Smart Electrical and Thermal Energy Supply for Nearly Zero Energy Buildings

The European Union (EU) intends to reduce the greenhouse gas emissions to 80-95 %
below 1990 levels by 2050. To achieve this goal, the EU focuses on higher energy efficiency mainly within the building sector and a share of renewable energy sources (RES)
of around 30 % in gross final energy consumption by 2030. In this context, the concept
of nearly zero-energy buildings (nZEB) is both an emerging and relevant research area.
Balancing energy consumption with on-site renewable energy production in a cost-effective
manner requires to develop suitable energy management systems (EMS) using demandside
management strategies.
This thesis develops an EMS using certainty equivalent (CE) economic model predictive
control (EMPC) to optimally operate the building energy system with respect to varying
electricity prices. The proposed framework is a comprehensive mixed integer linear
programming model that uses suitable linearised grey box models and purely data-driven
model approaches to describe the system dynamics.
For this purpose, a laboratory prototype is available, which is capable of covering most
building-relevant types of energy, namely thermal and electrical energy. Thermal energy
for space heating, space cooling and domestic hot water is buffered in thermal energy
storage systems. A dual source heat pump provides thermal energy for space heating
and domestic hot water, whereas an underground ice storage covers space cooling. The
environmental energy sources of the heat pump are ice storage or wind infrared sensitive
collectors. The collectors are further used to regenerate the ice storage. Photovoltaic
panels produce electrical energy which can be stored in a battery storage system. The
electrical energy system is capable of selling and buying electricity from the public power
grid. The laboratory test bench interacts with a virtual building model which is integrated
into the building simulation software TRNSYS Simulation Studio.
The EMS prototype is tested and validated on the basis of various simulations and under
close to real-life laboratory conditions. The different test scenarios are generated using
the typical day approach for each season.