Comparison of three residential ventilation systems in practical operation

Abstract

Three mechanical ventilation systems which are widely used in Swiss apartment buildings were investigated experimentally. The first system is the Central Bidirectional ventilation (CBi), where each dwelling is equipped with an AHU, supply air in bedrooms and extract air in wet rooms. In the second one, the Centralized Unidirectional ventilation (CUn), each dwelling is equipped with a unidirectional ventilation unit, extract air in wet rooms and outdoor air transfer devices in bedrooms. In third system bedrooms and living rooms are equipped with non-ducted Decentralized Ventilation Units (DVU), additional the wet rooms are served by independent extract air fans. In two related projects, the three ventilation systems were investigated in a total of 33 flats in 17 apartment buildings, with a total of 98 supply air rooms and 86 extract air rooms. Multiple products were represented in all systems. The cleanliness of the ventilation systems was assessed visually and the frequency of filter changes was queried. With few exceptions, the hygienic condition was good to satisfactory. In 80% of the CBi systems and DVU's, filters were changed at least once a year. In the CUn systems, this was done only half as often. The air flows were measured in all rooms as found. Then the filters were changed and the ventilation components were cleaned. Finally, all airflows were measured again. The values in the as-found and clean condition were compared with the requirements of the Swiss standards and the imbalance was determined. Overall, the CBi systems proved to work quite stable and robust. With the CBi systems, the supply air flow rate was 6% lower in the as-found condition than after cleaning. For the other two systems, significant reductions in supply air flow rates due to clogged filters and devices were observed (reduction in CUn 17%, in DVU 30%) as well as significant imbalances (as-found CUn 62%, DVU 42% / clean CUn 24%, DVU 21%). One conclusion is that CUn and DVU systems would require significantly more filter changes and cleaning to ensure the same reliable operation conditions as the CBi system. For the system CBi, the specific electric power input was determined and compared with measurements in projects carried out 9 and 12 years earlier. In the sample of the new project, the SPI was about one third lower than in the older projects. This effect is mainly interpreted as product improvements.