Carbon monoxide levels were monitored at intersections, bus stops, in enclosed parking garages, and in vehicles. Variation in CO levels was then compared with traffic variables. The effect of traffic volume, traffic delay, site location, time of day and meteorological variables were investigated during ambient testing. Incoming and outgoing vehicle volume as well as the effect of the time of day were studied during garage testing. Finally CO variation with vehicle speed, acceleration, road grade and vehicle specific power (VSP), a variable that measures a vehicles engine load per unit mass were investigated during in vehicle tests. The type of vehicle, the surrounding environment and time of day were also considered.
Two studies were performed at two different locations. One study was done in Singapore during the fall of 2009 and one in Cincinnati where tests were done from the winter to the summer of 2010. Similar tests were performed at both locations. Ambient monitoring in Singapore was performed around the NUS (National University of Singapore) campus at bus stops within the campus and around the perimeter of the campus. Ambient testing in Cincinnati was done during winter and spring time at a large intersection. An enclosed parking garage was studied at both locations as well as in vehicle tests. Singapore buses were studied while personal vehicles and city buses were studied in Cincinnati.
Consistent correlations between CO and traffic counts were not seen for the most part at ambient testing sites. A 5 minute interval was used and test periods were typically 1 to 2 hours long. Bus delay at busy bus stops showed consistent positive correlations with CO at the Singapore site. Vehicle delay counted by hand at intersections showed a positive correlation in some cases but was not consistently over each test period. The most consistent pattern around CO concentrations was a peak just after an acceleration period of a traffic cycle (after a green light for an approach with a large queue). Morning tests showed the highest CO levels during the ambient tests. Ambient CO levels ranged from 0 to 1.5 ppm in most cases at both locations. A 3 day test near a major highway showed that CO concentrations during peak periods were elevated when compared to non peak periods.
Better correlations between traffic and CO were observed in the parking garages. High levels of traffic both at the 5 minute interval and across testing periods consistently showed the highest CO levels. CO measurements at the Singapore car park were done just outside the enclosed garage and although low levels of CO were observed the correlation with traffic was the highest of any of the test sites. Good correlations were also seen at the parking garage in Cincinnati although variations in idling time, a wider range of vehicle ages and a more complex entrance may have lead to less consistent correlations than those observed in Singapore. The highest levels were observed in the evening hours as cold start emissions and long idle times led to higher levels of CO.
Elevated levels of CO were observed in all the vehicles tested relative to the ambient environment. CO increased by 0.2 to 0.6 ppm in most cases. A direct correlation with speed, acceleration, road grade and VSP was not observed over a 5 second interval that was tested. Large buses such as a double decker and double long buses in Singapore showed higher levels of CO than normal 12 meter long buses that were taken. The highest levels of CO were seen during cold starts in the morning in personal cars. CO was shown to exponentially decrease with distance driven in the morning which is the same pattern seen by other studies when looking at direct vehicle emissions.