Content uploaded by Margaret Bell
Author content
All content in this area was uploaded by Margaret Bell
Content may be subject to copyright.
400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760 Web: www.sae.org
2005-01-1620
Quantifying the Effects of Traffic Calming on
Emissions Using On-road Measurements
Basil Daham, Gordon E. Andrews, Hu Li and Mark Partridge
Energy & Resources Research Institute, University of Leeds
Margaret C. Bell and James Tate
Institute of Transport Studies, University of Leeds
Reprinted From: General Emissions 2005
(SP-1944)
2005 SAE World Congress
Detroit, Michigan
April 11-14, 2005
SAE TECHNICAL
PAPER SERIES
The Engineering Meetings Board has approved this paper for publication. It has successfully completed
SAE’s peer review process under the supervision of the session organizer. This process requires a
minimum of three (3) reviews by industry experts.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.
For permission and licensing requests contact:
SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: permissions@sae.org
Tel: 724-772-4028
Fax: 724-772-4891
For multiple print copies contact:
SAE Customer Service
Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-1615
Email: CustomerService@sae.org
ISSN 0148-7191
Copyright © 2005 SAE International
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.
Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract to Secretary, Engineering Meetings Board, SAE.
Printed in USA
ABSTRACT
The objective of this work was to determine the effect of
one form of traffic calming on emissions. Traffic calming
is aimed at reducing average vehicle speeds, especially
in residential neighborhoods, often using physical road
obstructions such as speed bumps, but it also results in
a higher number of acceleration/deceleration events
which in turn yield higher emissions. Testing was
undertaken by driving a warmed-up Euro-1 spark ignition
passenger car over a set of speed bumps on a level
road, and then comparing the emissions output to a non-
calmed level road negotiated smoothly at a similar
average speed. For the emissions measurements, a
novel method was utilized, whereby the vehicle was
fitted with a portable Fourier Transform Infrared (FTIR)
spectrometer, capable of measuring up to 51 different
components in real-time on the road. The results
showed that increases in emissions were much greater
than was previously reported by other researchers using
different techniques. When traffic-calmed results were
compared to a smooth non-calmed road, there were
substantial increases in CO2 (90%), CO (117%), NOx
(195%) and THC (148%). These results form the basis
for a good argument against traffic calming using speed
bumps, especially for aggressive drivers. Slowing traffic
down with speed restrictions enforced by speed
cameras is a more environmentally friendly option.
INTRODUCTION
In the UK, over 3000 people die every year in traffic
accidents, 25% of them pedestrians [1], mainly due to
excessive speed in congested urban roads. Due to
public safety fears regarding the levels of traffic currently
present on our inner city and town road systems, several
measures have been put in place over the years that
have tried to allay these fears. This was done either by
decreasing the volume of traffic on the roads (e.g.
congestion charging, city of London) or reducing the
speed of the traffic (using speed bumps or speed
cameras) or by diverting heavy vehicles away from the
small roads which are commonly used as short cuts by
haulers. Not only are the fears to do with pedestrian
safety (especially children) but they are also about the
environment. A measure that has been widely used as a
general ‘all-purpose’ solution in the UK is traffic calming.
This method is strongly supported by the public and the
evidence of this can be seen throughout the country with
speed bumps, roundabouts, bottlenecks and speed
cushions now commonplace within cities, towns and
villages. Traffic calming helps drivers make their speed
appropriate to local conditions through measures that
are self-enforcing.
The UK schemes undertaken to produce traffic calming
are covered under the Traffic Calming Act 1992, which
amended the Highways Act 1980 by the addition of
Sections 90G, 90H and 90I which allows works to be
carried out ‘…for purposes of promoting safety and
preserving or improving the environment…’. These
regulations were again further amended by the Roads
(Traffic Calming) Regulations 1993, which came into
effecting in August of that year and were introduced to
allow local highway authorities the power to construct
particular measures for traffic calming purposes which
are not otherwise clearly authorized. Hence the increase
in the amount of traffic calming devices present on our
roads today.
There are problems faced with traffic calming; some
schemes such as speed cameras or bottlenecks can
work out expensive, while speed bumps are low cost
and easily installed. Within villages where there is very
little alternative road network available, reduction in
traffic volume will be negligible and the effects these
traffic calming measures have on the environment is
often not considered both in terms of air quality and
noise.
Taking all this into account, the methodology behind the
selection of the correct traffic calming measure is of
paramount importance and is unique for each stretch of
road. The planning, consultation and execution of the
correct measure must be done accordingly,
2005-01-1620
Quantifying the Effects of Traffic Calming on Emissions
Using On-road Measurements
Basil Daham, Gordon E. Andrews, Hu Li and Mark Partridge
Energy & Resources Research Institute, University of Leeds
Margaret C. Bell and James Tate
Institute of Transport Studies, University of Leeds
Copyright © 2005 SAE International
remembering that any scheme undertaken must be for
the long term. A good traffic calming scheme will blend
well into the environment, and will continue to operate
with little fuss or concern [2].
