by Alex Sorton, Associate Director
Engineering Division, Traffic Institute, Northwestern University
INTRODUCTION
There is little empirical tested data for the development of a criteria to determine the bicycle compatibility of roadways. A method that holds great promise for the evaluation of roadways in urban and suburban areas is the bicycle stress level concept.
All bicyclists that ride on streets seek to minimize mental stress. They want to avoid conflict with motor vehicles and the strain of having to concentrate for long periods of riding along narrow, high speed,high volume roads. The bicycle stress level concept assigns values ranging from 1 to 5 were assigned to traffic variables of curb lane volume, speed and width. Level 1 indicates that the variables are not a problem for bicyclists, while level 5 suggests there are major problems. The highest and lowest stress levels were based on the review of the traffic engineering literature, the rational being that if conditions are bad for motorists, they will be worse for bicyclists.
The stress level concept was partially validated by a group of volunteer bicyclists who watched video taped segments of roadways showing a wide range of on street variables and rated them according to the stress level concept described previously.
The chart below relates the five bicycle stress levels to the types of bicyclists appropriate for each, based on their riding competency and preferred riding environment. It suggested that this procedure is not intended for use with bicyclists under the age of 10 years of age, who should only ride under supervision when or near the streets.
Table 1
Suggested Interpretation of Bicycling Stress Level
| Stress Level | Interpretation |
| 1 (very low) | Street is reasonably safe for all types of bicyclists (except for children under the age of 10) |
| 2 (low) | Street can accommodate experienced and casual bicyclists, and /or may need altering* or have compensating conditions** to accommodate youth bicyclists. |
| 3 (Moderate) | Streets can accommodate experienced bicyclists, and/or may need altering* or contains compensating conditions** to accommodate casual bicyclists. Not recommended for youth bicyclists. |
| 4 (High) | Streets may need altering* and/or have compensating conditions** to accommodate experienced bicyclists. Not recommended for casual or youth bicyclists. |
| 5 (Very High) | Streets may not be suitable for bicycle use. |
* "Altering" means that streets may be widened to include wide curb lane, paved shoulder addition, etc.
** "Compensating condition" can include street with wide curb lanes, paved shoulders, bike lanes, low curb lane volume, etc.
Determining the suitability of urban street segments for bicycle use, is based on the bicyclists perceptions of the on-street variables of curb lane traffic volume, speed of motor vehicles, and curb lane width. This is done by determining the stress level for each variable for the on-street segments. The stress level of the three variables are added and averaged to find the overall stress level. This is an approximation of how bicyclists perceive these variables while riding on the streets.
Curb Lane Volume vs. Stress Level
The quantity and character of motor vehicle traffic flow in the curb lane are primary determinants affecting bicycle compatibility.
Peak hour volumes are determined by the following relationship.
Curb lane volume is found by dividing the peak hour volume by the number of through lanes on the street segment. This assumes a 50/50 split on two-way streets. If the split is different, then use the known split for analysis.
Shown below is the suggested stress level for volume in the curb lane.
| Stress Level | Curb Lane Volume (veh/hr./lane) |
| 1 | = or <50 |
| 2 | 150 |
| 3 | 250 |
| 4 | 350 |
| 5 | = or >450 |
Curb lane Width vs. Stress Level
This variable is critical because it delimits the bicyclist's operating space. The suggested relationship between curb lane width and stress level is shown below.
| Stress Level | Curb Lane Width in meters (feet) |
| 1 | = or > 4.6 (= or >15) |
| 2 | 4.3 (14) |
| 3 | 4.0 (13) |
| 4 | 3.7 (12) |
| 5 | = or < 3.3 (= or > 11) |
Curb Lane Speed vs. Stress Level
The degree of the speed effect on bicyclists on curb lane depends on motor vehicle speed and size. The suggested stress level based on motorist speed is shown below.
| Stress Level | Speed of Motor Vehicles in kph (mph) |
| 1 | = or < 40 (= or <25) |
| 2 | 50 (30) |
| 3 | 60 (35) |
| 4 | 65 (40) |
| 5 | = or > 75 (= or > 45) |
The following example will explain the stress level concept as discussed
previously.
