Fault detection in district heating substations

Abstract: This thesis gives a summary of different faults that occur in district heating substations and suggests ways of detecting them. The aim of the research has been to achieve fault detection using only the heat meter and the sensors connected to it, namely two temperature sensors and one flow meter. The incentive for fault detecting district heating substations comes from an effort to keep them in proper working order so as to be able to keep a low return temperature in the network and thus be able to lower the supply temperature. Hypothetically, if the supply temperature in Sweden were lowered by only one degree, the annual savings would be about SEK 50 million. The thesis also entails brief accounts of district heating and heat metering along with two models of heat exchangers. The first paper in this thesis presents a way of fault detecting the temperature sensors by detecting the noise amplitude. The second paper describes how to separate the energy usage for heating and domestic hot water production using only one heat meter. Energy separation is very interesting, since separate measurement of domestic hot water energy usage and heating usage is recommended by the Swedish National Board of Housing, Building and Planning. The third paper describes how to estimate hot water consumption using the hot water energy consumption and measurement of cold water and hot water temperature. It also contains an error analysis of the estimate. The error analysis suggests that the estimate using a class 1 heat meter is as accurate as the measurement using a class-3 flow meter. Thus, the two estimation methods presented can be used instead of installing extra heat meters or flow meters. This makes it possible to implement the separate measurement(s) according to the recommended guidelines at a lower cost.