Traffic calming has now revolutionized thinking with
regards to town and city planning. The ability of a certain
scheme to reduce speeds at any part of a road network,
while in some circumstances improving capacity, has
been exploited globally. Measures that have been
undertaken are shown in table 1, along with their
proposed effect and some problems that have been
encountered [3].
Table 1: Road traffic calming measures [3]
Device Max comfortable
speed (mph)
Associated
problems
Road-top
hump
21-25 Original, cheap and
still effective tool on
urban roads, but
rough and noisy
Speed table 15-25 Slope of ramp
determines control
speed
Speed
cushion
20-30 Effect dependant on
exact size
Speed limit
sign
No direct control Reliant on drivers
compliance
Pinch point Controlled by
opposing flow
Dependent on
opposing flow
(priority writing not
usually
recommended)
Chicane Varies hugely Control based on
forced level of
lateral curvature of
vehicle paths
Road
narrowing
Any May reduce speed
slightly, but may
have a large effect
when it becomes a
pinch point
Mini
roundabout
21-25 Entirely dependent
upon geometry and
turning flows
Speed
camera
Any Expensive
The particular traffic calming device that was
investigated in this paper was the speed cushion. The
pollution problems associated with the braking and
accelerating of a passenger vehicle in order for it to deal
with the device correctly were investigated. The method
behind the size and shape of these cushions is down to
the size of the vehicle that dominates the traffic on the
roads. Hard sprung vehicles such as busses,
ambulances and fire engines are more affected by the
vertical deflection than cars or small van. Therefore the
width of the cushion is such that it is able to differentiate
between the types of vehicle, so vehicles such as the
ambulance are held up less and buses are unaffected if
the driver aligns the bus correctly. These vehicles will
feel some lurching, however not as much as a small car.
The outer edges of the speed table are rounded so the
bus or emergency vehicle does not suffer as much as a
small car that has to pass over a steeper incline,
therefore feeling more deflection, resulting in the need
for slower speeds [3]. It is recommended that the
gradient on and off the cushion should not be more than
1:8 due to the grounding of smaller vehicles on the
speed table and for the same reason the height should
not exceed 75mm. The length of the cushion should be
between 1.7 and 2.5 meters to avoid discomfort while a
width of 1.9 meters offers greater effectiveness for
slowing a vehicle down [4].
The cushions are situated in the center of the car’s path,
with no gap between them that may allow drivers to
avoid them. In order to cause less damage or
inconvenience to the driver, they are required to line the
car up correctly which in itself means that the speed of
the car must be reduced. It is the effect of this slowing
down process and the acceleration away from the speed
cushion on the levels of emissions produced by a
passenger vehicle that was investigated in this study. In
the road investigation carried out, there were seven
bumps per kilometer so the spacing between the speed
bumps was on average 140 meters, which is higher than
is usually encountered on the roads.
PREVIOUS WORK
As far as the authors are aware, on-road real-world
emissions data quantifying the effect of speed bumps
has not been published so far. Since this study is the
first of its kind, there was no literature to compare its
results to. Nevertheless, in some respects a traffic
calming investigation is similar to studies concerned with
driving behavior. This is because aggressive drivers
tend to be on and off the throttle more often and more
aggressively compared to normal drivers. A normal or
calm driver tends to be smoother, therefore producing a
smooth speed-time profile similar to a non-traffic calmed
road. The aggressive driver has a speed-time profile
similar to a traffic calmed road since acceleration and
braking events will be more frequent. Consequently,
parallels can be drawn between driver behavior studies
and traffic calming studies.
De Vliger’s work [5] investigated driver behavior and
found that aggressive driving produced a dramatic
increase in CO and THC emissions, but less so for NOx.
CO emissions were up to three times higher for
aggressive drivers, while HC and NOx were up to two
times higher. Fuel consumption was generally 30-40%
higher for aggressive urban driving compared to rural
and motorway traffic. Average trip speeds remained
almost the same.
In a similar study performed by Rapone [6] comparing
congested and free flowing traffic conditions, HC was
found to be 12 times higher, NOx was 5 times higher and
CO was 4 times higher. This test used a small-engined,
instrumented car to obtain on-road data which was then
reproduced on a chassis dynamometer for emissions
analysis.
The most comprehensive and authoritative study carried
out on the impact of traffic calming measures was set up
by the Charging and Local Transport Division of the
DETR. It commissioned a three-year study on the
impacts of traffic calming measures on exhaust
emissions from passenger vehicles. The study was
carried out by the TRL and included in it was an analysis
of nine types of traffic calming measures using many
types of vehicles [3]. It was the first study of its kind and
the results are important in assessing the impact of
traffic calming measures on the environment and the
local community. It was a wide reaching study that took
nine different measures into account, assessing the
emissions produced, speed, safety and delays caused to
emergency vehicles. The test procedure involved using
a LIDAR (Light Detection and Ranging) system to
produce speed-time profiles for the vehicles passing
through each of the schemes. Afterwards, the impacts
on the emissions were determined using the driving
cycles and a chassis dynamometer with constant
volume sampling. The pollutants measured were CO,
CO2, HC, NOx and particulates. The results, which are
summarized in table 5 for two types of vehicles, clearly
show that the calming measures increase the emissions
of the pollutants. Catalyst cars were shown to be most
sensitive to traffic calming methods, although they
tended to have the lowest absolute emissions rates
compared to the diesel and non-catalyst vehicles which
were also studied.
