handling in customer installations as a tool to reduce their return temperature levels. Reduced return temperatures would make it possible to improve the efficiency of the district heating systems and decrease their environmental impact. Faults in the customer installations cause high return temperatures that increase the return temperature of the entire district heating system. Thus, to reduce the system return temperatures, such faults must be detected, diagnosed, and corrected. The focus of the thesis has been to understand and identify the specific challenges that arise when integrating data analysis of district heating customer data as a natural part of the utilities’ fault handling processes and find (parts of) the solutions to these challenges. The reason for this is that faults in the customer installation manifest themselves in district heating customer data. Thus, it is possible to analyse this data to detect faults.

The results show that a multitude of different faults may occur in a customer installation. However, some faults are more common than others, and these faults should be prioritized when developing fault detection methods using data analysis. There is also a need to develop a mutual way to name faults that occur in the installation. Today, the district heating utilities use different words when talking about the same faults, making it hard to compare results from different district heating systems. It also makes it hard to create labelled data sets, which are needed to develop and validate successful fault detection tools based on data analysis. However, two different data analysis methods for fault detection have been developed without access to labelled data. Both methods show that it is possible to utilize customer data analysis to detect faults in customer installations.

To realize the full potential of the fault detection tools, the utilities need to utilize them on a larger scale within their organization. Important organizational aspects include identifying a clear stakeholder in the fault handling process and calculating the economic value of eliminating faults. There is also a need to change the current fault handling processes to better align with automated fault detection tools. Therefore, an important result in this thesis is a suggestion for how a fault handling process based on data analysis may be designed.

District heating utilities may use the results of this thesis to improve their fault handling processes of customer installations, thereby reducing their return temperatures and improving the possibilities of district heating being part of the future energy systems.









Energy Sciences, Lund University


  • Energy Engineering


  • Customer installations
  • Customer data
  • Data analysis
  • District heating
  • Experience from industry
  • Fault detection
  • Fault handling
  • Fault labelling
  • Return temperatures
  • Substation performance
  • Taxonomy





20 augusti 2021




Lecture hall KC:A, Kemicentrum, Naturvetarvägen 14, Faculty of Engineering LTH, Lund University, Lund. Zoom:


  • Svend Svendsen (Prof.)