Natural Ventilation of Urban Offices: a Summary of Findings from the Refresh Project

Many buildings in temperate climates rely on natural ventilation for fresh air, which is also a sustainable, low energy design choice. However, predicting natural ventilation rates for buildings surrounded by other buildings is difficult due to complex flow patterns in the urban canopy. Indoor Environmental Quality (IEQ: air quality, “stuffiness” – humidity and CO₂ concentrations - thermal comfort) also depends on how people interact with the building as well as its location. The Refresh project (Universities of Reading, Leeds, Southampton, Birmingham and Surrey, UK) explored the impact of urban microclimate on building ventilation for optimal performance of occupants. A range of methods was used (measurements, wind-tunnel modelling, CFD, qualitative interviews, controlled EEG tests) across idealised and realistic buildings to understand better how people respond to poor IEQ, and how natural ventilation can help to improve it.

The Refresh Cube Campaign (RCC) at Silsoe, UK was the first field study of natural ventilation for an idealised array of cubical "buildings" under realistic weather conditions and ran for over 9 months. Engineering standards often include design data only for isolated cubical buildings, which motivated the “sheltered building” experiment. It was found that the cross ventilation rate for the array case was reduced by 50-90% compared to the isolated building when the wind was within ± 50° of being perpendicular to the window. Single-sided ventilation models, commonly used for design, underestimated the RCC field data by a factor of 10-20. Levels of turbulence are much higher for the RCC experiments than previous work – results suggest that turbulence enhances the mean ventilation rate when window size is relatively small (1% of wall area). This is a typical situation for buildings in dense urban areas.

Flow patterns around the test building in the array switched between different states for the same background wind direction. This “bistable” behaviour is due to complex interactions between building wakes and can lead to errors in predicted ventilation rates. Unsteady CFD simulations showed that an internal jet appeared for most wind angles for cross ventilation for the isolated building. For the array case, the internal jet was very weak; however, internal mixing seemed to be improved by the influence of unsteady flow outside the window.

Work was done on how people respond to poor IEQ, and how technology could help. Three studies took place: (1) A study on the social determinants of IEQ was done by conducting interviews with office occupants in southern England. Poor perceived IEQ was found not to be a reliable trigger for individual actions to open windows or adjust radiators, due in part to politeness and social norms. Adjustments to windows or heating were made through negotiation with others, especially those who “owned” windows by sitting next to them. (2) To inform these social interactions around IEQ, a second study deployed an ambient IEQ monitor in 11 offices, measuring CO₂, temperature and relative humidity. Due to its focus on only one pollutant (CO₂) and simple “traffic light” display, participants understood the relationship between its readings, air quality, and the need to ventilate the room. The findings suggest that ambient displays can act as a stepping-stone towards improving IEQ literacy and more informed interactions which can improve air quality in offices. (3) A controlled experiment was done to test the effect of “fresh” (low) and “stale” air (high CO₂ concentrations) on cognitive performance. EEG was used to monitor brain state and was analysed to provide an objective measure of sleepiness. Cognitive performance was marginally affected even after short exposures (<40 minutes) to high CO₂ which supports hypotheses from recent literature that poor IEQ can impact performance of office-workers prior to them being aware of it.