The results found in the TRL report were compared to
an average speed model (MEET) [4]. While the MEET
model tended to underestimate CO and overestimate
NOx and CO2, it was found that the %change in going
from a non-calmed road to a calmed road was very
similar for the TRL and MEET data for all the pollutants.
EXPERIMENTAL PROCEDURE
A EURO1 vehicle was used for this study as they still
constitute a fair proportion of the UK vehicle fleet and
hence are still major contributors to air pollution in cities.
It takes about 16 years for 90% of vehicles sold in any
one year to be no longer in use [7] and this period is
becoming longer for modern vehicles. Thus the work on
EURO1 vehicles has significance in terms of their
current use in city driving and hence their impact on air
quality. It will be at least 2013 before 90% of EURO1
vehicles are an insignificant proportion of city traffic.
Future work will investigate EURO2, EURO3 and
EURO4 vehicles.
The device used for measuring on-road emissions in this
investigation was a novel system built around a Temet
FTIR. This system is described in detail by Daham et al.
[8]. It uses a compact FTIR installed in the boot of the
car along with a fuel flow measuring device in order to
calculate the total emissions on a g/km basis. The
repeatability of the instrument is more than adequate for
making comparisons between different drive cycles. In
previous work, the FTIR was validated against other
measurements systems and shown to be within 7% for
steady state and within 20% for transient cycles in terms
of the accuracy of drive cycle mass emissions.
Three baseline runs were initially performed while trying
to be as smooth as possible on the throttle in order to
maintain a constant 30mph (~50km/h), which was the
speed limit of the road under investigation. The results
from these three runs were averaged in order to obtain
the emissions for a non-calmed level road with a 30mph
speed limit.
After the baseline 30mph runs were completed, the car
was driven over the speed bumps with appropriate
braking and accelerations events. Even though the
speed cushions were designed to permit an average car
to pass over them at the required speed of 20-30mph,
for this study the car was slowed down to 10mph and
then accelerated back to 20-30mph in 2nd gear. This was
done in order to simulate an 80mm round-top road hump
which is one of the worst types of speed bumps. Speed
cushions allow a vehicle to pass over them at 30mph if
the car is positioned correctly, whereas with road-top
speed humps, the car must be brought to a very low
speed in order to avoid discomfort to the passengers
and damage to the vehicle. The action of many drivers
at speed bumps is to slow down before the bump and
accelerate off the bump as simulated in the present
work. The average of the three traffic-calmed runs was
obtained and compared to the baseline result at 30mph.
The drive cycle was simply a round trip along the traffic
calmed road in non-rush hour traffic. The road used for
testing contained seven speed cushions in total. After
the seven speed cushions were passed, the car was
turned around in a side road for a return trip. Therefore
each 2.2km run contained a total of fourteen speed
cushions in addition to the turnaround point where speed
was almost zero. The average distance between speed
bumps was 140 meters.
The distances for the calmed and non-calmed runs are
identical since the same road was used. The only
difference being that for the traffic calmed runs, the car
was slowed to about 10mph and reaccelerated in 2nd
gear, thus mimicking the normal action of a driver over a
speed bump.
RESULTS
Smooth road results
Figure 1 plots the speed-time profile as well as throttle
position for three runs over the non-traffic calmed level
road. It can be seen that the first two runs were
consistent, but the third run was affected by other traffic
at the beginning of the run. The section in the middle of
the graph where speed is zero is where the car is turning
around, in a side road, to go back to the starting point.
Figure 2 shows how the emissions varied during these
three smooth runs. It is noticed that there was variability
in the emissions, but for the most part it was within
acceptable upper and lower limits. It can also be seen
from figure 2 that the first run was the smoothest run
with the least overall emissions. A brief numerical
analysis of the three runs is given in table 2, with the
European EURO1 regulations being listed for
comparison. The total CO2 and average speed of the
third run confirmed that there was something different
compared to the first two runs. In spite of this, the third
run was included in the average since it didn’t seem to
deviate excessively in terms of emissions. A relationship
between CO2 and average speed can be seen in the
three runs with higher CO2 being emitted for lower
average speeds.
Run 1 was chosen to be presented in the subsequent
analysis since it is the most consistent run with the least
speed and throttle position variations. Figure 4 shows
how the mass based emissions vary with throttle
position, rate of throttle position change and road speed.
The emissions are constant for the most part except for
three peaks; one at the beginning, one in the middle and
one at the end. These peaks respectively correspond to
getting on to the main road, turning around, and getting
off the main road. CO seems to be most affected by
these three transients compared to the other pollutants.
This can be clearly seen in figure 3 where the gradient of
the CO plot changes drastically when the car gets on to
the test road and when it turns around at around the
150-second mark.
