Amid. Am/. & Prev. Vol. 24. No Printed in Great Britain.
I, pp. 17-28.
SAFETY MEASURES
0001-4575192 S.OU + .OO 0 1992 Pcrgamon press plc
1992
EFFECTS OF SPEED REDUCING IN DANISH RESIDENTIAL AREAS ULLA ENGEL and LARS K. THOMSEN
Danish
Council
of Road
Safety
Research,
Ermelundsvej
101, DK-2820
Gentofte.
Denmark
Abstract-On May 1. 1977 a new code was introduced into the Danish Road Traffic Act. The result was a change in layout and speed limits in a great number of residential streets; in most cases the streets were transformed into 30 km/h streets and in few cases into 15 km/h streets. In addition to speed signs, both types of streets were equipped with speed reducing measures. Based on experiences from a selection of experimental streets, mostly 30 km/h streets, different, but very positive effects were found. Overall there was a reduction in the mean speed in these areas of 11 km/h. On the 223 km, 30 km/h streets there was a reduction of 77 accidents and 88 casualties within a period of three years. These reductions were caused by the implementation of speed signs, speed reducing measures, and a reduction in traffic. On the basis of 44 experimental streets, where traffic was recorded both before and after the changes, the reduction in risk of casualties, i.e. the number of casualties per road user km, was 72%, while the risk of accidents seemed to be unchanged. Considering serious injuries alone, a very high reduction of 78% was found. Accidents included in the study consist of all police reported accidents, i.e. accidents with personal injury as well as damage only accidents.
BACKGROUND
Following a change in the Danish Road Traffic Act, which came into force on November 1, 1978, it was made legal under certain circumstances to change the status of streets from “traffic streets, ” i.e. streets with priority for vehicles, to “living areas” with priority for pedestrians. The implementation of “living areas” demanded the implementation of different physical measures, among others an advisory speed limit of 1.5 km/h (see Fig. 1). However, in order to meet demands from the public as well as the local authorities, a second type of “living area” was created. Here, the demands were an advisory speed limit of 30 km/h (see Fig. 1) and some, but fewer, restrictions concerning physical measures. The costs of the implementation of this street type were much lower compared to the costs of the implementation of 15 km/h streets. It was the 30 km/h streets, that became popular and most frequent, especially in existing residential areas. Paradoxically, the 30 km/h streets involved no change in the traditional priority for vehicles. The evaluation is therefore mainly based on a street type that was only an indirect outcome of the change in the Road Traffic Act. This paper shows briefly the main parameters included in the research project together with photos of experimental streets and countermeasures. Second, the paper presents the results of the different studies of the project, and finally, two examples are given showing how to calculate expected speed changes on the basis of a regression model. A summary report contains the most important findings. The report has an extensive summary in English (Engel and Thomsen, 1990b).
MAIN
PARAMETERS
The evaluation of effects is based on two main studies concerning accidents and motor vehicle speeds. The “accident evaluation” includes all police reported accidents, both accidents with personal injury and damage only accidents. The “casualty evaluation” includes only the personal injury accidents. The research design in both studies is a before-and-after study with control group, and the before-and-after periods were three years each in the accident studies. 17
‘h SI:reets and 30 km/h
streets
Two different studies concerned the accident frequency. A total of 729 experimental streets has been investigated before and after the changes in status (i.e. local reorganisation) were implemented. From this study we learned about the local reorganization of streets in Denmark between November 1, 1978 and July 31, 1983 and about the changes in the number of road accidents per road km. (Engel and Thomsen 1989a). Simultaneously a second study was performed. A total of 44 experimental streets and 52 control streets were investigated more intensely. The group of experimental streets consisted of almost all streets with changed status in Denmark from July 1, 1982 until June 30, 1983. The control streets were selected to match each of the experimental streets. From this study we Icarned about the changes in the number of road accidents per road-user km. (Engcl and Thornsen 1989b).
On 41 of 44 experimental streets and 13 of 52 control streets the speeds of motor vehicles were recorded. No significant change in the mean speed on the control streets was found from the before to the after period. Hence. further studies were based only on data from the experimental streets. The speed of X.504 motor vehicles is included in this part of the study. The speed measurements were collected continuously over a distance of up to 200 m. Hence, it was possible to observe changes in not only the mean speed but also in “speed profiles” before and after the implementation of the physical so-called speed reducing measures. Figure 2 presents an example of the results of speed measurements taken at an experilnent~~l site. From this study we learned about the changes in speeds as a consequence of the implementation of different countermeasures (Engel and Thomsen 199Oa). Experimental streets Figures 3-X illustrate the variety of street types selected with the variety in types of countermeasures.
