
Bus stop spacing has a major impact on transit performance. Stop spacing affects both access time and line-haul time, and therefore affects the demand for transit service. In general, there is a tradeoff between: (a) closely spaced, frequent stops and shorter walking distance, but more time on the vehicle and (b) stops spaced further apart and longer walking distance, but less time on the vehicle.
Some of the findings of optimization studies are:
Although analytical studies to determine optimal stop spacing provide some useful guidelines, stops must ultimately be sited to serve major trip generators and attractors in the service area. To the degree that a BRT is emulating a light rail system, a useful benchmark is the actual stop spacing of LRT systems in the U.S. (see table).
| Mean distance between stops | ||
|---|---|---|
| System | m | ft. |
| Memphis (1) | 284 | 931 |
| Philadelphia (2) | 500 | 1639 |
| Denver | 566 | 1856 |
| Newark | 627 | 2056 |
| Cleveland | 651 | 2134 |
| Buffalo | 714 | 2341 |
| Sacramento | 964 | 3161 |
| Portland | 1052 | 3449 |
| Salt Lake City (3) | 1142 | 3744 |
| San Jose | 1181 | 3872 |
| Baltimore | 1265 | 4148 |
| Pittsburgh | 1323 | 4338 |
| Boston (4) | 1488 | 4879 |
| Dallas | 1523 | 4993 |
| St Louis | 1526 | 5003 |
| San Diego | 1664 | 5456 |
| Los Angeles | 1742 | 5711 |
| Source: Janes Urban Transport Systems, 1999-2000 edition. Notes: 1. System serves a downtown distributor function only. 2. Computed for tunnel section only. 3. Under construction, based on planned system. 4. Average of subway (Metro) and LRT combined. |
||
Over time there is a tendency for additional stops to be added to bus routes, as requests for service in front of more places are accepted. When stops are as frequent as every or every other city block, it may be useful to comprehensively re-examine the location of all stops. In addition to reducing the number of stops, citing stops so as to improve service (as discussed below) can be a component of a BRT project.
Limited-stop service is used frequently on high-demand bus corridors in combination with local service. BRT projects can also involve adding limited-stop service, perhaps overlayed over existing service.
For BRT systems which include (a) bus detection and active signal priority or (b) queue jumper lanes, bus stops should be at the far side. This permits effective use of these priority measures to clear the bus through the intersection with minimal delay. Otherwise, the added bus dwell time variability from a near side stop would complicate, if not preclude, bus detection and green phase extension.
A near side stop would also prevent effective use of a queue jumper lane (with or without an advanced bus signal) since the adjacent queue of through traffic would already be discharging from the stop line by the time the bus was ready to depart from the near side stop. Instead of having a "jump" on the queue of traffic in the adjacent through lanes, the bus would have to merge with it. There would be very few gaps of adequate size because of the compressed queue of traffic discharging from the intersection at a saturation rate of flow. The bus would experience a delay equal to the time for the queue to clear the intersection, or the sum of this clearance time and the cross traffic green time if the bus is forced to wait to the next signal cycle.
| Stop Type |
Advantages |
Disadvantages |
|---|---|---|
|
Near Side |
|
|
|
Far Side |
|
|
|
Mid block |
|
|
| Source: Table A-4, Appendix A, TCRP, original source: K. Fitzpatrick et al., Guidelines for Planning, Designing, and Operating Bus-Related Street Improvements. FHWA/TX-90/1225-2F, Texas Transportation Institute, College Station, TX. August 1990. | ||
One can calculate the bus merge delay upon reentering a traffic stream as a function of both the adjacent lane’s traffic flow and the critical gap length needed by the bus operator to merge. The point with significant delay is above 450 vehicles per hour per lane (vphl) (TCRP, p. D-43). For a four mile trip, the cumulative delay can be in excess of ten minutes.
"(1) The driver of a vehicle shall yield the right of way to a transit vehicle traveling in the same direction that has signalled and is reentering the traffic flow.The Florida version of this rule applies only to buses stopped at "a specifically designated pullout bay." In British Columbia, the rule applies only on roadways with a speed limit of 60 km/h (37 mph) or less and also specifies that a driver must only yield when it is safe to do so. The introduction of the rule in May 1999 was accompanied by a "Yield to Bus" public awareness campaign.
(2) Nothing in this section shall operate to relieve the driver of a transit vehicle from the duty to drive with due regard for the safety of all persons using the roadway."
In addition to a reduction in bus merging delay at each stop, other significant benefits of a priority merge rule include: reduced waiting times for passengers at bus stops due to reduced irregularity of the service, decreased travel time for passengers, less stress on bus operators, and less impact on bus operations due to traffic congestion. A priority merge rule is consistent with providing a Bus Rapid Transit service.
Bus bulbs are a section of sidewalk that extends from the curb of a parking lane to the edge of the through lane. When used as a bus stop, the buses stop in the traffic lane instead of moving into the parking lane.

Adding a bus bulb permits installing a bus shelter even on a narrow sidewalk.
Advantages of a bus bulb include:

With a bus bulb, buses stop in the travel lane and therefore there is no delay in re-entering the traffic stream.
Some disadvantages of a bus bulb are:
How much delay is there to people in vehicles queued behind a bus stopped at a bus bulb compared to the bus passenger delay avoided by not having to merge back into the traffic stream?
For an indication of the tradeoff between the delay imposed on persons in a queue of vehicles behind a stopped bus versus the person- seconds of delay avoided by avoiding a bus merge maneuver, consider the following example. Assume an average passenger vehicle occupancy of 1.1, and a discharge headway of 2.5 seconds per vehicle. For a traffic flow of 700 vphl, and a critical gap size of 10 seconds, the bus merge delay is 24.2 seconds. Assuming a net departure bus occupancy of 30 persons after loading at the bus bulb during a dwell time of 20 seconds, the person-seconds of delay avoided equals (24.2) * (30) = 726 seconds or 12.1 minutes. The delay to the queued vehicles is 42.7 person-seconds The net gain or reduction in person delay is therefore (-726)+(42.7) = -683.3 seconds or 11.4 minutes saved.
There are several manufacturers who specialize in specialized, modular shelters. Whether adapting a manufactured shelter or using a custom design, some general factors to consider include:
TCRP. Transit Cooperative Research Program. Report 19, Guidelines for the Location and Design of Bus Stops. Transportation Research Board. National Academy Press. Washington, DC, 1996.