Table 2: Smooth runs statistics
Run 1 Run 2 Run 3 Avg. Euro1
Time (s) 315 331 356 334 784
Fuel (kg) 0.172 0.182 0.181 0.178 n/a
Dist. (km) 2.238 2.291 2.245 2.258 11
Cat. temp
pre test
(°C)
305 293 320 306 (Cold)
Cat. temp
post test
(°C)
405 401 393 400 n/a
Avg. Speed
(km/h)
41.16 40.10 36.52 39.26 18.7
(urban)
33.6
(overall)
g CO2/km 294 307 358 320 n/a
g CO/km 1.72 3.45 2.30 2.49 2.72
g NOx/km 1.07 1.33 1.23 1.21 0.42*
g THC/km 0.12 0.14 0.16 0.14 0.55*
*EURO1 specifies a total NOx+THC of 0.97g/km, but the EURO3
HC/NOx ratio is used for the sake of comparison with experimental
data
0 50 100 150 200 250 300 350
0
10
20
30
40
Tim e (s)
Run 1
0
10
20
30
0
10
20
30
40
Speed (km/h)
Run2
0
10
20
30
Throttle position (%)
0
10
20
30
40
Speed
T h ro ttle position
Run 3
0
10
20
30
Figure 1: Speed and throttle position for the three smooth runs
0 50 100 150 200 250 300 350
0.00
0.05
0.10
0.15
0.20
CO
NOx
THC
Time (s)
Run 1
0
2
4
6
8
10
12
14
0.00
0.05
0.10
0.15
0.20
CO, THC, NOx (g/s)
Run 2
0
2
4
6
8
10
12
14
CO2 (g/s)
0.00
0.05
0.10
0.15
0.20 Run 3
0
2
4
6
8
10
12
14
CO2
Figure 2: Emissions comparison for the three smooth runs
0 50 100 150 200 250 300 350
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
CO
NOx
THC
Time (s)
Total CO, NOx, THC (g)
0
100
200
300
400
500
600
700
CO2
Total CO2 (g)
Figure 3: Cumulative emissions plot of smooth run 1
0 50 100 150 200 250 300 350
0.000
0.002
0.004
0.006
0.00
0.02
0.04
0.06
0.0
0.1
0.2
0
5
10
0
2
4
6
12
14
16
18
-20
0
20
0
10
20
30
0
20
40
Time (s)
THC (g/s)
NOx (g/s)
d(TP)/dt
Throttle Position (% )
Speed (km/h)
Air-fuel ratio
CO (g/s)
CO2 (g/s)
Fuel flow (kg/hr)
stoich.
Figure 4: Analysis of smooth run 1 (d(TP)/dt is rate of change of throttle position)
Traffic-calmed road results
Figure 5 plots the speed-time profile in addition to the
throttle position for three runs performed on the traffic-
calmed road. Coincidently, the first run was again the
cleanest of the three. Run 2 had too long an idle time
when turning around. This was due to traffic briefly
blocking the exit to the main road. During run 3 there
was a complete stop around the 136-second mark due
to other traffic. The numerical analyses of the three runs
are listed in table 3 and it can be seen that the
repeatability is very good in spite of the slightly different
traffic conditions. The average values of the three runs
will be compared to the average values of the smooth
runs in table 4.
A graphical representation of the emissions for the three
traffic-calmed runs is shown in figure 6. Looking at run 1
and neglecting the first two CO2 peaks (which
correspond to getting on to main road) it can be seen
that there are seven distinct CO2 peaks before turning
around. These correspond to the seven speed bumps in
the drive cycle. And of course there are another seven
peaks for the return journey. All the pollutants show
these distinct peaks, with the exception of THC since the
scaling isn’t conducive to its much smaller magnitude. It
should be noted that the y-axis scales of figure 6 are
approximately 1.5 times the scales of the non-calmed
road in figure 2.
Table 3: Speed bump runs statistics
Run 1 Run 2 Run 3 Avg.
