Katharina Boudier

• Local thermal comfort plays a growing role in terms of occupant satisfaction and in the energy performance of a building. To improve thermal comfort of occupants, decentralized heating and cooling systems are getting more interest in the research community but also in the market. Some studies have also shown that they can reduce the heating and cooling demand of buildings. These systems can be for example, an office chair with heating and cooling function, a thermoelectric heating and cooling wall or even a table fan. Building simulation software is used to optimize the energy performance of buildings during the planning process, but just a few programs enable detailed comfort calculations. Programs which allow the usage of decentralized heating and cooling systems are totally missing. This dissertation presents a newly developed adaptive building controller for combined, central and decentralized systems inside the building simulation software ESP‐r. The building controller adapts the setpoint‐temperatures of the central heating and cooling system and regulates the usage of decentralized systems, based on the thermal sensation and comfort values of a virtual thermal manikin in the considered building zone. This work shall contribute to the application of detailed comfort values within building simulation, which is also necessary to consider decentralized heating and cooling systems like the thermoelectric movable partition or the office chair with heating and cooling function. The first step by the development of the adaptive controller, was the coupling of PhySCo, a "Physiology, Sensation and Comfort" model with the building simulation software Esp-r. The physiology model within PhySCo uses the values of room temperature, mean radiant temperature, air velocity, relative humidity, solar radiation as well as personal parameters such as clothing and activity level. The model calculates skin and core temperatures for 16 individual body parts under consideration of thermophysiological control mechanisms such as sweating, shivering, vasodilatation and vasoconstriction of the blood vessels. These values are used to calculate local and overall sensation and comfort values. Within the adaptive building controller, the local and overall sensation and comfort values are used to control the setpoint‐temperatures of the central heating and cooling system, as well as to regulate the decentralized heating and cooling systems. The adaptive building controller with a wide deadband, with setpoints of 18 to 26 °C was in simulation studies compared with a basic controller with a fixed and narrow setpoint range of 21 to 24 °C. The evaluation focused on the overall comfort values and on a possible reduction of the heating and cooling energy demand of the central system. The results showed, that the adaptive controller could keep the comfort values at the same level as the basic controller and reduced the heating demand at the same time noticeably. The cooling load could also be reduced compared to the basic controller, but the reduction was much smaller compared to the heating load.In the next step, the decentralized heating and cooling systems were added to the adaptive building controller. First the thermoelectric heating and cooling wall, followed by the office chair with heating and cooling function, then both systems were added to the adaptive controller and tested together. The results show clearly that by adding the decentralized heating and cooling systems, the comfort could get further improved and at the same time, the heating and cooling energy demand could get reduced. It was noticeable that during the summer simulation period, comfort was increased, though the increase was rather small. As an additional support, a fan was simulated, which increased the air velocity for the heat‐sensitive head. Thereby, the comfort could be further increased, thus the cooling of the central system was also reduced. In addition, an increase of the upper set point to 30 °C was possible without reducing the comfort level. The adaptive building controller enables a detailed comfort analysis within the building simulation software. It can also be used for the planning of decentralized heating and cooling systems and their effect on thermal comfort.