the explosive global growth of bus rapid transit (BRT)

recent study from ITDP  surveys the growth of BRT around the world over the past decade.  

BRT Infographic


Note that IDTP thinks of BRT as something that matches the performance of rail using buses.   ITDP's BRT standard excludes many of the projects that the US Federal Transit Administration calls BRT, which amount to premium buses in mixed traffic with minimal speed and reliability features.*  

China has created the largest quantity of true BRT systems, but of course in per capita terms it's Latin America that is building true BRT most intensively.  Fast-developing middle-wealth countries like China, India, Mexico, and Brazil are the sweet spot for BRT because (a) car ownership is still moderate, (b) government power tends to be consolidated enough that decision making is easy, (c) there is simply not enough money to build massive rail transit systems, at least not quickly and at the necessary scale.  

This news is also interesting in light of the forthcoming Rio de Janeiro conference on climate change, and the rumours that China may be ready to commit to reducing emissions, putting pressure on India to do the same.  Latin America, where many countries of similar wealth already have relatively strong climate change policies, is the perfect site for this conversation.

The other interesting stat is how rapidly the BRT revolution has moved.  Of all the true BRT in the world, 75%  was built in the last decade, mostly in middle-income countries, and the pace shows no signs of abating.

Fortunately, those middle income countries amount to a big share of the world, which could mean a real impact on global transportation impacts over time.


* (I tend to agree with ITDP's concern that the overly weak use of the term BRT is making it hard to talk about the original point of the BRT idea, which was to mimic what rail rapid transit does in terms of speed, frequency, and reliability.  This meaning is inherent in the "R" in BRT, which means "rapid".)

25 Responses to the explosive global growth of bus rapid transit (BRT)

  1. M1EK December 3, 2014 at 10:55 am #

    I think it’s not complete to talk about the reasons BRT works in those countries without mentioning the much lower cost of labor (both in absolute pay terms and in things like unionization and labor rules).

  2. Kantor57 December 3, 2014 at 11:54 am #

    My impression is that BRT is a good third world solution for third world countries; not so good elsewhere because of labor cost and politics.
    In general BRT is trying to fit a square peg in a round hole: busses are very good at being busses (more flexibility of the routes, more capability of going around obstacles and so on..) and rail is very good at being rail (speed, capacity and much more confort).

  3. Shane Phillips December 3, 2014 at 12:28 pm #

    The above points about labor are important to consider, though I don’t believe that completely disqualifies use of BRT in the U.S. and other high-cost countries.
    On a related note, in the case of these middle-income countries, the great thing about true BRT is that the dedicated right of way can eventually be converted to rail, relatively cheaply and easily, when and if the ridership/economics favor it.

  4. Goosoid December 3, 2014 at 1:42 pm #

    Well here in Auckland, New Zealand (certainly not a low wage, third world country) we have had a top quality BRT system (the Northern Busway,_Auckland) for 6 years and it has been a phenomenal success.
    It is already exceeding 2020 targets and the only problem is that it is working so well, people are being left behind at stops by full buses. Buses are running at 2-3 minute intervals at peak times. It has really changed the transport picture in the north of the city.
    I know it is easier to believe there is something special about your city that means change is not possible, but there really isn’t. BRT is an awesome and cheap way to get public transport going. Our busway was graded for light rail so it can easily be upgraded to LRT when the capacity is at maximum.

  5. EngineerScotty December 3, 2014 at 2:28 pm #

    Worth emphasizing again, but a quick summary of the advantages of BRT that do not apply to rail. (Rail, of course, has its own advantages, which I don’t list here).
    * Potentially cheap to install, particularly if you can reuse existing pavement for an exclusive lane (i.e. taking it from traffic).
    * As buses can run in mixed traffic, it is harder for an incident on a BRT line to shut down the line or require external bridging or portage. (Exceptions might be made for grade-separated BRT lines that have few points where vehicles can enter or leave the system).
    * As buses can run in mixed traffic, it is far easier to “phase” BRT projects without disrupting service or requiring transfers. (Replacing half a bus line with rail is a good way to piss off riders at the end of the line, as they will need to transfer; assuming oddball buses aren’t sued, a local bus simply shifts to the exclusive right-of-way as it is completed).
    * As buses can run in mixed traffic, branching topologies (“open BRT”) are possible.
    * As buses can far more easily pass each other than rail can (without passing tracks/sidings), it is easier to mix express or skip-stop service along a BRT route with local service.
    * As BRT can run in mixed traffic, BRT projects can avoid building expensive structures to cross things like rivers or other obstacles–instead, the road network is used. (Of course, doing this often implies requiring the BRT line share a bottleneck with general traffic, which can dissolve the advantages of building BRT in the first place).