Time (s) 343 369 349 354
Fuel (kg) 0.237 0.239 0.248 0.241
Dist. (km) 2.217 2.232 2.224 2.224
Cat. temp pre
test (°C)
276 240 259 258
Cat. temp
post test (°C)
475 450 454 460
Avg. Speed
(km/h)
37.44 35.11 36.91 36.45
g CO2/km 586 633 600 607
g CO/km 6.08 5.52 4.63 5.41
g NOx/km 3.20 3.75 3.76 3.57
g THC/km 0.35 0.37 0.31 0.34
0 50 100 150 200 250 300 350
0
10
20
30
40
50
60
Time (s)
Run 1
0
20
40
60
0
10
20
30
40
50
60
Speed
Throttle position
Speed (km/h)
Run 2
0
20
40
60
Throttle position (%)
0
10
20
30
40
50
60
Run 3
0
20
40
60
Figure 5: Speed and throttle position for the three speed bump runs
0 50 100 150 200 250 300 350
0.00
0.05
0.10
0.15
0.20
0.25
0.30
CO
NOx
THC
Tim e (s)
Run 1
0
5
10
15
20
25
0.00
0.05
0.10
0.15
0.20
0.25
0.30
CO, NOx, THC (g/s)
Run 2
0
5
10
15
20
25
CO2 (g/s)
0.00
0.05
0.10
0.15
0.20
0.25
0.30 Run 3
0
5
10
15
20
25
CO2
Figure 6: Emissions comparison for the three speed bump runs
Figure 7 plots the post-catalyst emissions, throttle
position, rate of change of throttle position and road
speed for the first traffic-calmed run. The first run was
chosen for the detailed graphical representation since it
was the cleanest run of the three. Throttle position
seems to have a major effect on NOx emissions. As a
speed bump is approached, the throttle is closed and the
level of NOx produced is very low. Then as the vehicle
passes over the speed bump and accelerates away,
thus opening the throttle, the level of engine-out NOx
increases owing to the higher combustion pressure and
temperature. This increase in engine-out NOx is the
principal reason for the post-catalyst NOx peaks shown
in figure 7. Another important factor is the momentary
decrease in catalyst efficiency that results from a brief
lean period experienced immediately after any sudden
throttle application. This would be less of a problem on a
fresh catalyst, but on a high mileage vehicle as used in
this study, air-fuel ratio deviations away from
stoichiometry can drastically affect catalyst efficiency.
One final potential contributor to the post-catalyst NOx
peaks is catalyst temperature fluctuations whilst
negotiating the speed bumps. Since the catalyst in this
study was hot and already lit off, then catalyst efficiency
changes as a result of higher combustion temperatures
are not likely.
A similar trend can be seen for CO because the air-fuel
mixture is slightly enriched when the ECU detects a
sudden change in throttle position, as shown in figure 7.
This is done so that the car accelerates smoothly and
effectively when the driver demands a power increase
by depressing the throttle. This fuel enrichment strategy
is worse in older cars compared to the newer generation
of EURO3 and EURO4 cars, where the ECU is
programmed to maintain stoichiometry for as long as
possible without sacrificing driveability. This is possible
in direct injection systems, but for port fuel injection,
some enrichment is necessary to overcome the brief
period when there is more air than fuel in the intake
manifold. THC follows the same trend as CO for the
same reasoning. CO2 follows the throttle position plot as
well as the fuel flow plot since throttle position is
proportional to engine load which is proportional to fuel
flow rate as mentioned previously. Thus when the load
increases, the fuel injected increases and hence more
CO2 is produced from the combustion process of this
fuel.
The numbers in square brackets to the left of figure 7’s
y-axis are an indication of the scaling. Each number is
the ratio of the y-axis scale after calming to the same
scale before calming. It can be noted that for traffic-
calming most of the scales had to be doubled, and for
NOx tripled, in order for the peaks to be visible.
Figure 8 shows a cumulative plot of the emissions of run
1. For all the pollutants, a flat region can be seen where
the car was turned around. This is a low power condition
and therefore very little emissions were produced
relative to the main drive on the traffic calmed road. This
is in contrast to figure 3 for the non-calmed road, where
a flat region was observed during the drive on the main
road and a sharp increase was recorded (especially for
CO) while turning the car around. In figure 8, the
numerous jagged edges on the plots correspond to all
the speed bumps encountered. The times where there
are sharp increases in the emission levels correspond
with the passing of each speed bump, which in turn is
followed by the leveling off of emissions as the car
travels between the speed bumps.
0 50 100 150 200 250 300 350
0.000
0.005
0.010
0.0
0.1
0.2
0.0
0.1
0.2
0.3
0.4
0
10
20
0
2
4
6
12
14
16
18
-50
0
50
0
20
40
60
0
20
40
60
Time (s)
THC (g/s)
NOx (g/s)
[3]
[2]
CO2 (g/s)
CO (g/s)
Fuel flow (kg/hr)
d(TP)/dt
Throttle position (%)
Air-fuel ratio
stoich.
Speed (km/h)
[1]
[1]
[2]
[2]
[2.5]
[1.5]
[2]
Figure 7: Analysis of speed bump run 1 (numbers in square brackets to the left of y-axis
are the ratio of ‘speed bump’ axis scale to ‘smooth run’ axis scale; d(TP)/dt is rate of
change of throttle position)
0 50 100 150 200 250 300 350
0
2
4
6
8
10
12
14
CO
NOx
THC
Time (s)
Total CO, NOx, THC (g)
0
200
400
600
800
1000
1200
1400
CO2
Total CO2 (g)
Figure 8: Cumulative emissions plot of speed bump run 1
Table 4: %change due to speed bumps
Smooth run Bumps run % change
Time (s) 334 354 +6
Fuel (kg) 0.178 0.241 +35
Dist. (km) 2.258 2.224 -1.5
Avg. speed
(km/h)
39.26 36.46 -7.1
g CO2/km 320 607 +90
g CO/km 2.49 5.40 +117
g NOx/km 1.21 3.57 +195
g THC/km 0.14 0.34 +148
Table 4 is a comparison of the various parameters
calculated previously for the smooth runs and the traffic-
calmed runs. As can be seen from the results, speed
bumps have a dramatic effect on the levels of pollution
entering the atmosphere and the percentage change
varies depending on the pollutant in question. The
catalyst temperatures were left out of this table as they
had no bearing on related performances due to the fact
that the catalyst was hot for each run so the efficiency of
the catalyst was more or less the same for all the runs.
