Designing a BRT on Flatbush is complicated
The details will make it or break it
by Annie Weinstock and Walter Hook
In July, NYCDOT released bold designs for central median bus lanes on Flatbush Avenue between Livingston and Grand Army Plaza. For many reasons that we have detailed out before, we wholeheartedly support DOT moving forward with the central-median option.
In our recent work on Where NYC Should Build BRT, Flatbush Avenue ranks as the 7th most promising BRT corridor in NYC. While the NYC DOT’s claim of 132,000 daily bus riders is exaggerated (we estimate fewer than 25,000), Flatbush is still on the short list for full BRT.
However, the current design still doesn’t reach the minimum criteria to be considered a BRT. It also has some critical design issues which are likely to result in massive amounts of delay for users of the corridor. It is incredibly important that the new corridor be built as an actual BRT and that the design problems get resolved, or the opportunity on this critical corridor will be lost. Further, it could be a make or break for BRT in NYC.
In this piece, we review the critical areas which, if not addressed, will undermine the benefits to the bus passengers while causing significant additional delay to motorists. In our next piece, we will take a more visionary stance, assuming that DOT takes their time and gets these immediate details right.
The current design does little to speed boarding
One of the main benefits of SBS has been all-door boarding. At-level boarding, which has never been tried in NYC, would save even more time. So far, there is no indication that Flatbush bus stops will have all-door or at-level boarding.
During the PM peak hour, there are approximately 12 passengers boarding each bus at Atlantic Center. Each boarding takes an average of 3 seconds. The occasional person with a wheelchair or stroller needs much longer. We recommend a full BRT station at Atlantic Center with all-door, at-level boarding. Doing so would reduce average travel times by 30 seconds per trip at a minimum and would improve reliability. Doing so at all stops in this corridor would further reduce delay, and create a cohesive sense of corridor. The current plan does not address this delay.
NYC DOT made necessary compromises on stop placement… but adds stops to the B41-Ltd
Two bus routes will use the full section of Flatbush being designed — the B41-local and the B41-Ltd — while the B45 and B67 will use a portion of this corridor. The B41-local makes seven stops between Grand Army Plaza and Atlantic, while the B41-Ltd goes direct without making any stops.
To allow all bus routes to benefit from the new center bus lanes, the stopping pattern on all these bus routes needs to be the same.¹ Thus, DOT made a compromise, wherein there are only two stops between Plaza St. and Atlantic: one at Bergen and one at Park Pl. The result is that B41-local passengers save time by stopping at fewer stations, while B41-Ltd passengers lose time by stopping at more. B45 passengers have no change and B67 passengers stop at one less station.
As Bergen Street and Atlantic Center are very close together, we suggest eliminating the Bergen Street station. This would further speed up the trip for bus passengers and would win back some space, as discussed later.
The design only partially resolves one of the most critical areas of bus delay
All of the bus routes using the northern section of Flatbush turn onto Livingston St, use the Livingston St busway, and terminate in Downtown Brooklyn. Thus, the Flatbush project must ensure not only that buses can travel smoothly on Flatbush, but also that the connection to the Livingston Busway is smooth. That connection is one of the worst congestion points on all of Flatbush.
The crux of this problem is caused by two problematic left turns (buses and mixed traffic) at the northern end of the Flatbush corridor:
- Northbound Flatbush onto Livingston (the point at which all bus routes connect between the Flatbush and Livingston busways)
- Southbound Flatbush onto Lafayette (an opposing nearby left turn that creates needless delay at both locations)
The turn lanes for both are too short to handle the turning volumes, meaning that they back up and block the next lane over. The NYC DOT design lengthens both of these turning lanes, reducing the risk that the turning queue will back up into the mixed traffic lanes. This is an improvement.
However, this doesn’t fully resolve the problems at these intersections. Buses today often get stuck trying to make the northbound left onto Livingston. Both mixed traffic and buses must enter Livingston together in a single shared lane before the buses can proceed to the busway. Because this shared lane frequently backs up, buses often cannot enter the Livingston busway. As a result, it sometimes takes more than one signal cycle for the buses and mixed traffic to make the left turn, causing a major bottleneck on Flatbush. The design doesn’t solve this.
The best way to solve this would be to eliminate the mixed traffic left turn onto Livingston, maintaining it for buses only. The easternmost block of Livingston can then become a bus-only block. This would solve the problem currently caused by so many vehicles trying to enter the narrow first block of Livingston. There are only minimal drivers with destinations on this very block and there are other ways of accessing the blocks to the west.
As we discussed in our recent report, the bus stop on the next block of Livingston is delaying both mixed traffic and buses on Livingston. If it were moved to this new bus-only block, that could alleviate these problems, especially if it were upgraded to a full BRT station.
In the case of the southbound left onto Lafayette, the dedicated left turning signal phase adds 20–30 seconds of delay for all vehicles at that intersection. This includes northbound buses on Flatbush which must stop during the left turn phase. This hasn’t been changed in the new design.
