To meet increased needs of transports in the Stockholm region and reduce the problems with traffic congestion in central parts, a 21 km long by-pass (18 km in a tunnel) is planned. The bypass is expected to reduce traffic and emissions in central Stockholm, but at the same time tunnel users could be exposed to high concentrations of air pollutants from traffic. Thus to control the air quality in the tunnel system, air pollution guideline values have been proposed. The current study is initiated and funded by the Swedish Transport Administration (Trafikverket), and the aim is to assess the potential health impacts of applying different NOX guideline values (1000, 2000, 3000 and 4000 μg/m3 as hourly average max values all-over the tunnel system). The passengers’ exposure was estimated based on annual average NOX exposures, time spent in the tunnel and the number of tunnel users. Health impacts were assessed following health impact assessment principles using equations and WHO’s software AirQPlus.

With minimal ventilation and maximal traffic amounts during rush hours the NOX hourly average concentrations could raise up to 3500 μg/m3 and even when the planned maximum ventilation would be in use, the maximum concentration would stay as high as 1789 μg/m3. Thus, it is in principle with planned the technology impossible to meet the lowest proposed guideline value of 1000 μg/m3 in the whole tunnel system. However, the effects would be with this guideline still the smallest, resulting annually in 22.2 (CI 95% 16.8–30.1) more premature deaths and 480.4 (95% CI 364.1–650.6) years of life lost (assuming travellers to come from the age group 30–74). If the guideline value would be 2000 μg/m3, the exposure would annually in the same age group cause 35.2 (CI 95% 26.7–47.6) premature deaths with 760.9 (480.4–650.6) years of life lost. With the lowest guideline level, passing the whole tunnel during rush hours on working days would increase mortality risk by 7.4% (95% CI 5.5-10.1), on average corresponding to a life expectancy decrease by 0.27 (95% CI 0.20-0.37) years for people aged 30–74 years.

Among different tunnel links, the biggest exposure is expected in link 5N, where 28.6-37.2% (depending on limit value scenario) of the total health impact could be generated. The link 3N has high NOX concentration, large number of passengers and long exposure time (time spent in the tunnel link). Even the NOX concentrations are expected to be highest in links 411 and 314, the exposure time there would be shorter and the number of exposed passengers smaller. For the separate links the differences in exposure between limit value scenarios could also vary largely: while the difference was big for link 5N, it was rather small for link 7N.

If we compare these results with the previously estimated beneficial effect on the health of the local population due to decrease of urban air pollution exposure (expecting annually 23.7 (95% CI 17.7–32.3) fewer premature deaths), only with most favourable assumptions as less older persons using tunnel and with highest ventilation the tunnel could have smaller negative health effects compared to the alternative current open road E4. In all other cases the health effects in the by-pass tunnel Förbifart Stockholm are expected to be higher. Also the exposure levels in the tunnel are expected to be somewhat higher compared to previous analysis due to more 5 enhanced dispersion modelling for the tunnel, including also ramps in the impact assessment and predicting higher numbers of cars than previously.

To reduce problems with traffic congestion and meet increased needs of transports, a 21 km long by-pass (18 km in a tunnel) is planned. The by-pass is expected to reduce traffic emissions in central Stockholm but at the same time tunnel users could be exposed to high concentrations of air pollutants from traffic.

For the reduction in urban air pollution concentrations, the change in annual ambient NO_X and PM₁₀ levels were modelled using 100x100 m grids and the population (1 628 528 inhabitants) average exposure was calculated for Greater Stockholm area. The tunnel exposure was estimated based on annual average NO_X and PM₁₀ levels, time spent in tunnel and number of persons using the tunnel. Health risks were calculated based on health impact assessment principles using equations and the WHO AirQ software. In these calculations the E-R coefficient for non-external mortality was 8% per 10 μgm^-3 increase of NO_X (vehicle exhaust indicator) and for daily number of deaths 1.68% per 10 μgm^-3 increase of non-exhaust (road dust) PM₁₀.

It appeared that for the general population there would be annually 23.7 (95% CI 17.7–32.3) premature deaths less; mainly from lower exposure to vehicle exhaust (indicated by NO_X) and somewhat from a reduction in coarse particles (indicated by PM₁₀), contributing 23.2 and 0.5 fewer deaths, respectively. Other adverse health effects of exposure are also expected to be reduced. At the same time, tunnel users will be exposed to vehicle exhaust components in terms of NO_X up to near 2000 μgm^-3 during rush-hours. Passing the whole tunnel twice on working days would correspond to an additional annual NO_X exposure of 9.6 μgm^-3. Assuming there would be on average approximately 55 000 vehicles per day each way and 1.3 persons in each vehicle from the range 30–74 years of age, this exposure would result in 20.6 (95% CI 14.1–25.6) more premature deaths. If there would be more persons per vehicle or older and more vulnerable people travelling, the adverse effect of exposure in traffic could become larger. Hence, the effects in reality may be different as these results are based on now presented scenarios.