ANALYSIS
STRESS LEVEL
This street is not compatible for young and casual bicyclists. It may be compatible for experienced bicyclists. The street was restriped to provide for bicycle lanes.
RESTRIPED STREET
STRESS LEVEL
With the bicycle lanes, the street may be compatible with casual and
experienced bicyclists.
The initial validation of the stress level concept was done in Madison, Wisconsin where twenty three street segments were videotaped. These street segments represented the range of traffic variables that bicyclists encountered while riding on the Madison streets. The range of the variables were curb lane volumes of 60 to 670 vph, speeds of 40 kph (25 mph) to 65 kph (40 mph), and curb lane width of 3.3 to 5.5 meters (11 to 18 feet).
Several members of the Chicago Bicycle Federation were made aware of the bicycle stress level concept while attending a Bicycle Planning and Facility Design Workshop at the Traffic Institute at Northwestern University in Evanston, Illinois. The Chicago Bicycle Federation is composed of an alliance of bicycle club members, bicycle advocates and local representatives of the bicycling industry in the Chicago area. Their primary goal is to promote bicycling as a legitimate alternative mode of transportation for shorter distance trips in the Chicago metro area. A vital component in achieving this goal is to make sure streets are bicycle compatible. The bicycle stress level concept provides a means for identifying which streets are or are not bicycle compatible and may provide information on making them bicycle friendly.
The officer of the Chicago Bicycle Federation requested that their members be allowed to view the Madison videotape and rate the on-street variables. Since time was limited for taking the survey, only ten street segments were chosen for the rating process. The range of these variables was similar to the range mentioned previously. The bicyclists were asked to view video clips of the selected street segments and then rate the three variables of vehicular volume, curb lane width and speed of motor vehicles based on their perceptions via stress level concept. After rating the three variables, they were asked to indicate whether they would or would not ride their bicycle on each of the segments. If they indicated they would ride on this street, they were then asked what distance they would ride their bicycle on that street.
Prior to watching the video clips, participants were instructed to indicate how comfortable they would feel riding their bicycle with each on-street variable which they would see in the clips. They were told that a 1 indicates they are very comfortable riding with this variable condition and a 5 indicates they would not want to ride with this variable condition under any circumstances. They were further instructed to rate the specific variables condition between 2 to 4 for conditions they felt did not meet the extremes.
There were 43 adult bicyclists that took part in the rating evaluation.
Fifteen of the participants were casual bicyclists and twenty-eight were
experienced bicyclists.
Survey Results
The earlier Madison survey results indicated good correlation between the bicyclists perceptions of stress level for volume and speed. That the bicyclists stress level increased level increased with increase of volume and the increase of speed of motor vehicles. The perception of stress level for width did not have a good correlation. While the stress level of the bicyclists did increase when curb lane widths decreased, but the relationship was very inconsistent for all types of bicyclists.
Therefore, in the Chicago survey an attempt was made to determine why there was not good correlation between curb lane width and stress level in Madison.
In the Chicago survey three of the street segments had a curb lane width
of 3.9 meter (13 feet) with curb lane volumes ranging from 150 to 650 vph
and motor vehicle speed ranges of 50 to 65 kph (30 to 40 mph). Figure 1
shows the results of volume in the curb lane versus stress level for curb
lanes 3.9 meters (13 feet) wide, as the volume increase from low to high
the stress level increases accordingly.
Figure 2 shows similar results for speeds of motor vehicles in curb lanes 3.9 meters (13 feet) wide, as speed increase so do the stress levels of bicyclists. This seems to indicates that the bicyclists' perception of the stress level of curb lane width is dependant on the volume and speed of motor vehicles. Figures 1 and 2 also indicate that casual and experienced bicyclists perception of the roadway variables are different. The casual bicycles stress level for curb lane widths is about one stress level higher than experienced bicyclists. Further research should be done to determine whether these results are duplicated for other curb lane widths.