This was not surprising considering the car was fully
warmed up before testing.
The results revealed in this study are compared against
the results obtained by the TRL when they carried out
their own investigation [4] into the effects of speed
bumps using various vehicles driven over various types
of traffic calming devices. Even though TRL conducted a
study of 1.7m and 1.9m wide speed cushions, it was
decided that their 80mm round-top speeds hump study
was more representative of the speed profiles recorded
in the present work. This was because the vehicle in this
study was slowed to ~10mph while negotiating the
speed cushion, and this is normally only necessary for a
round-top speed hump. For this scheme (80mm round-
top speed humps), the TRL tested two different medium-
sized, EURO1 certified, catalyst-equipped cars that are
comparable to the vehicle used in the present work. The
1995 Ford Mondeo and 1996 Vauxhall Astra vehicles
were both 1.6-liter petrol cars, while the test vehicle
used in this study was a 1992 Ford Orion EURO1 petrol
1.8-liter. A comparison is shown in table 5. For its
investigation, TRL conducted two test runs per vehicle
and it must be noted that the variability between these
two runs for the Mondeo vehicle was much higher than
the variability for the Astra vehicle. This means that the
Mondeo results are not as reliable as the Astra results.
Compared with the Mondeo TRL results, this study
yielded almost twice the CO2, three times the CO, four
times the THC and five times the NOx for a traffic-calmed
road versus a non-calmed road. The results are closer
when a comparison is made with the Astra vehicle. Even
though the TRL study included a Ford vehicle, it is not
appropriate to make a direct comparison between the
TRL data and the current study since the cars are
slightly different in terms of their mileage and ECU
strategy.
The discrepancy in results between this study and the
TRL study could be due to the fact that the TRL used a
rolling road dynamometer and therefore the rates of
acceleration were limited due to slippage between the
tire and the roller. This would explain the much higher
NOx obtained in this study since real-world testing does
not have the same limitations on acceleration as
dynamometer testing. Another difference between the
two studies is the speed bump spacing. The speed
bumps in the TRL investigation were spaced 60 meters
apart on average, whereas they were 140 meter apart in
the present work. This allowed the car to accelerate to a
higher speed and therefore producing higher NOx than
the TRL study. Yet another difference is that the vehicle
used in this study was close to fully laden (thus
producing a higher load on the engine) due to the heavy
equipment, whereas dynamometer testing is not usually
based on a fully laden car. The final reason is the
different ECU strategies which are used by the different
manufacturers of the vehicles tested. The data
presented in this investigation is probably a
representation of an unsmooth driver who is in a hurry to
negotiate a traffic-calmed road driving a heavily laden
car. Smoother driving will always produce cleaner
emissions even if there are speed bumps to negotiate.
Table 5: Comparisons with TRL data [4]
This study TRL
Mondeo
TRL Astra
Avg. Speed
(km/h)
-7.1 -67 -67
g CO2/km +90 +43 +28
g CO/km +117 +41 +169
g NOx/km +195 +37 +48
g THC/km +148 +34 +185
It’s worth noting that for the same 80mm round-top
speed hump scheme, the TRL measured much smaller
changes in emissions for non-catalyst petrol cars and
diesel cars. These results are listed in table 6 along with
the results from the catalyst equipped car. All vehicles
are medium sized, with the catalyst-equipped petrol car
and the diesel car being EURO1 certified.
Table 6 : Comparison of TRL cat, non-cat
and diesel cars [3]
Petrol Non-
catalyst
Petrol
Catalyst
Diesel
Avg. Speed
(km/h)
-67 -67 -67
g CO2/km +32 +43 +34
g CO/km +25 +41 +111
g NOx/km +16 +37 +53
g THC/km +55 +34 +53
Non-regulated hydrocarbons
It must be noted that all the THC results reported using
the FTIR are not representative of a true total
hydrocarbon measurement. This is because the FTIR
does not count the C-H bonds as does a conventional
FID analyzer. The FTIR simply identifies all the
hydrocarbons it can (30 in this case) and then sums
them to derive a methane-based THC count. Based on
previous experience [8] the THC results from the FTIR
need to be multiplied by a factor of three in order to be a
rough approximation of a FID. In this study, it was more
important to investigate the change in emissions rather
than the absolute level of emissions. For this purpose,
the THC readings from the FTIR were not corrected in
this report.
