Open BRT
BRT, or bus rapid transit, can be done in one of two ways: closed and open. Closed systems imitate rail lines, in that there is a BRT route along the entire length of the corridor; open ones instead take a trunk route, upgrade it with dedicated lanes and other BRT features, and let routes run through from it to branches that are not so equipped, perhaps because there is less traffic on the branches. I complained 14 years ago that New York City Transit was planning closed BRT in the form of SBS on Hylan Boulevard on Staten Island, a good route for open BRT. Well, now the MTA is planning BRT on the disused North Shore Branch of the Staten Island Railway, arguing that it is better than reactivating rail service because buses could use it as an open corridor – except that this is a poor corridor for open BRT. This leads to the question: which corridors are good for open BRT to begin with?
Trunks and branches are good
Open BRT can be analogized to a Stadtbahn system, fast in the core and slow outside it. Like a Stadtbahn, it works best where several branches can converge onto a single route, where the high traffic both requires higher capacity and justifies higher investment; just as grade separation increases the throughput of a rail line, BRT treatments increase those of a bus through greater separation from other traffic and regularity of service.
Unlike a Stadtbahn, open BRT remains a bus. This means two things:
- The trunk route must itself be a strong surface route. It had better be a wide street with room for physically separated bus lanes, or else a city center route that could be turned into a transit mall. A Stadtbahn system puts the fast central portion underground and could do it independently of the street network, or even run under a slow narrow street like Tremont Street in Boston.
- The connections from the trunk route to the branches must themselves be strong bus links. If the bus needs to zigzag on narrow residential streets to get between two wider arterials, then it will be unreliable and slow even if one of the wider arterials gets dedicated lanes. A Stadtbahn system can tunnel a few hundred meters here and there to ensure the onramps are adequate, but a surface bus system cannot, not without driving its cost structure to that of a subway but with few of the benefits of underground running.
The North Shore Branch could pass a modified version of criterion 1, but fails criterion 2. In general, former rail lines are bad for such BRT systems, since the street network was never set up for such connections. In contrast, street networks with a central artery and streets of intermediate importance between it and residential side streets emanating from it, which were never used for grade-separated rail lines, are more ideal for this treatment.
Grids are bad
Street grids eliminate the branch hierarchy of traditional street networks. There is still a hierarchy of more and less important grid streets – in Manhattan, the avenues and two-way streets are wider and more used for traffic than the one-way streets – but there is little branching. Bus networks can still branch if they move between streets, which happens in Manhattan, but it’s not usually a good idea: Barcelona’s Nova Xarxa uses the grid to run mostly independent bus routes, each route mostly sticking to a grid arterial, and the extent of branching on the Brooklyn, Queens, and Bronx bus networks is limited to a handful of short segments like the Washington Bridge.
In situations like this, open BRT would not work. Hylan is possibly the only route in New York that has any business running open BRT. For this reason, our Brooklyn bus redesign proposal, and any work we could do for Queens, Manhattan, or the Bronx, eschews the open BRT concept. The buses are upgraded systemwide, since features like off-board fare collection and wider stop spacing are not really special BRT features but are rather normal in, for example, the urban German-speaking world. Center bus lanes are provided wherever there is need and room. There is more identification of a bus route with the street it runs on, but it isn’t really closed BRT, which is a series of treatments giving the BRT routes dedicated fleets and stations, for example with left-side doors to board from metro-style island platforms like Transmilenio.
What this means more broadly is that the open BRT is not a good fit for most of North America, with its grid routes. Occasionally, a diagonal street could act as a trunk if available, but this is uncommon. Broadway is famous for running diagonally to the Manhattan grid, but that’s not a BRT route but a subway route.
Pedestrian Observations 
North America is gridded, but I feel Open BRT could work in many of the Western U.S. cities that developed their road networks with wide trunks that and branching streetcars. For example:
In LA: on Venice, for ~4 miles west of Downtown, before opening to Venice / Pico / San Vicente Blvds.
In SF: on Market St, which is fed by both busses and streetcars (and the eastern-most stretch is quasi protected lanes already.)
In Seattle you had the bus tunnel…
Market Street already has subway-surface rail – it’s a great route for a trunk-and-branch system but is also such a strong corridor that it needs (and has) more than open BRT.
You highlight the problem with BRT in general. It works poorly on traditional urban street networks (gridded or not) with lots of intersecting streets, but that is where you want busses. If you have very wide highway like roads, the immediate urban environment is instead poorly suited for buses, and connectivity to the system is poor. In the latter cases, it is never going to be attractive transportation relative a car (north america) or a scooters (much of the global south).