The bicyclists after viewing each segment and rating the three variables
were asked whether they would ride on a particular roadway segment. Table
1 and 2 show the results to this question. The proposed overall stress level
of each street segment was used to determine whether the casual and experienced
bicyclist would or would not ride their bicycle on a particular segment.
The proposed overall stress level is found by finding the average of the
three individual variables for each street variables. The proposed overall
stress level is based on the concept reviewed earlier. It was used to compare
the differences in perception between the casual and experienced bicyclists.
It has a common base in which to evaluate the two types of bicyclists.
Table 1: Variable Attributes vs Overall Stress Level Casual Bicyclists
| Volume (vphpl) |
Speed (mph) |
Width (feet) |
Overall S.L. Casual |
Overall S.L. Proposed |
% Will not ride |
| No bike lane | |||||
| 350 | 35 | 13' | 3.8 | 3.3 | 36 |
| 450 | 35 | 13' | 3.8 | 4.0 | 36 |
| 650 | 40 | 13' | 4.3 | 4.3 | 64 |
| Bike lane | |||||
| 500 | 30 | 12' + 8' bl | 1.6 | 2.6 | 0 |
| 750 | 40 | 12' + 8' bl | 1.6 | 3.3 | 0 |
Table 2 Variable Attributes vs. Overall Stress Level Experienced Bicyclists
| Volume (vphpl) |
Speed (mph) |
Width (feet) |
Overall S.L. Casual |
Overall S.L. Proposed |
% Will not ride |
| No bike lane | |||||
| 350 | 35 | 13' | 2.9 | 3.3 | 14 |
| 450 | 35 | 13' | 3.1 | 4.0 | 21 |
| 650 | 40 | 13' | 3.8 | 4.3 | 32 |
| Bike lane | |||||
| 500 | 30 | 12' + 8' bl | 1.5 | 2.6 | 0 |
| 750 | 40 | 12' + 8' bl | 1.6 | 3.3 | 0 |
Table 1 indicates that 36 % of the casual bicyclists indicated they would not ride their bicycles when the proposed overall stress level was between 3.3 and 4.0. When the proposed overall stress level was 4.3, 64 % of the casual bicyclist indicated they would not ride their bicycle on this street segment. Table 2 indicates that 14 % of the experienced bicyclist indicated they would not ride their bicycles on a street segment when the proposed overall stress level was 3.3. When the proposed stress level was 4.0 and 4.3 then 21 % and 32 % of the experienced bicyclist indicated they would not ride their bicycles on these streets segments.
Table 1 and 2 also shows that when a 2.4 meter (8 foot) bicycle lane exists on streets that have high traffic volumes and fast motor vehicles both casual and experienced bicyclist rated the overall stress level about 1 to 1.6 lower than the proposed overall stress level. The differences between the street segments with or without bike lanes in tables 1 and 2 is that the curb lane width without a bike lane is 3.9 meters (13 feet) wide, whereas, the street with a bike lane the curb lane is 6.1 meters (20 feet) wide and has been striped with a 2.4 meter (8 feet) bike lane. This seems to indicates that the stress level perception of bicyclists of street with bicycles lanes is much lower than for streets without bicycle lanes. Further all bicyclists in both groups indicated they would ride on the street segments with bicycle lanes even when the volumes were as high and speed were as fast as on the previous segments without bicycle lanes. This seems to indicate that streets with bicycle lanes will attract casual and experienced bicyclists, where as roadways with similar characteristics but without bicycle lanes may not attract some of the same bicyclists.
The effects of bicycle lanes on the perceived overall stress level of bicyclists must be studied in more detail. Further studies should be carried out to determine whether bicycle lanes less than 2.4 meters (8 feet) wide or roadways with paved shoulders with an edge stripe will show similar results.
It is the author's opinion that caution should be used in using the results
of this limited research. The studies conducted so far only observed the
effects of traffic on bicyclists at mid block locations. The midblock location
may not indicate how easy or difficult it is for bicyclists using the streets
do to conflicts with motor vehicles. Intersections may present more challenging
problems for bicyclists due to turning traffic, exclusive right turn lanes
and other factors.