for 30 km/h
streets
together
RESULTS
Accidents per kilometer of roud This part of the project concerns 10 km of 15 km/h streets and 223 km of 30 km/h streets. The control group consisted of all urban streets in Denmark belonging to the local government auth(~rities-a~1 together, 18,935 km of streets. The accidents-all police reported accidents-and casuatties-police reported accidents with personal itrjury-are related to the length of each street. No traffic figures were avaiiable for this retrospective study. No significant changes were found in the 15 km/h streets. The main effect was a significant change in accidents in 30 km/h streets of -24% (corresponding to 77 fewer accidents in three years). In the same street type there was a change in casualties of - 45% (corresponding to 88 fewer casualties in three years) (see Table 1). In street sections adjoining the 30 km/h streets-outer areas in Tables 1 and 2 there in three years) and was a change in accidents of - 18% (i.e. 150 fewer accidents
I
I +-
20
U. ENGEL and L. K. THOMSON
Fig. 3. 30 km/h
street,
St. St. Blichersvej.
Arhus
in casualties of -21% (i.e. 106 fewer casualties in three years). These after the trends in the control group have been taken into consideration. Accidents per road user km This part of the project contains five 15 km/h streets and thirty-nine all together 30 km of experimental streets. The control group contains 35 km, all together.
Fig. 4. 30 km/h
street,
Mulius
Erichsensvej,
Aalborg.
are the results
30 km/h streets. 52 streets, about
The accidents and casualties are related to the number of road user km travelled in each street, one day in the periods 6 A.M.-10 A.M.and 12 A.M.-~P.M. The designation “road users” covers all categories of traffic including also pedestrians, pedal cyclists, and moped riders. The main effect is a significant change in the number of casualties per road user km of -72Y0, with confidence limits ranging from -4% to - 92%, due to the change in street status (see Table 2). Furthermore, there was a significant change in the number of seriously injured of
Fig. 6. 30 km/h street, Trevangsvej. Farum. AAP 24:1-c
22
U. ENGEL and L. K. THOMSEN
Fig. 7. 30 km/h
street,
Carit
Etlarsvej.
.bhus.
- 78%, with confidence limits ranging from - 26% to - 93%. The number of casualties in the experimental and the control street is shown in Table 3. All figures mentioned above are results after the trends in the control group have been taken into consideration.
Motor vehicle speed Thirty km/h streets must be designed in a way that discourages motor vehicle drivers from driving at speeds exceeding 30 km/h. To obtain such behaviour, speed reducing
Fig. 8. 30 km/h
street,
Glentevej.
Odense
Safety Table
I. Change
effects
of speed
reducing
measures
23
in accidents and casualties per road km on fifty 15 km/h streets streets distributed on inner areas’ and outer areas2 % of change
type
Area
Change disregarding control group
15 km/h streets
Inner’ Outer*
- .76% -0.64%
30 km/h streets
Inner1 Outer’
- 24.92%’ - 19.43%3
Change/Street
in accidents
% of change
Change including control group 0.48% 0.61% -23.97%3 - 18.42%3
and 679 30 km/h
in casualties
Disregarding control group
Including control group
- 100% - 39.92%’
- 100% - 25.40%
- 56.04%j -36.51%
-45.42%) -21.16%3
‘Inner area, i.e. the part of the street limited by the speed-limit sign. ‘Outer areas. i.e. parts of the street located just outside the inner area. ‘Change significant on the 5% level.
measures have to be implemented with a distance of a maximum of 100 m according to design standards. Table 4 presents some results of speed measurements taken at the centre of the countermeasures at distances of 50 m from the center. Results are shown for each of seven types of countermeasures. From the table it can be seen that the greatest change in speed is caused by humps, even at a distance of 50 m.
Factors with influence on speeds
A regression model for the speed changes was created. To identify the most significant factors that influence speed change from the before period to the after period, speed data from 1,002 sections were used. By use of the model it is possible to calculate the expected change in speed in the carriageway, as a consequence of the implementation of a specific type of countermeasure. Below, the factors most important to the speed change are listed. Characteristics of countermeasure:
(4 Height of hump
(ii) Type of lateral dislocation (iii) Type of street narrowing (iv) Distance from point in the carriageway to primary countermeasure and the countermeasure (v) Distance between primary countermeasure
just before (vi) Distance between just after
primary countermeasure
and the countermeasure
Table 2. Change in number of casualties per road user km in inner areas and outer areas of 44 experiment streets Type of area
Casualties per road user km Change
significant
Inner
Outer
- 72%’
+96%
on the 5% level.