Moreover, by retaining and lengthening this turn lane in the new designs, there is no longer any room for a northbound dedicated bus lane on the last block of Flatbush before Livingston. This increases the risk that backups of northbound left turns onto Livingston will block the bus lane on the previous block.
These problems can be solved by simply removing the southbound left turn onto Lafayette. This Fort Greene-bound traffic can use Fulton instead, turning right onto Rockland to reach nearly the same point.
NYC DOT is dramatically reducing lane capacity which is likely to cause new problems
Mixed traffic on Flatbush frequently backs up in both directions from Livingston all the way to 4th Avenue. Flatbush is an important trucking route and, because it is not part of the Brooklyn street grid, there are few alternate routes for trucks bound for Manhattan or the BQE northbound.
Additionally, as we all know well, NYC drivers have no compunction about blocking intersections when they have been waiting a long time to reach the intersection. This further delays bus turning movements, particularly at Livingston Street. Problematically, the NYC DOT design, which decreases the number of traffic lanes throughout much of the corridor, increases the risk of traffic congestion, slowing trucking movements and increasing intersection blockages, in the stretch between 4th Avenue and Livingston.
We generally applaud the radical nature of dramatically reducing mixed traffic capacity, but it is being reduced more than is necessary, increasing the risk of needlessly antagonizing truckers and drivers.
- The southbound approach to 4th Avenue
Currently, southbound Flatbush has two through and one right turn-only lane approaching 4th Avenue. The right-turn lane frequently backs up because the block between Flatbush and Atlantic is extremely short, and many trucks are making this turn.
The NYC DOT design retains this dedicated right turn but reduces the mixed traffic through lanes from two to one. This is likely to exacerbate the bottleneck in the southbound direction, angering motorists and truckers AND delaying buses that can’t turn through the blocked upstream intersection at Livingston.
It is critical that NYC DOT maintain two southbound mixed traffic lanes and the right turn-only lane here. The presence of two new bus lanes makes this difficult, but not impossible. The lanes can be narrowed, and a few feet of sidewalk space could be taken from Flatbush at the approach to the Apple Store at Ashland. However, this loss can be more than offset by creating additional new sidewalk space on Ashland, which is not a through street and carries very limited traffic (see image below).
Adding a second left turn lane on 4th Avenue between Flatbush and Atlantic should also be considered to further mitigate the risk of this area backing up, especially because so many trucks are making this movement.
- Flatbush northbound between Atlantic and 4th Ave
Flatbush northbound between Atlantic and 4th Ave has also been reduced to one lane in the designs. There is a lot of traffic, including trucks, turning right from Atlantic onto Flatbush, plus one bus route (B45) that will be adversely affected if this turn backs up significantly.
- Flatbush south of Atlantic
South of Atlantic, mixed traffic is also reduced to one lane in the proposed design. While traffic, including trucks, is not as heavy here, it is still worth studying whether it might be possible to reduce the number of lanes to two instead of one. This could potentially be done by eliminating the Bergen St. station and carefully placing the Park Pl. station by shaving space away from the two triangles at Park Pl. This possibility should at least be studied.
Bus stop placement around Atlantic needs more consideration
With four bus routes using the section of Flatbush north of Atlantic, special care is needed to ensure that the stations are long enough to handle the volume of buses stopping during the peak period. Without doing this, buses may queue up behind the station, and even back up into the intersection, while waiting to approach the platform. This has presented serious problems in other cities where insufficient consideration was given to station sizing.
NYC DOT placed the southbound Atlantic Center station on the west side of the street between Atlantic and Ashland. The station length is limited by the fact that the block is shorter on the west side of the street than on the east side. Moving this station to the east side, and turning it into a northbound station, would allow for a longer station, reducing the risk of buses backing up at the station.
The southbound station would then be located in front of Barclay’s Center between Atlantic and Pacific, where the currently-proposed northbound station is located.
The details matter, especially for NYC’s first real attempt at BRT
As we’ve presented here, there are many details that need to be considered for this project, and any one of them done wrong could create significant problems.
NYC DOT should be applauded for putting such a bold plan on the table. However, the boldness is not always so strategically placed, and meanwhile they have chickened out on measures like removing the left turns that should be a win-win for mixed traffic and bus passengers. All of this requires in-depth analysis and should not be rushed.
The MTA just put $166M towards the design of the Interborough Express. The Flatbush bus lane design was done for peanuts in just a few months. Rushing forward with this plan without more careful analysis risks discrediting the very notion of BRT in NYC. We are all for the abundance agenda — but not at the expense of a well-functioning project. NYC DOT should slow down, at least enough to get these details right. This is an historic opportunity to build the first real BRT in NYC, and it should be given its full due.
¹In the world’s best BRTs, bus-only passing lanes at stations allow for varied stopping patterns, but this is not the case on Flatbush.
