Master's Thesis Luise Middelhauve
Robust optimisation of thermal energy systems on city district scale
At the moment, the trend towards a distributed energy system regarding energy generation and trading
takes place. In this work, robust energy unit sizing and distribution within a city district is investigated
economically. Thereby, an existing model is used for mixed linear integer optimisation in Python along with
the solver Gurobi. Therein, various energy units and local networks can be chosen to supply the thermal
and electrical demand.
An approach for robust optimisation is implemented for the uncertainties within the thermal demand and
the economical situation. To reduce runtime the annual demand profile is replaced by characteristic demand
periods. This leads to a separation of the robust approach. The thermal demand is optimised strictly
robust. Further uncertainties such as the efficiency or the gas price is considered separately. A method is
introduced, which enables the handling of the degree of robustness.
A robust optimisation of a real city district verifies the methodology. Furthermore, the solution space is
explored by relaxation of the energy balance. The so found energy distribution and sizing is evaluated economically
and ecologically with the aid of various price- and demand developments.
The model identifies a robust solution, which increases only 1 % in annual costs. Furthermore, a robust
solution is detected, which reduces greenhouse gas emission by 60 % compared to the initial solution.
Thereby, the energy system is 19 % more expensive in annual costs. The economic evaluation shows the
gas price has a bigger impact on thermal energy systems than the electricity price. Within the investigated
interval, the gas price influences the capacity of thermal storages. The electricity price influences the decision
regarding the installation of PV panels.