located located
U. ENGEL and L. K. THOMSEN
24 Table 3. Casualties
on 44 experimental
streets and 52 control streets of severity and time period Inner
Experimental
(inner areas)
distributed
on degree
areas
streets
Control
streets
Fatal
Serious injuries
Before After
0
27
1
I
13 5
40 13
1 0
x 10
10 7
1Y 17
Total
1
34
18
53
1
18
17
36
Characteristics (vii)
Light injuries
Total
Fatal
Serious injuries
Light injuries
Total
of speed before:
Mean speed in the point of the carriageway
The model and the figures of the different elements in the model is shown in Table 5. The model is used by filling in the relevant data for a given point at the road and summarizing the effects of the relevant elements. The main findings as a result of model calculations are as follows: (i)
The height o f a h ump has proved to be of greatest importance to speed change. Per 1 cm in height, there is an expected speed reduction of one km/h. Hence, a hump 10 cm high will produce a speed reduction of 10 km/h. (ii) The presence of a narrowing of the carriageway will produce a speed reduction of 4.7 km/h and so will the presence of a double lateral dislocation. A single lateral dislocation will produce a reduction in speed of only 2 km/h. Table 4. Speed changes in three positions (-50 m, center of the countermeasure. +50 m) for a maximum of eight countermeasures. given together with the 95% confidence limits Percentiles Countermeasure
Change in speed (km/h)
2.5
97.5
-50 m hump, circle segment hump, elevated junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
~ 13.7 - 13.7 -6.5 -26.8 12.1 -3.7 - 1.7
~ 17.6 ~ 16.5 ~ 12.7 -32.1 - 15.5 -7.0 px.7
-Y.8 ~ IO.‘) -0.3 --21.5 - 8.7 -0.4 5.3
0 m. center of countermeasure hump, circle segment hump, elevated junction hump, elevated area before junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
-7.9 - 14.7 3.2 -21.2 -8.1 4.1 0.8 -3.4
- 10.5 - 17.4 ~ 10.5 -26.1 - 10.3 1.8 - 1.3 -8.1
ps.3 12.0 16.Y - 16.3 -5.9 6.4 3.0 1.3
50 m hump, circle segment humn. elevated junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
3.6 _ 8.3 -23.6 - 16.8 3.2 -9.5 - 14.1
-22.2 -24.2 -47.3 - 27.0 - 14.2 - 28.2 -23.2
15.0 7.6 0.1 -6.6 20.6 9.2 - s.0
Safety
effects
of speed
reducing
measures
25
Table 5. Statistical model and its variables for calculation of expected speed changes at individual road section. The degree of explanation of the model is 61%. The level of significance for each variable is below .Ol% Speed 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
change = 29.058 km/h (constant factor) -0.6451 km/h x mean speed before (km/h) 0.00376 km/h x distance in metres to the primary countermeasure 0.0005352 km/h x (distance in metre)? to the primary countermeasure - 148.32 km/h x LN(l + l/Distance in metre from the primary countermeasure termeasure before) -81.50 km/h x LN(l + l/Distance in metres from the primary countermeasure termeasure after) - 1.001 km/h x height in cm of hump (if a hump is present) -2.017 km/h (if single lateral dislocation is present) -4.724 km/h (if double lateral dislocation is present) -4.680 km/h (if street narrowing is present)
to the counto the coun-
Regarding the distance between countermeasures, it has been possible to reach a decrease in speed down to the recommended level of 30 km/h in distances more than 50 m away from the countermeasure by the use of different types of measures presented in this study. Comparisons between estimated speed changes and observed speed changes show the quality of the model. Disparities of 5-10 km/h are not uncommon. The model explains 61% of the total variation. Separate studies show that the results mentioned above are valid for passenger cars as well as for delivery vans and lorries. The long-term effect of the countermeasures has been investigated as well. On selected experiment streets, speed measurements were taken not only one year after but also two years after the implementation of the countermeasures. No systematic changes were found between the speed measurements taken in these two periods. Speed was measured on 13 control streets, for 1,948 motor vehicles. There was a tendency to a slight increase in the mean speed from the before period to the after period. This general trend was not included in the estimations of effects of the countermeasures in the experiment streets, and, hence, the results of the countermeasure may have been slightly underestimated in the study. How to calculate speed changes
Two examples will be given of how to use the regression model in real life. Example 1: Countermeasure: Narrowing of the carriageway (10) Mean speed before: 65 km/h (2) Distance from the car to the primary countermeasure: 50 m (3) (4) Distance from the primary countermeasure to the countermeasure before: 100 m (5) Distance from the primary countermeasures to the countermeasure after: 100 m (6) The figures are repeated in Fig. 9 with the figures in parentheses referring to the figures in Table 5. Figure 10 illustrates the situation after the implementation of the countermeasure (see also Table 6). The result of the implementation of a narrowing of the carriageway is an expected decrease in mean speed at a distance of 50 m from the measure of 18.31 km/h, when the mean speed before was 65 km/h and the secondary countermeasures (i.e. other speed reducing measures or a junction or a speed limit sign) were situated at a distance of 100 m from the primary countermeasure; i.e. in the car, the mean speed dropped from 65 km/h to 46.69 km/h.