Figure 3 shows the relationship between the distance the casual and experienced
bicyclists would ride on street segments with different overall stress levels.
This figure indicates that on streets where bicyclists perceive the overall
stress level increasing they will ride shorter distances. It must be remembered
that the overall stress level is based upon the three individual variable
of curb lane volume, width, and speed. That is, as the volumes increase,
widths decrease, and speeds increase the bicyclists perceive the overall
stress level increase and either will not ride on these streets or will
ride a shorter distance on them. Again, the experienced bicyclists has a
higher tolerance when compared to the casual bicyclists.
Lastly, the bicyclists were asked to rank each of the three variables in order of importance to them in reaching a decision to ride on a given street. They were told they could rank one, two or all three variable as equal in importance. They were told a three indicates that the variable was the most important, a two it was moderately important, and a one that the variable was the least important. The casual bicyclists rated all three variables almost of equal importance. While the experienced bicyclists rated the width variables as the most important variable with volume and speed being the second most important variables. Several of the experienced bicyclists wrote on their questionnaires that width of the curb lane is the most important variable and that a wider curb lane compensates for higher volumes and speeds.
These result seem reasonable. The casual bicyclists probably does not ride on busy streets very much and thus rating all three variables equally seems reasonable to them. The experienced bicyclist ride their bicycles on busy street and based on their experience of riding on these streets realize that width is the most important variable.
Comparison of the Madison and Chicago Bicyclists
Figures 4, 5, and 6 show the difference in perceptions of the bicyclists
in Madison, Wisconsin and Chicago, Illinois for the three variables. At
the lower volume and speed ranges there is about one stress level difference
between the two groups. At the higher values for volume and speed there
is little difference. The curb lane width variable indicated about one half
a stress level difference at the wider and mid range widths and little difference
for the narrower widths.
These figures show that the Chicago bicyclist have higher tolerances for the on-street variables as compared to the Madison bicyclists. This is not unexpected because larger cities usually tend to have more traffic for longer time periods on their streets in comparison to smaller cities. Thus, bicyclists in larger cities would expect to have higher tolerance levels than bicyclists in smaller cities or towns.
REFERENCES:
1. Transportation Research Board, Special Report 209 Highway Capacity Manual, Washington, D.C. 1985
2. New Jersey Department of Transportation, Bicycle Compatible Roadways: Planning and Design Guidelines, Trenton, N.J., 1982
3. McHenry, S.R., Evaluation of Wide Curb Lanes As Shared Lane Bicycle Facilities, Maryland Department of Transportation, 1984
4. Victoria Ministry of Transportation, Planning and Design of Bicycle Facilities, Melbourne, Australia, 1988
5. Van Valkenberg, P., Methodology for Evaluating the Suitability of Two-Lane Two-way Paved Rural Roads for Shared Bicycle/Motor Vehicle Use, Wisconsin Division of Tourism, Madison, Wisconsin 1982 (Unpublished Report)
6. Morrall, J.F., Passing Lane Research Study for the Trans Canada Highway in Banff National Park - Phase II, Transportation Division of Engineering and Architects, University of Calgary, 1984
7. Wisconsin Department of Transportation, Guidelines for Developing Rural Bike Routes, Madison, Wisconsin 1975
8. Geelong Bikeplan, Geelong, Australia 1978
9. Harwood, D.W., Effective Utilization of Street Width on Urban Arterials, NCHRP Report 330, Transportation Research Board 1990.
10. Guideline for the Control of Direct Access to Arterial Highways, Federal Highway Administration, Report No. FHWA-RD-
76-86, 1975.
11. Bicycle Planing and Facility Design Worksop, Traffic Institute Northwestern University, 1992.
[Editor's note: After publishing this paper in 1995, the author revisited the differences between the Madison and Chicago data in light of the cyclists' answers to other questions. It is now his opinion that the Madison "experienced" bicyclists were more like the "casual" bicyclists than previously thought. He bases this conclusion on their stated propensity to ride on sidepaths and sidewalks.]
*Thanks to Alex Sorton, deputy director of the Traffic Institute for allowing us to use the above.