Analysis of Ventilation Strategies for ZEB
MetadataShow full item record
Powerhouse Kjørbo is an office building renovated to become a zero emission and energy positive building. It has a displacement ventilation system, which aims at providing a good indoor climate with low energy consumption. Previous measurements indicate that the ventilation effectiveness is lower than expected. There might be several causes, like stagnant zones, obstructions to the air flow or short-circuiting between air supply and extract. This thesis aims to investigate these issues. A reduced-scale building model was built in a laboratory to resemble the office landscape in the prototype (Powerhouse Kjørbo). Tracer gas measurements, velocity mapping, and temperature measurements were conducted to examine the ventilation efficiency, gain understanding of the air flows and validate the model against the prototype in order to achieve similarity. Fieldwork was conducted in the prototype building. Tracer gas measurements, velocity mapping, duct traversing, smoke visualization and registration of presence were performed to examine the ventilation efficiency, gain understanding of the air flows and determine flow rates and average occupancy. The tracer gas measurements in the prototype show that the air change efficiency is lower than expected for displacement ventilation. This indicate existence of short-circuiting or stagnant zones. The local air change indexes indicate displacement characteristics in the zones, but mixing of the stratified air layers seem to occur and the bookshelves are assumed to work as obstructions to the air flow. Comparison of local indexes suggest that short-circuiting and stagnant zones also occur in areas of the building apart from the office landscape. A minor air leakage within the heat exchanger is still present. The tracer gas measurements in the model show that the air change efficiency is similar to the prototype, but the local air change indexes differ significantly. Excessive air leakage and suboptimal temperature differences are the most likely causes. Improvements made to the air-tightness were confirmed by the tracer gas measurements. Similar to the prototype, the bookshelves in the model seem to pose as obstructions to the air flows. The temperature measurements in the model reveal a temperature difference which is relatively stable, but insufficient to match the Archimedes number to the prototype. The cooling capacity of the air handling unit, air leakage, and uninsulated ducts are identified as the main causes. The velocity mapping and smoke visualization in the prototype indicate that the diffuser discharge is unstable, most likely caused by the large diffuser area. The issue affect the air distribution and may cause short circuiting. The adjacent zone were determined to be one meter, and no occupants or objects are found within this zone. The velocity mapping in the model reveal that the diffuser discharge flow towards different parts of the office landscape than in the prototype. This is assumed to affect the similarity. Comparison of the results conclude that the model and the prototype do not share satisfying similarity yet. The local air change indexes differs too much, the air flow patterns are not similar enough and the air leakage in the model is unacceptably high. However, there are indications that the air flows in certain zones are behaving similarly. It is believed that the model and prototype can share similarity after improvements have been made to the air leakage, temperature difference and control of the simulated air coming from the part of the prototype excluded from the modeling.