26
U. ENGEI. and L. K. THOMSEN
I
65 KM/H I 1 I+100 M
f
50M
100 M
A
-+
I I JUNCTION “COUNTERMEASURE” (SECONDARY)
JUNCTION “COUNTERMEASURE” (SECONDARY)
CENTER OF “COUNTERMEASURE” (PRIMARY)
Fig. 9. Street plan for Example 1 before the implementation of a countermeasure. The change speed will be estimated for the position of the car, marked in the figure.
in mean
30 KM/H
Fig. 10. Street
plan for Example
1 after the implementation
of the countermeasure
30 KM/H
30 KM/H
I Fig. 11. Street
plan for Example
2 after the expansion
of countermeasure
Safety
effects
Table 6. Calculation Element no.
reducing
of speed change
Value
Element
Present
-4.680
Table 7. Calculation
(Example
27 1)
effect
km/h
of speed change
Value Present 15 cm 65 km/h 50 metres 50’ ” 100 ” 100 ” Present
measures
- 4.680
-0.6451 0.00376 km/h km/h x x 6550km/h m 0.0005352 km/h x 5O’m -148.32 km/h x LN(l + 11100) -81.50 km/h x LN(1 + 11100) 29.058 km/h Total
50 65 m km/h 50’ 100 m 100 m Present
Element no.
of speed
Element
(Example
= = = = =
-41.9315 0.1880 1.3380 -1.4758 -0.8110 29.068 -18.31 km/h
2)
effect
-4.680 km/h -1.001 km/h x 15 cm -0.6451 km/h x 65 km/h 0.00376 km/h x 50 metres 0.0005352 km/h x 50: metres -148.32 km/h x LN(l + l/100) -81.50 km/h x LN(l + 11100) 29.058 km/h Total
= = = = =
=
-4.680 - 15.015 -41.9315 0.1880 1.3380 - 1.4758 -0.8110 29.058 -33.33 km/h
With the countermeasure selected, the objective of a maximum speed of 30 km/h was not achieved. In the second example, the countermeasure was more restrictive by installation of a hump of height 15 cm. Example 2: Countermeasure: Narrowing of the carriageway (10) Countermeasure: Hump with the height 15 cm (7) Mean speed before: 65 km/h (2) Distance from the car to the primary countermeasure: 50 metres Distance from the primary countermeasure to the countermeasure (5) Distance
from the primary
countermeasures
to the countermeasure
(3) (4) before: after:
100 m
100 m (6)
The figures are repeated in Fig. 9 with the figures in parentheses referring to the figures in Table 5. Figure 11 illustrates the situation after the implementation of the hump. The results of the narrowing of the carriageway and adding a hump of a height of 15 cm produced an expected speed change of -33.33 km/h, from 65 km/h to 31.67 km/h. In this case a mean speed below 30 km/h was nearly obtained (see Table 7).
Acknowledgements-The start of this research project was facilitated by the Danish Technical Research Council with a donation of 130,000.- Danish Kroner in 1982. Many students and young candidates have been employed in the project. Special thanks are given to Erik Stern Nielsen and Morten Agerlin Petersen.
REFERENCES Engel. U.; Thomsen, L. K. $40 gaders sikkerhed. En analyse af politirapporterede ulykker gader baseret pa ulykkestztheder for hhv. efter aendring af gadestatus. Documentation Copenhagen: Danish Council of Road Safety Research; 1989a.
pB danske $40 Report 1189.
28
U. ENW
L
and L. K.
THOMStN
Engel. U.: Thomscn. L. K. $30 gadcrs sikkerhed. En analyse af politirapportercdc ulykker pa 14 WI gadcr og 52 kontrolgadcr baseret pa ulykkesfrekvenser for hhv. efter sndring af gadestatus. t~ocumt‘ntatlon Report 2i8Y. Copenhagen: Danish Council of Road Safety Research: 1YXYb. En&cl, U.; Thomsen. L. K. $40 gaders sikkerhed. En analyse af bilers korehastighcder pa 31 #-IO gadcr og 13 kontrolgader for hhv. efter zendring of gadestatus. Documentation Report 2!90. Copenhagen: Danish Council of Road Safety Research; 1990a. Engel, U; Thomsen. L. K. Effekter af fzerdselslovens 940. Anvendelse af $40 i Danmark samt effekter ph trafikulykker, risikoforhold og hastigheder; Sammenfatning. Summary Report 29. Copenhagen: Danish Council of Road Safety Research; 1990b.
