Occupant Interactions and Effectiveness of Natural Ventilation Strategies in Contemporary New Housing in Scotland, UK

The need to reduce carbon emissions and fuel poverty has led to increased building envelope air tightness, intended to reduce uncontrolled ventilation heat losses. Ventilation strategies in dwellings still allow the use of trickle ventilators in window frames for background ventilation. The extent to which this results in "healthy" Indoor Air Quality (IAQ) in recently constructed dwellings was a concern of regulators in Scotland. This paper describes research to explore this. First a review of literature was conducted, then data on occupant interactions with ventilation provisions (windows, doors, trickle vents) gathered through an interview-based survey of 200 recently constructed dwellings, and measurements made on a sample of 40 of these. The main measured parameter discussed here is CO2 concentration. It was concluded after the literature review that 1000 ppm absolute was a reasonable threshold to use for "adequate" ventilation. The occupant survey found that there was very little occupant interaction with the trickle ventilators e.g., in bedrooms 63% were always closed, 28% always open, and in only 9% of cases occupants intervened to make occasional adjustments. In the measured dwellings average bedroom CO2 levels of 1520 ppm during occupied (night time) hours were observed. Where windows were open the average bedroom CO2 levels were 972 ppm. With windows closed, the combination of "trickle ventilators open plus doors open" gave an average of 1021 ppm. "Trickle ventilators open" gave an average of 1571 ppm. All other combinations gave averages of 1550 to 2000 ppm. Ventilation rates and air change rates were estimated from measured CO2 levels, for all dwellings calculated ventilation rate was less than 8 L/s/p, in 42% of cases calculated air change rate was less than 0.5 ach. It was concluded that trickle ventilation as installed and used is ineffective in meeting desired ventilation rates, evidenced by high CO2 levels reported across the sampled dwellings. Potential implications of the results are discussed.