In traditional urban cities it is better to focus on high quality bus lanes everywhere, and good bus stop design (benefiting the entire bus network), rather than think of BRTs as systems.
It can work in (poorly planned) large dense but car oriented cities (e.g. in South East Asia), where large highways are the major movers of people, whether you like it or not, and a transit like system along those highways still serve a need. But that is a solution to previous bad decisions.
Lots of intersecting streets are only a problem in the presence of idiocy. Which is to say, they aren’t a problem, they are an excuse with which stupidity tries to explain away its own failure.
In a traditional urban street network:
It can be argued whether the result of the above is truly a “BRT”, since noise and safety concerns (the presence of parallel pedestrians) limit the stopless speed of the buses below what would be possible on a highway or the middle of a grand boulevard.
Wouldn’t the sensible thing to do to be to run trains on it?
Light rail?
I have the same question, what is better in a closed BRT compared with a tram? In Barcelona we have a tram in the occasional diagonal avenue that can act as a trunk, and it’s an unqualified success in ridership (and an unqualified failure at the politics to complete it).
It feels like New York is being graded on a curve here.
On the North Shore Branch? Yes, this should be a Staten Island Railway line.
Trains vs Buses is not the question everyone thinks it is. Rubber tires have a max speed that is slower than steel wheels. Everything else you can do with a train I can do with a bus – though sometimes that would require me design a new bus. Buses mix with cars much better than trains and so are a great answer in low traffic areas where it isn’t economical to pay for two sets of infrastructure (you need roads everywhere for things like trash, so may as well share that cost with buses when there isn’t much demand). When your system gets busy rails are cheaper to maintain than a road lane and since you are busy that lane can’t be shared anyway to rails make sense.
As your roads (not to be confused with system, though they correlate) get busy, grade separated “lanes/rails” dedicated just to transit allows your transit system to be faster than general traffic and this speed is important to users of the system. It doesn’t matter if this is bus or rail, but rail is generally enough cheaper as to be worth it. Rail also already has full automation technology – there is no reason this can’t be put on a bus but there is a lot of engineering cost needed for this (this is only in the closed to all other traffic sections, on the open road self driving is not a solved problem) Rail already has off the shelf trains of whatever length you need, while a large bus is something you would have to design.
For the above reasons I think all large (as in area, not population) cities need to have grade separate rail across them. However you will still have more bus routes and more people on buses than anything else.
North of 59th Street in Manhattan, Broadway has both a major subway line and a major bus route (M104).
Yeah, but it’s not really good for branching buses – what would you branch off of it, a route on Columbus and Amsterdam?
Perhaps something like this?
https://www.mta.info/document/7546
Right.
I think that Newark, NJ would be a good location for open BRT. Its current bus network is extremely messy, but it looks like many of the strongest routes are radial. A core north south route could run from the intersection of Elizabeth Ave and Meeker Ave up Elizabeth Ave, Clinton Ave, Broad St, and Bloomfield Ave to the intersection of Bloomfield and Mt Prospect. I think there’s just enough room to squeeze in center-running bus lanes on the whole route. There would be 3 main northern branches (north on Bloomfield Ave, Mt. Prospect Ave, and Broadway) and 3 main southern branches (west on Clinton Ave, south on Elizabeth Ave, and east on Meeker and Haynes Aves to Newark Airport). All of these are pretty strong routes currently.
That happens outside of Sim City It’s a pity they weren’t planning for automobiles and buses in the 18th Century.
They’ve had it, in peak direction, on Market Street, downtown, for so long that no one notices it anymore. It works well when it’s enforced. Broad Street doesn’t need it as much because Broad Street is… broad.
Precisely because Broad St is broad it would be a good idea to set aside some lanes exclusively for busses and not have them get stuck behind slow cars.
I don’t think that the bus network’s messiness can be explained by the street grid. I understand that you’re going to have some quirks, but this network has things like a bus route (#27) with THREE different northern terminals. There is no official map, but an unofficial one exists here: https://www.dougandadrienne.info/njbus/indexnnj.html
There’s also a short section of bus route #76 where the northbound and southbound routes are THREE blocks apart, bus #1 sometimes takes detours to visit a jail or a warehouse and sometimes doesn’t– so much complexity for no reason.
They could truncate all the routes so your sensibilities aren’t offended and the people who use the once an hour bus can call a cab. You going to reimburse them?
London shows that if you’re willing to tell cars to stuff it with the secret technology of painted bus lanes, you can make your 18th century street layout have very high bus ridership.
There are bus lanes outside London. And they are often camera enforced.
Apparently he’s offended that the the far reaches of some lines have more than one terminal. It would be silly to paint bus lanes for a bus that comes once an hour.