I, pp. 17-28.
SAFETY MEASURES
0001-4575192 S.OU + .OO 0 1992 Pcrgamon press plc
1992
EFFECTS OF SPEED REDUCING IN DANISH RESIDENTIAL AREAS ULLA ENGEL and LARS K. THOMSEN
Danish
Council
of Road
Safety
Research,
Ermelundsvej
101, DK-2820
Gentofte.
Denmark
Abstract-On May 1. 1977 a new code was introduced into the Danish Road Traffic Act. The result was a change in layout and speed limits in a great number of residential streets; in most cases the streets were transformed into 30 km/h streets and in few cases into 15 km/h streets. In addition to speed signs, both types of streets were equipped with speed reducing measures. Based on experiences from a selection of experimental streets, mostly 30 km/h streets, different, but very positive effects were found. Overall there was a reduction in the mean speed in these areas of 11 km/h. On the 223 km, 30 km/h streets there was a reduction of 77 accidents and 88 casualties within a period of three years. These reductions were caused by the implementation of speed signs, speed reducing measures, and a reduction in traffic. On the basis of 44 experimental streets, where traffic was recorded both before and after the changes, the reduction in risk of casualties, i.e. the number of casualties per road user km, was 72%, while the risk of accidents seemed to be unchanged. Considering serious injuries alone, a very high reduction of 78% was found. Accidents included in the study consist of all police reported accidents, i.e. accidents with personal injury as well as damage only accidents.
BACKGROUND
Following a change in the Danish Road Traffic Act, which came into force on November 1, 1978, it was made legal under certain circumstances to change the status of streets from “traffic streets, ” i.e. streets with priority for vehicles, to “living areas” with priority for pedestrians. The implementation of “living areas” demanded the implementation of different physical measures, among others an advisory speed limit of 1.5 km/h (see Fig. 1). However, in order to meet demands from the public as well as the local authorities, a second type of “living area” was created. Here, the demands were an advisory speed limit of 30 km/h (see Fig. 1) and some, but fewer, restrictions concerning physical measures. The costs of the implementation of this street type were much lower compared to the costs of the implementation of 15 km/h streets. It was the 30 km/h streets, that became popular and most frequent, especially in existing residential areas. Paradoxically, the 30 km/h streets involved no change in the traditional priority for vehicles. The evaluation is therefore mainly based on a street type that was only an indirect outcome of the change in the Road Traffic Act. This paper shows briefly the main parameters included in the research project together with photos of experimental streets and countermeasures. Second, the paper presents the results of the different studies of the project, and finally, two examples are given showing how to calculate expected speed changes on the basis of a regression model. A summary report contains the most important findings. The report has an extensive summary in English (Engel and Thomsen, 1990b).
MAIN
PARAMETERS
The evaluation of effects is based on two main studies concerning accidents and motor vehicle speeds. The “accident evaluation” includes all police reported accidents, both accidents with personal injury and damage only accidents. The “casualty evaluation” includes only the personal injury accidents. The research design in both studies is a before-and-after study with control group, and the before-and-after periods were three years each in the accident studies. 17
‘h SI:reets and 30 km/h
streets
Two different studies concerned the accident frequency. A total of 729 experimental streets has been investigated before and after the changes in status (i.e. local reorganisation) were implemented. From this study we learned about the local reorganization of streets in Denmark between November 1, 1978 and July 31, 1983 and about the changes in the number of road accidents per road km. (Engel and Thomsen 1989a). Simultaneously a second study was performed. A total of 44 experimental streets and 52 control streets were investigated more intensely. The group of experimental streets consisted of almost all streets with changed status in Denmark from July 1, 1982 until June 30, 1983. The control streets were selected to match each of the experimental streets. From this study we Icarned about the changes in the number of road accidents per road-user km. (Engcl and Thornsen 1989b).
On 41 of 44 experimental streets and 13 of 52 control streets the speeds of motor vehicles were recorded. No significant change in the mean speed on the control streets was found from the before to the after period. Hence. further studies were based only on data from the experimental streets. The speed of X.504 motor vehicles is included in this part of the study. The speed measurements were collected continuously over a distance of up to 200 m. Hence, it was possible to observe changes in not only the mean speed but also in “speed profiles” before and after the implementation of the physical so-called speed reducing measures. Figure 2 presents an example of the results of speed measurements taken at an experilnent~~l site. From this study we learned about the changes in speeds as a consequence of the implementation of different countermeasures (Engel and Thomsen 199Oa). Experimental streets Figures 3-X illustrate the variety of street types selected with the variety in types of countermeasures.
