Master's Thesis HU Xiaolin


Optimal Design of Decentralized Energy Conversion Units for Smart Microgrids

In this thesis, a mixed integer linear programming (MILP) model is created for optimal design (i.e..
technology selection and sizing, scheduling) of microgrid that meets the electricity and heat demand
of a cluster of residential buildings. The objective of the model is minimizing annual costs
which comprise the investment as well as demand- and operation-related costs. The model of
centralized approach is formulated as a whole large scale MILP. It is reformulated based on the
Dantzing-Wolfe decomposition method into a decentralized approach and solved by column generation
algorithm. The results of decentralized approach are compared with the centralized approach
which is used as a benchmark to assess the performance of decentralized approach. The comparison
is done for 4 scenarios with 3-12 buildings. Then the results of decentralized approach for
concept with microgrid are compared with the results of concept without microgrid to research the
economic and environmental impact of themicrogrid concept.
The results show that the centralized approach provides best coordination with a limited extendability.
The decentralized approach delivers a slightly lower but still good coordination within a
proximate runtime and can be applied to calculate large-scale microgrid. The results also shows
that the microgrid concept provides relatively good economical and environmental performance
for large-scale neighborhood and buildings with high demand.