The system is radial, once you get outside of the densest part of Downtown Newark the routes diverge and the traffic lightens. The peak direction bus lane is moderately self enforcing because there are so many buses it’s one long bus stop for blocks and blocks. It’s been that way for decades, the locals know about it and with some enforcement it works well.
I suppose NJTransit could abandon their system of the two digit number being the trunk and the letter being the obscure tendrils of interest to locals and give all the routes four digit numbers.
https://content.njtransit.com/sites/default/files/bus_schedules/T0025.pdf
Three western terminals with local, “GO” and rush hour express service to, to keep it simpler, two eastern terminals. Without parsing the schedule deeply I come up with the 25, GO25 and 375 becoming 18 different routes. Repeat for the routes with multiple terminals. Go ahead, give it a whirl
https://content.njtransit.com/sites/default/files/bus_schedules/T0001.pdf
The far reaches of some lines having multiple terminals is pretty unhinged. Call it the 27A and 27B if needed.
Why is it unhinged? Unless you think they should bulldoze a few blocks closer to downtown, for an enormous parking garage, so everybody can drive to the garage at the truncated end of the line.
The logic is that the main route is the number and the letter is the branch. The northbound 27B goes to Bloomfield. The northbound 27F goes to Forest Hill. The northbound 27N goes through Nutley to CliftoN. The 27R, a the short turn of the B that goes to BRanch BRook PaRk.
Quite useful if I’m at the Broad Street train station and want to get to City Hall. Or the Federal buildings. It’s any 13 whether it’s a 13T or a 13V. Or a 27 because all of the southbound 27s go to Irvington Terminal. On the way back it’s a 13 B,C,M or N and I don’t give a fuck what the letters stand for because I’m going to the train station not some obscure suburban destination. The same thing with the 27 and I’m not in the mood to check Google for the buses at City Hall that are also serve Broad Street.
I mean if they do have different letters that is pretty normal to be fair.
what do you think about Fulton st in Brooklyn as an open brt route, with DeKalb, Gates ad Halsey being the branches
It has the same problem as setting up BRT in Flushing and Jamaica – the town center loops are a very small share of the route’s length, and the branches are strong enough to merit bus lanes of their own.
I don’t understand why Nova Xarxa keeps appearing in discussions about BRT. It was a nice redesign from the old chaos of historical routes to a more rational and navigable network, but beyond that it has no BRT characteristics, and especially it is not rapid at all (the V15 rockets at a cap-flying commercial speed of 9,2 km/h!). The service is nice, accessible, good covering, frequent (on working days), but BRT it is not. The pro-car politician that cut the ribbon called it BRT in order to imply that “we don’t need the tram”, but that was just doublespeak.
It’s not a BRT network, it’s just a good bus redesign.
There is other downsides of open BRT that makes me not like it: it limits the frequency of the branches and makes a “knot” impossible.
If you are trying to get to the dense area that BRT is on it works, but if the route gets popular you have limited ability to upgrade frequency on that route because you have to get on the BRT when some other bus is already there, for a few this is okay, but eventually you get to congestion on the BRT. If you are sure you will never grow that busy (as would be the case in a small city) then go for it. However if the most optimistic projection says you outgrow the BRT you should just put in a train now – it will be about the same price today and scales much better in the future.
May trips in the suburbs are to other suburbs. The grocery store you want to go to is an hour walk, but only a few minutes by car. Many transit systems it would be 1.5 hours though because you have to go all the way downtown and then transfer to a different bus out. OpenBRT might let you transfer sooner, but there is no knot and so it is still a long wait for your transfer bus to come. If you bus a bus transit center where the buses get on the BRT and force everyone to transfer you make those other trips reasonable. (of course your train now needs to handle this busy time of the knot and will be empty between)
There is no easy answer. Most transit systems lack the money needed to give their less dense areas great service even though where such areas get great service they generate enough ridership to pay for it, and overall it is cheaper for everyone than their share of driving costs (both road and car).
The MTA could run Chiltern level service on its suburban/rural lines without spending too much money.
The approach to the Port Authority Bus Terminal would be a good open BRT project. First upgrade the Lincoln tunnel to 24/7 two lane bus only lanes. Then keep extending the lanes up the helix, maybe even begin constructing bus-only ramps from the various connecting roads.
Open BRT is maybe not so good for radial routes out of Jamaica and Flushing, but there are a lot of crosstown routes that head into those areas on common roads that could probably use it even if you extended the lines east. College Point > Main > Kissena in Flushing is a pretty obvious choice for one, as is Sutphin > Archer > Merrick.
For the Bronx you have buses converging on Fordham Road to Fordham Plaza, and to the 181st St bus terminal in Manhattan.