for 30 km/h
streets
together
RESULTS
Accidents per kilometer of roud This part of the project concerns 10 km of 15 km/h streets and 223 km of 30 km/h streets. The control group consisted of all urban streets in Denmark belonging to the local government auth(~rities-a~1 together, 18,935 km of streets. The accidents-all police reported accidents-and casuatties-police reported accidents with personal itrjury-are related to the length of each street. No traffic figures were avaiiable for this retrospective study. No significant changes were found in the 15 km/h streets. The main effect was a significant change in accidents in 30 km/h streets of -24% (corresponding to 77 fewer accidents in three years). In the same street type there was a change in casualties of - 45% (corresponding to 88 fewer casualties in three years) (see Table 1). In street sections adjoining the 30 km/h streets-outer areas in Tables 1 and 2 there in three years) and was a change in accidents of - 18% (i.e. 150 fewer accidents
I
I +-
20
U. ENGEL and L. K. THOMSON
Fig. 3. 30 km/h
street,
St. St. Blichersvej.
Arhus
in casualties of -21% (i.e. 106 fewer casualties in three years). These after the trends in the control group have been taken into consideration. Accidents per road user km This part of the project contains five 15 km/h streets and thirty-nine all together 30 km of experimental streets. The control group contains 35 km, all together.
Fig. 4. 30 km/h
street,
Mulius
Erichsensvej,
Aalborg.
are the results
30 km/h streets. 52 streets, about
The accidents and casualties are related to the number of road user km travelled in each street, one day in the periods 6 A.M.-10 A.M.and 12 A.M.-~P.M. The designation “road users” covers all categories of traffic including also pedestrians, pedal cyclists, and moped riders. The main effect is a significant change in the number of casualties per road user km of -72Y0, with confidence limits ranging from -4% to - 92%, due to the change in street status (see Table 2). Furthermore, there was a significant change in the number of seriously injured of
Fig. 6. 30 km/h street, Trevangsvej. Farum. AAP 24:1-c
22
U. ENGEL and L. K. THOMSEN
Fig. 7. 30 km/h
street,
Carit
Etlarsvej.
.bhus.
- 78%, with confidence limits ranging from - 26% to - 93%. The number of casualties in the experimental and the control street is shown in Table 3. All figures mentioned above are results after the trends in the control group have been taken into consideration.
Motor vehicle speed Thirty km/h streets must be designed in a way that discourages motor vehicle drivers from driving at speeds exceeding 30 km/h. To obtain such behaviour, speed reducing
Fig. 8. 30 km/h
street,
Glentevej.
Odense
Safety Table
I. Change
effects
of speed
reducing
measures
23
in accidents and casualties per road km on fifty 15 km/h streets streets distributed on inner areas’ and outer areas2 % of change
type
Area
Change disregarding control group
15 km/h streets
Inner’ Outer*
- .76% -0.64%
30 km/h streets
Inner1 Outer’
- 24.92%’ - 19.43%3
Change/Street
in accidents
% of change
Change including control group 0.48% 0.61% -23.97%3 - 18.42%3
and 679 30 km/h
in casualties
Disregarding control group
Including control group
- 100% - 39.92%’
- 100% - 25.40%
- 56.04%j -36.51%
-45.42%) -21.16%3
‘Inner area, i.e. the part of the street limited by the speed-limit sign. ‘Outer areas. i.e. parts of the street located just outside the inner area. ‘Change significant on the 5% level.
measures have to be implemented with a distance of a maximum of 100 m according to design standards. Table 4 presents some results of speed measurements taken at the centre of the countermeasures at distances of 50 m from the center. Results are shown for each of seven types of countermeasures. From the table it can be seen that the greatest change in speed is caused by humps, even at a distance of 50 m.
Factors with influence on speeds
A regression model for the speed changes was created. To identify the most significant factors that influence speed change from the before period to the after period, speed data from 1,002 sections were used. By use of the model it is possible to calculate the expected change in speed in the carriageway, as a consequence of the implementation of a specific type of countermeasure. Below, the factors most important to the speed change are listed. Characteristics of countermeasure:
(4 Height of hump
(ii) Type of lateral dislocation (iii) Type of street narrowing (iv) Distance from point in the carriageway to primary countermeasure and the countermeasure (v) Distance between primary countermeasure
just before (vi) Distance between just after
primary countermeasure
and the countermeasure
Table 2. Change in number of casualties per road user km in inner areas and outer areas of 44 experiment streets Type of area
Casualties per road user km Change
significant
Inner
Outer
- 72%’
+96%
on the 5% level.