Table 7: %change in non-regulated HC's
Smooth
run
Bumps run % change
Toluene
(g/km)
0.002 0.014 600
Formaldehyde
(g/km)
0.006 0.014 133
Acetaldehyde
(g/km)
0.001 0.002 100
1,3-Butadiene
(g/km)
0.003 0.021 600
Benzene
(g/km)
0.013 0.052 300
One of the main advantages of an FTIR is its ability to
speciate 30 out of the ~160 hydrocarbons present in the
exhaust [9]. These non-regulated hydrocarbons such as
benzene and 1,3-butadiene can cause cancer and other
serious health problems [10], and therefore they are
taken into consideration when assessing air quality.
Figures 9 and 10 show graphs of five important
hydrocarbons plotted against road speed and throttle
position for the smooth and speed bump runs
respectively.
For the smooth run, it can be seen that benzene
dominates the analysis with peaks that are 5-10 times
higher than the other four pollutants. Toluene seems to
have peaks only at the beginning and the end of the
drive cycle, with almost no toluene being emitted when
traveling at a constant speed. Formaldehyde is different
from the other four, with no obvious peak, but rather a
series of peaks that are seem to be proportionally
related to throttle position. Acetaldehyde has a very
similar trend to formaldehyde but with an overall smaller
magnitude. Benzene and 1,3-butadiene have similar
trends and they seem to produce most of their emissions
at idle and low power conditions.
For the traffic-calmed run in figure 10, Toluene has
started to show up during the drive cycle, but the two
peaks at the beginning and the end of the drive cycle are
still present. Formaldehyde now has two distinct peaks,
each peak corresponding to the beginning of the drive in
each direction. Acetaldehyde also shows peaks during
the drive cycle, whereas in the smooth run the peaks
were produced at low power conditions when the car
was being turned around. Benzene and 1,3-butadiene
show a similar trend with most of the peaks produced
during the drive cycle as expected.
The numbers in square brackets to the left of the y-axis
in figure 10 are ratios. Each number is a ratio of the
maximum y-axis scale of the traffic-calmed run to the
maximum y-axis scale of the smooth run. The numbers
are a rough indication of the increase in the peaks of
each of the parameters shown in figures 9 and 10. The
pollutant that seems to be most affected by the traffic
calming is formaldehyde with a scale 10 times as high
as the smooth run. Toluene and 1,3-butadiene are also
strongly affected by the traffic calming with a factor of 5.
Benzene and acetaldehyde are least affected, but they
are still approximately twice the magnitude of the
smooth run results. This is made clearer in table 7,
which indicates that the highest total emission increases
are for toluene and 1,3-butadiene, with a seven-fold
difference between smooth and traffic calmed results.
The aldehydes are least affected by traffic calming, but
they are still double the smooth run levels.
0 50 100 150 200 250 300 350
0.000
0.001
0.002
0.0000
0.0002
0.0004
0.0000
0.0001
0.0002
0.0000
0.0001
0.0002
0.0000
0.0002
0.0004
-20
0
20
0
10
20
30
0
20
40
Time (s)
Speed (km/h)
Throttle position (%)
d(TP)/dt
Toluene (g/s)
Formaldehyde (g/s)
Acetaldehyde (g/s)
1,3-Butadiene (g/s)
Benzene (g/s)
Figure 9: Non-regulated hydrocarbons from smooth run
0 50 100 150 200 250 300 350
0.000
0.002
0.004
0.000
0.001
0.002
0.0000
0.0002
0.0004
0.000
0.001
0.002
0.000
0.001
0.002
-50
0
50
0
20
40
60
0
20
40
60
Time (s)
Speed (km/h)
Throttle position (%)
d(TP)/dt
Toluene (g/s)
Formaldehyde (g/s)
Acetaldehyde (g/s)
1,3-Butadiene (g/s)
Benzene (g/s)
[1.5]
[2]
[2.5]
[5]
[10]
[2]
[5]
[2]
Figure 10: Non-regulated hydrocarbons from speed bump run 1 (numbers in square brackets to
the left of y-axis are the ratio of ‘speed bump’ axis scale to ‘smooth run’ axis scale)
DISCUSSION
A comparison of exhaust emissions was made between
a traffic-calmed and a non-traffic calmed scenario on the
same road. The road had a set of seven speed cushions
which were mild enough to negotiate at a constant
average speed of about 25mph. Baseline data was
obtained for a 2-way journey along the road at a
constant speed. Data was then obtained while driving
across the same speed cushions as if they were the
more aggressive road-hump type of speed bumps.
Even though the average speeds of the calmed and
non-calmed runs were similar, a large change in
emissions was recorded. Had the non-calmed speed
been higher than 25mph as was initially planned, it
would have made the difference (compared to the traffic-
calmed run) even greater. This is because a vehicle
produces fewer emissions as the average speed
increases since the engine operates in a more efficient
regime at higher speeds (up to ~40mph). At speeds
higher than ~40mph, the aerodynamic drag of the
vehicle tends to push emissions back up [11].
