Modelling Reliability of Supply and Infrastructural Dependency in Energy Distribution Systems

Abstract

This thesis presents methods and models for assessing reliability of supply and infrastructural dependency in energy distribution systems with multiple energy carriers. The three energy carriers of electric power, natural gas and district heating are considered.

Models and methods for assessing reliability of supply in electric power systems are well documented, frequently applied in the industry and continuously being subject to research and improvement. On the contrary, there are comparatively few examples of formal reliability assessment models and methods applied to natural gas and district heating systems. This work aims at contributing to bridge this gap, considering the structural, operational and physical similarities and differences between the systems. A method for evaluating the reliability of supply in natural gas distribution systems is presented, based on state-of-the-art reliability calculations from the electric power domain. Furthermore, a novel modelling approach incorporating pipeline storage in reliability evaluation of high-pressure natural gas pipeline systems is presented.

Parallel energy infrastructures depend on each other at different levels, two of which are addressed in this work. First, by introducing a second energy carrier in an area dominated by electric power, the type of energy end-uses served by the electric power system is affected. An optimisation problem is formulated, finding the optimal allocation of switchgear in an electric power distribution system. It is shown how changes in energy end-uses cause changes in the expected customer interruption costs, which in turn affects the optimisation problem. Second, the dependency of district heating systems on electric power is modelled. Network models for the two systems are coupled, and the consequences of higher-order power system failures are quantified for both systems.

The methods and approaches presented in this thesis are demonstrated by use of simple examples, and applied to test networks and case studies.