Master's Thesis Stefan Klask


Determining an optimized distribution of thermal energy plants on district level using an optimization algorithm

The use ofmodern thermal power plants with high primary energy utilization factors, coupled with
the compound of buildings with decentralized networks, can replace the existing heat supply by
boiler and reduce the generation of greenhouse gas emissions. Due to the variety of different plant
technologies and interconnections heuristic planning of the distribution entails a risk of considering
suboptimal solutions. This work shows the implementation of a mixed integer programming
problem to determine an optimized distribution of power plants on city district level. The objective
function minimizes either the total costs or greenhouse gas emissions by preserving the thermal
and electrical energy balances. The model includes thermal energy generators (boiler, CHP plant,
heat pump), photovoltaic systems, storage systems (hot water tank and battery) and decentralized
grid structures on thermal and electrical level. The consideration of long-term scenarios uses large
amounts of input data. The thesis describes the application of demand day and demand hour tools,
which allow minimizing the amount of data by clustering similar time steps. To install pipelines of
local heating network in already existing areas, the consideration of the street network is recommended
as a laying route. The usage of a graph with nodes and edges can implement the geographical
conditions to determine an short and realistic link between buildings. The last chapter shows the
virtually creation of a real city district that includes 13 buildings and a proposed solution of an optimized
distribution. Depending on the modernization of the buildings the result shows potential
cost savings up to 17%, while reducing greenhouse gas emissions by about 33%.