The results obtained from this study were compared to a
similar investigation carried out by the TRL. The present
work yielded much higher changes in emissions
compared to the TRL study. One reason is that the TRL
study used a rolling road dynamometer to reproduce
drive cycles that were obtained from real-world driving
speed profiles. Consequently, the emissions produced
were not obtained on-road and therefore might have
been limited in terms of acceleration due to slippage
between the tire and the rolling road. Another reason for
the increase is the unsmooth nature of the driver
negotiating the speed bumps in this study. Normally,
well-designed speed cushions do not require the driver
to slow to 10mph as was done in this study. That much
of a retardation is only necessary for the more
aggressive round-top speed-hump type of traffic
calming. Another factor is the heavy weight of the car.
The car used in this study was almost fully laden with
equipment and two people on board. Even though the
vehicles used in this and the TRL studies are similar in
size, ECU strategies used by different manufacturers
have a large influence on the levels of emissions
produced. For all the reasons mentioned, it was not
surprising to see that the results from this study do not
agree very well with the TRL report.
Speed cushions do limit speeds to around 30mph as
evidenced by the fact that the car was driven over them
at a constant average speed of ~23mph without much
discomfort to the occupants. Therefore this study is
more a representation of the effect of round-top speed
humps on emissions rather than speed cushions.
It can be argued the traffic calming in this case was not
as effective as the TRL reported, with a 7% reduction in
average speed versus a 67% reduction. This is not
necessarily true since a car would be able to maintain a
higher speed than was done in this study if the speed
cushions had not been present. Even if a comparison
had been made between a 50mph non-traffic calmed run
and a 25mph traffic calmed run, the results would not
have been significantly different, and might even have
exaggerated the %change in emissions.
A EURO1 vehicle was used in this study, but in future
work, similar tests will be performed on EURO2, 3 and 4
vehicles as part of an ongoing project to measure and
model real-world traffic emissions.
CONCLUSION
Emissions for a traffic calmed road employing speed
humps were shown to be 2-3 times as high as a non-
calmed road negotiated smoothly. This was measured
on a mass basis using an FTIR installed in-vehicle on a
EURO1 petrol-fuelled passenger car. CO2 was found to
increase by 90%, CO by 117%, NOx by 195% and THC
by 148%. Five toxic species of hydrocarbons were also
examined and found to increase dramatically due to
speed bumps. As far as the authors are aware, this is
the first on-road study of the real-world effects of traffic
calming on exhaust emissions. The use of the FTIR for
emissions measurements can provide quantitative
hydrocarbon speciation data which can potentially be
used to calculate ozone forming potentials in future.
ACKNOWLEDGMENTS
The authors would like to thank the UK EPSRC for a
research grant, GR/M88167/01, for a JIF (Joint
Infrastructure Fund) award for the LANTERN project
(Leeds health, Air quality, Noise, Traffic, Emissions
Research Network). We would also like to thank the
EPSRC for the research grant, GR/S31136/01, for the
RETEMM (REal-world Traffic Emissions Measurement
and Modeling) project award in support of this work as
well as the FUTURES grant GR/S90881/01. Basil
Daham would like to thank the University of Leeds for a
research scholarship. Thanks are also due to Bob
Boreham for his technical expertise during this work.
REFERENCES
1. http://www.thinkroadsafety.gov.uk/statistics.htm
2. ITE, Traffic calming: State of the Practice,
ITE/FHWA, pp. 66-74, August 1999
3. Sawers, C.P., Introduction to Traffic Calming,
http://www.mini-roundabout.com/calming, Pentratt
and MoorWeb, April 2004
4. TRL, The Impact of Traffic Calming measures on
vehicle exhaust emissions, TRL report 482, 2001
5. De Vlieger et al., Environmental Effects of Driving
Behaviors and Congestion Related to Passenger
Cars, Atmospheric Environment, vol 34, pp.4649-
4655, March 2000
6. Rapone et al., Driving behavior and emission results
for a small size gasoline car in urban operation, SAE
Technical paper 2000-01-2960
7. Hans Peter and Christian Cozzarini, Emissions and
Air Quality, SAE Reference book R-237, 2000
8. Daham et al., Application of a Portable FTIR for
Measuring On-road Emissions, SAE Technical
Paper, SAE 2005-01-0676, 2005
9. Villinger et al., Dynamic monitoring of differentiated
hydrocarbons in direct engine exhaust: a versatile
tool in engine development, SAE Technical paper
960063
10. Anon, Control of Emissions of Hazardous Air
pollutants from Mobile Sources, EPA Federal
Register vol. 66 no.61 pp1720, March 29 2001
11. Commission of the European Communities, “The
European Auto Oil Programme” A report by the
Directorate Generals for Industry, Energy and
Environment, Civil Protection and Nuclear Safety
(XI/361/96), 1996
CONTACT
Please direct all correspondence to Basil Daham,
Energy & Resources Research Institute, University of
Leeds, Leeds LS2 9JT, UK. basildaham@yahoo.com
ACRONYMS
MEET: Methodologies for Estimating Air Pollutant
Emissions from Transport.
MODEM: MODeling of EMissions and consumption in
urban areas
ITE: Institute of Transport Engineers
FHWA: Federal HighWay Administration
TRL: Transport Research Laboratory
FID: Flame Ionization Detector
DETR: Department of the Environment, Transport and
the Regions
THC: Total HydroCarbons