located located
U. ENGEL and L. K. THOMSEN
24 Table 3. Casualties
on 44 experimental
streets and 52 control streets of severity and time period Inner
Experimental
(inner areas)
distributed
on degree
areas
streets
Control
streets
Fatal
Serious injuries
Before After
0
27
1
I
13 5
40 13
1 0
x 10
10 7
1Y 17
Total
1
34
18
53
1
18
17
36
Characteristics (vii)
Light injuries
Total
Fatal
Serious injuries
Light injuries
Total
of speed before:
Mean speed in the point of the carriageway
The model and the figures of the different elements in the model is shown in Table 5. The model is used by filling in the relevant data for a given point at the road and summarizing the effects of the relevant elements. The main findings as a result of model calculations are as follows: (i)
The height o f a h ump has proved to be of greatest importance to speed change. Per 1 cm in height, there is an expected speed reduction of one km/h. Hence, a hump 10 cm high will produce a speed reduction of 10 km/h. (ii) The presence of a narrowing of the carriageway will produce a speed reduction of 4.7 km/h and so will the presence of a double lateral dislocation. A single lateral dislocation will produce a reduction in speed of only 2 km/h. Table 4. Speed changes in three positions (-50 m, center of the countermeasure. +50 m) for a maximum of eight countermeasures. given together with the 95% confidence limits Percentiles Countermeasure
Change in speed (km/h)
2.5
97.5
-50 m hump, circle segment hump, elevated junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
~ 13.7 - 13.7 -6.5 -26.8 12.1 -3.7 - 1.7
~ 17.6 ~ 16.5 ~ 12.7 -32.1 - 15.5 -7.0 px.7
-Y.8 ~ IO.‘) -0.3 --21.5 - 8.7 -0.4 5.3
0 m. center of countermeasure hump, circle segment hump, elevated junction hump, elevated area before junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
-7.9 - 14.7 3.2 -21.2 -8.1 4.1 0.8 -3.4
- 10.5 - 17.4 ~ 10.5 -26.1 - 10.3 1.8 - 1.3 -8.1
ps.3 12.0 16.Y - 16.3 -5.9 6.4 3.0 1.3
50 m hump, circle segment humn. elevated junction hump, plateau and circle segment hump, plateau lateral dislocation, single lateral dislocation, double narrowing of carriageway
3.6 _ 8.3 -23.6 - 16.8 3.2 -9.5 - 14.1
-22.2 -24.2 -47.3 - 27.0 - 14.2 - 28.2 -23.2
15.0 7.6 0.1 -6.6 20.6 9.2 - s.0
Safety
effects
of speed
reducing
measures
25
Table 5. Statistical model and its variables for calculation of expected speed changes at individual road section. The degree of explanation of the model is 61%. The level of significance for each variable is below .Ol% Speed 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
change = 29.058 km/h (constant factor) -0.6451 km/h x mean speed before (km/h) 0.00376 km/h x distance in metres to the primary countermeasure 0.0005352 km/h x (distance in metre)? to the primary countermeasure - 148.32 km/h x LN(l + l/Distance in metre from the primary countermeasure termeasure before) -81.50 km/h x LN(l + l/Distance in metres from the primary countermeasure termeasure after) - 1.001 km/h x height in cm of hump (if a hump is present) -2.017 km/h (if single lateral dislocation is present) -4.724 km/h (if double lateral dislocation is present) -4.680 km/h (if street narrowing is present)
to the counto the coun-
Regarding the distance between countermeasures, it has been possible to reach a decrease in speed down to the recommended level of 30 km/h in distances more than 50 m away from the countermeasure by the use of different types of measures presented in this study. Comparisons between estimated speed changes and observed speed changes show the quality of the model. Disparities of 5-10 km/h are not uncommon. The model explains 61% of the total variation. Separate studies show that the results mentioned above are valid for passenger cars as well as for delivery vans and lorries. The long-term effect of the countermeasures has been investigated as well. On selected experiment streets, speed measurements were taken not only one year after but also two years after the implementation of the countermeasures. No systematic changes were found between the speed measurements taken in these two periods. Speed was measured on 13 control streets, for 1,948 motor vehicles. There was a tendency to a slight increase in the mean speed from the before period to the after period. This general trend was not included in the estimations of effects of the countermeasures in the experiment streets, and, hence, the results of the countermeasure may have been slightly underestimated in the study. How to calculate speed changes
Two examples will be given of how to use the regression model in real life. Example 1: Countermeasure: Narrowing of the carriageway (10) Mean speed before: 65 km/h (2) Distance from the car to the primary countermeasure: 50 m (3) (4) Distance from the primary countermeasure to the countermeasure before: 100 m (5) Distance from the primary countermeasures to the countermeasure after: 100 m (6) The figures are repeated in Fig. 9 with the figures in parentheses referring to the figures in Table 5. Figure 10 illustrates the situation after the implementation of the countermeasure (see also Table 6). The result of the implementation of a narrowing of the carriageway is an expected decrease in mean speed at a distance of 50 m from the measure of 18.31 km/h, when the mean speed before was 65 km/h and the secondary countermeasures (i.e. other speed reducing measures or a junction or a speed limit sign) were situated at a distance of 100 m from the primary countermeasure; i.e. in the car, the mean speed dropped from 65 km/h to 46.69 km/h.
