Environmental Engineering Reference
In-Depth Information
After analysis of the initial results it was decided to reduce the horizontal
distance between sampling points to just 1 m to improve maneuverability and safety
while measuring. A new set of ten trips in both directions were made on the 3rd and
4th of April 2008 under similar meteorological conditions. Part of the trajectory is
shown in Fig.
1
. The total trajectory is exactly 4 km long and was cycled in about
12 min with an average speed of 19.6 km/h (SD 1.4 km/h). Measurements were
compared by subjecting average particle numbers per trajectory for each device to
a Wilcoxon matched pairs signed ranks test.
Results
The ratio and 95% CI between measurements taken during the pilot test by both
instruments were 1.00 (0.94-1.05) for case a, 0.98 (0.86-1.09) for case b and 1.01
(0.94-1.09) for case c respectively.
We concluded that the set-up was reliable and reproducible although a small
correction was sometimes needed to off-set systematic differences between instru-
ments. There is no effect of either sampling height or sampling direction on the
results. The results of the measurements will therefore adequately reflect the con-
centration in the breathing zone of the cyclist and his/her external exposure to UFP.
The facts that results are independent of sampling direction confirm the results in
[4]
. Here it is also demonstrated that results are not influenced by cycling speeds
below 21 km/h.
Comparison of UFP with PM10 and PM2.5 Measurements
Results from one of the preliminary tests shown in Fig.
3
clearly indicate that peaks
of PM10 and UFP often occur at different sites on a test trajectory. Analysis of the
video footage revealed that PM10 peaks were often related to construction activities
while UFP was more closely related to traffic (although UFP peaks attributed to
construction equipment also can be seen at construction site 1).
Exposure of UFP to Cyclists on a Cycling Track
The pilot test revealed an unexpected, but important difference over a horizontal
distance of 2 m. Measurements taken at the side of the road were on average 10%
higher than those taken simultaneously at the other side of the bicycle (Ratio 1.096;
stdev 0.044). Average particle numbers during the 3 days of the pilot test were
around 30,000 cm
−3
. Non-parametric tests revealed a difference that was barely
significant (Critical value of 0 for p = 0.05 at N = 6 in the Wilcoxon Matched pair
signed ranks test) and only after correction for an unusually large systematic difference
