U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-RD-98-108
Date: February 1998
Capacity Analysis of Pedestrian and Bicycle Facilities
Recommended Procedures for the "Bicycles" Chapter of the Highway Capacity Manual
3. INTERRUPTED BICYCLE FACILITIES
This section focuses on operational analyses of interrupted bicycle facilities, including signalized and unsignalized on-street designated bicycle facilities with and without exclusive right-turn lanes for motor vehicle traffic. An example of a bicycle lane treatment at a signalized intersection having an exclusive right-turn lane is shown in Figure 4.
The concept of "control delay" is proposed as the service measure of effectiveness for interrupted bicycle facilities. Control delay is the portion of the total delay incurred by bicyclists passing through an intersection that is caused by the intersection traffic control, and includes initial deceleration, queue move-up time, stopped delay (i.e., the actual time stopped) and final acceleration delay. Control delay differs from total delay in that control delay does not include the delay caused by factors other than the intersection traffic control.
Delay is very important to bicyclists because bicyclists are completely exposed to the elements. Also, excessive delays to bicyclists on designated bicycle facilities may cause them to disregard traffic control devices or use alternate routes that are not intended for bicycle use. Once bicycle delay is determined, it can also be incorporated with vehicle and pedestrian delays to get a multimodal LOS for the intersection.
Only intersections with on-street bicycle facilities will be addressed in this document. It is acknowledged that interruptions exist between off-street facilities and crossing streets or other off-street facilities, but these types of intersections are not common in the United States and have not been extensively researched.
3.1 Signalized Intersections
A signalized intersection covered by these procedures is one where there is a designated on-street bicycle lane on at least one approach.
It is proposed that control delays be estimated from the uniform delay portion of the delay model in the Highway Capacity Manual (HCM)(TRB, 1994) as currently applied to motor vehicles for signalized intersections. The HCMcurrently uses stopped delay instead of control delay for signalized intersections, but that is likely to be changed to control delay (signal delay) in future editions.
The typical width of an on-street bicycle lane for which this recommended analysis applies is between 1.2 and 1.8 m (4 and 6 ft). A wide range of capacities and saturation flow rates have been reported around the world for these types of facilities. The ideal saturation flow rate may be as high as 2,600 bicycles/h of green, based on our observations of signalized intersections with significant bicycle traffic as described in the Research Reportfor this project (Rouphail et al., 1997). However, very few intersections provide ideal conditions for bicycles. Adjustment factors for less than ideal conditions have not been sufficiently documented to date to make any meaningful recommendations at this time. Until adjustment factors are developed, it is recommended that a saturation flow rate of 2,000 bicycles/h of green be used as an average value for most intersections.
Using a saturation flow rate of 2,000 bicycles/h of green assumes that right-turning motor vehicles yield the right of way to through-bicyclists as required by law. Aggressive right-turning traffic could reduce this value.
It is then recommended that the capacity of an on-street bicycle facility at a signalized intersection be computed as follows:
Control delay is then computed as follows:
The delay equation shown here differs slightly from the delay equation contained in Chapter 9 of the 1994 Update to the 1985 HCMbecause it computes control delay as opposed to stopped delay. This equation applies to both through and right-turning bicycles. It also applies to those left-turning bicycles making a "pedestrian style" left turn (i.e., in two stages, with bicycles traveling adjacent to the pedestrian crosswalks of the two intersecting streets). Advanced bicyclists who leave the bicycle lane and make left turns with motor vehicles are not covered by this procedure. Users of this procedure should also note that right-turning bicycles at intersections with heavy pedestrian flows will often experience additional delay depending on the configuration of the approach.
It is then recommended that the LOS be determined based on control delay, as shown in Table 6. These values are taken from the unsignalized chapter of the HCM. These are lower than the values in the signalized chapter for motor vehicles. However, lower delays are justified because bicycles are exposed to the elements.
TABLE 6 Level of Service (LOS) for interrupted bicycle facilities
At most signalized intersections, the only delay to through bicycles is caused by the signal itself because bicycles have the right of way over turning vehicles during the green phase. One possible exception is at signalized intersections, which force bicycles to weave with right-turning motor vehicle traffic on the intersection approach. This could cause additional delay to bicycle traffic at high motor vehicle volumes, although there is a lack of prior research in this area to confirm this effect. The research team was unable to effectively study the potential for a weaving effect due to a lack of suitable locations in the United States, as reported in the Research Reportfor this project (Rouphail et al., 1997). Therefore, at this time, it is impossible to make any recommendations as to the additional delay that may be caused by weaving-type configurations.
3.2 Unsignalized Intersections
An unsignalized intersection covered by these procedures is one where there is a designated on-street bicycle lane on at least one of the minor approaches.
The analysis procedures recommended for unsignalized intersections are for the minor approaches that are controlled by STOP signs. Bicycles on the major approaches are not delayed at most unsignalized intersections because they have the right of way over turning vehicles. One possible exception is at unsignalized intersections, which force bicycles to weave with right-turning motor vehicle traffic on the intersection approach. This could cause additional delay. However, at this time, it is impossible to make any recommendations as to the additional delay that may be caused by weaving-type configurations because of a lack of prior research in this area.
It is also assumed that bicycles on a minor approach turning right from one designated bicycle lane to another are not delayed because they do not have to wait for gaps in motor vehicle traffic. Experienced bicyclists making left turns from either the minor or major approach often leave the bicycle lane and queue with motor vehicles. It is impossible to make any recommendations as to the analysis of these types of left turns due to a lack of prior research in this area. Many bicyclists make "pedestrian style" left turns, which involve crossing the street twice. These bicycles are then effectively through-traveling-bicycles and should not be counted as left turns.
Very little research has been conducted regarding the evaluation of bicycle "critical gaps," and no research regarding "follow-up times" could be located for bicycles as used in the HCMfor computing control delay for motor vehicles at unsignalized intersections. Gap distributions have been reported by both Ferrara (1975) and Opiela et al. (1980) for bicycles crossing two-lane major streets. However, the research team is uncomfortable recommending the 3.2-s critical gap reported by Opiela et al. or a critical gap based on Ferrara's data because either would be much lower than the current critical gap used in the HCMfor motor vehicles in the same situation.
It is felt that the methodology currently used in the HCMfor motor vehicles at unsignalized intersections is also applicable to bicycles. Once critical gaps and follow-up times for bicycles are determined, it is recommended that the average control delay for bicycles be computed using the delay equation in Chapter 10 of the HCM. One caution deserves mention here. Bicycles differ from motor vehicles in that they normally do not queue linearly at a STOP sign. As a result, multiple bicycles often accept a single available gap. This fact will probably impact the determinations of bicycle follow-up times. Unfortunately, no prior research documenting and quantifying this behavior could be located.
It is then recommended that users determine the LOS based on control delay, as shown in Table 5, which is based on the values currently given in the HCMfor motor vehicles at unsignalized intersections. Due to a lack of prior research in this area, the research team cannot make any recommendations regarding delay and LOS for bicycles at all-way stop intersections at this time.