26
U. ENGEI. and L. K. THOMSEN
I
65 KM/H I 1 I+100 M
f
50M
100 M
A
-+
I I JUNCTION “COUNTERMEASURE” (SECONDARY)
JUNCTION “COUNTERMEASURE” (SECONDARY)
CENTER OF “COUNTERMEASURE” (PRIMARY)
Fig. 9. Street plan for Example 1 before the implementation of a countermeasure. The change speed will be estimated for the position of the car, marked in the figure.
in mean
30 KM/H
Fig. 10. Street
plan for Example
1 after the implementation
of the countermeasure
30 KM/H
30 KM/H
I Fig. 11. Street
plan for Example
2 after the expansion
of countermeasure
Safety
effects
Table 6. Calculation Element no.
reducing
of speed change
Value
Element
Present
-4.680
Table 7. Calculation
(Example
27 1)
effect
km/h
of speed change
Value Present 15 cm 65 km/h 50 metres 50’ ” 100 ” 100 ” Present
measures
- 4.680
-0.6451 0.00376 km/h km/h x x 6550km/h m 0.0005352 km/h x 5O’m -148.32 km/h x LN(l + 11100) -81.50 km/h x LN(1 + 11100) 29.058 km/h Total
50 65 m km/h 50’ 100 m 100 m Present
Element no.
of speed
Element
(Example
= = = = =
-41.9315 0.1880 1.3380 -1.4758 -0.8110 29.068 -18.31 km/h
2)
effect
-4.680 km/h -1.001 km/h x 15 cm -0.6451 km/h x 65 km/h 0.00376 km/h x 50 metres 0.0005352 km/h x 50: metres -148.32 km/h x LN(l + l/100) -81.50 km/h x LN(l + 11100) 29.058 km/h Total
= = = = =
=
-4.680 - 15.015 -41.9315 0.1880 1.3380 - 1.4758 -0.8110 29.058 -33.33 km/h
With the countermeasure selected, the objective of a maximum speed of 30 km/h was not achieved. In the second example, the countermeasure was more restrictive by installation of a hump of height 15 cm. Example 2: Countermeasure: Narrowing of the carriageway (10) Countermeasure: Hump with the height 15 cm (7) Mean speed before: 65 km/h (2) Distance from the car to the primary countermeasure: 50 metres Distance from the primary countermeasure to the countermeasure (5) Distance
from the primary
countermeasures
to the countermeasure
(3) (4) before: after:
100 m
100 m (6)
The figures are repeated in Fig. 9 with the figures in parentheses referring to the figures in Table 5. Figure 11 illustrates the situation after the implementation of the hump. The results of the narrowing of the carriageway and adding a hump of a height of 15 cm produced an expected speed change of -33.33 km/h, from 65 km/h to 31.67 km/h. In this case a mean speed below 30 km/h was nearly obtained (see Table 7).
Acknowledgements-The start of this research project was facilitated by the Danish Technical Research Council with a donation of 130,000.- Danish Kroner in 1982. Many students and young candidates have been employed in the project. Special thanks are given to Erik Stern Nielsen and Morten Agerlin Petersen.
REFERENCES Engel. U.; Thomsen, L. K. $40 gaders sikkerhed. En analyse af politirapporterede ulykker gader baseret pa ulykkestztheder for hhv. efter aendring af gadestatus. Documentation Copenhagen: Danish Council of Road Safety Research; 1989a.
pB danske $40 Report 1189.
28
U. ENW
L
and L. K.
THOMStN
Engel. U.: Thomscn. L. K. $30 gadcrs sikkerhed. En analyse af politirapportercdc ulykker pa 14 WI gadcr og 52 kontrolgadcr baseret pa ulykkesfrekvenser for hhv. efter sndring af gadestatus. t~ocumt‘ntatlon Report 2i8Y. Copenhagen: Danish Council of Road Safety Research: 1YXYb. En&cl, U.; Thomsen. L. K. $40 gaders sikkerhed. En analyse af bilers korehastighcder pa 31 #-IO gadcr og 13 kontrolgader for hhv. efter zendring of gadestatus. Documentation Report 2!90. Copenhagen: Danish Council of Road Safety Research; 1990a. Engel, U; Thomsen. L. K. Effekter af fzerdselslovens 940. Anvendelse af $40 i Danmark samt effekter ph trafikulykker, risikoforhold og hastigheder; Sammenfatning. Summary Report 29. Copenhagen: Danish Council of Road Safety Research; 1990b.