  6. al December 3, 2014 at 2:38 pm #

    With advances in vehicle design, and vehicle automation, one could see road train type BRT in 1st World countries as replacements for Light Rail. These vehicles could be turbo diesel, CNG, diesel-Hybrid and or electric powertrains. One could see a system with a human diver/monitor on the lead vehicle with robot buses following close behind. The congested sections could have dedicated separated lanes with trolley wires overhead. Outside the congested corridors, the trailing buses could branch off using the diesel gensets.

  7. Medes December 3, 2014 at 8:11 pm #

    My problem with BRT is that almost all of its “advantages” are things that reduce speed and efficiency (eg mixed traffic running on certain segments). I see BRT as a great alternative to rail (with the same infrastructure) when ridership does not warrant rail’s high-capacity. Sadly however whenever these systems are built, planners jump at the opportunity to save costs by slowing the buses down. A good example is the brand new sbX in San Bernadino, where the Buses are subject to traffic conditions for half the route and have to wait for lights just like everyone else.

  8. M1EK December 4, 2014 at 5:24 am #

    “the great thing about true BRT is that the dedicated right of way can eventually be converted to rail, relatively cheaply and easily,”
    No, this is far from true. It’s actually very difficult – both technically and politically. Technically, you have to end bus service on the guideway for a long period of time while installing the rails. Politically, there will be resistance to a second investment in the same corridor instead of in other corridors.
    I think the “BRT can convert into rail” talking point is practically a con job at this point. It presents more obstacles for rail than help.

  9. Carsinogenic December 4, 2014 at 6:21 am #

    it would be great to see along side this chart, data on automobiles
    In Canada. according to Stats Canada we spend per year on the social costs of driving close to $200 BILLION per year

  10. Neil December 4, 2014 at 9:06 am #

    Labour costs have an impact on the economics of BRT vs Rail for high capacity routes: ie, running larger trains less often vs running busses more often. Once automation technology is sufficiently evolved to replace the driver, then the labour cost element mostly disappears. But the capital cost of building is so much less that there’s still many situations where BRT makes more sense.
    I mostly find that BRT is handicapped in wealthy countries, North America in particular, by the overly loose use of the term. This is really driven home by EngineerScotty’s comment of supposed BRT advantages. “As BRT can run in mixed traffic, BRT projects can avoid building expensive structures to cross things like rivers or other obstacle.” Which is all well and good except that BRT can’t run in mixed traffic. Buses can, but they cease to be rapid transit when they do.
    In reality, this is a big part of the drive to rail: it’s a lot easier to sell a 100% dedicated right of way on rail projects, and once it’s built it’s not going anywhere. Getting true BRT out of politicians in North America is next to impossible. And even if you can get a dedicated right-of-way, keeping it dedicated is very difficult in the face of the car lobby.
    People then confuse these political barriers as technological ones, and instead of just electing better politicians, they sell their politicians on the more expensive solution instead.

  11. Kantor57 December 4, 2014 at 12:39 pm #

    ” the only problem is that it is working so well, people are being left behind at stops by full buses. Buses are running at 2-3 minute intervals at peak times.”
    and this simply means that BRT was a wrong choice to start with; if with 2-3 minutes headways you leave people behind you have a big problem of capacity… a problem that can only be solved with light rail (longer trains…)

  12. Bolwerk December 4, 2014 at 5:05 pm #

    As “rapid transit” has a specific meaning rather at odds with even their idea of BRT, ITDP should really choose another term before they complain about dilution. How about something less vague like FRB (Fixed ROW Bus service)?

  13. Paulo December 4, 2014 at 10:50 pm #

    Exactly. while vehicle capacity isn’t an issue here in the states, it is in Guangzhou, and Jakarta, which means heavy rail is the answer… Something the itdp, and this blog, don’t understand. Brt will work fine in Eugene… Not the densest cities in the world.

  14. EngineerScotty December 5, 2014 at 5:09 pm #

    “As BRT can run in mixed traffic, BRT projects can avoid building expensive structures to cross things like rivers or other obstacle.” Which is all well and good except that BRT can’t run in mixed traffic. Buses can, but they cease to be rapid transit when they do.
    Fair enough; I’m speaking here of lines that are predominantly exclusive-ROW, but shift to mixed traffic running to get around some specific obstacle. Whether or not this is a useful compromise is not a decision that can be prescribed generally–it requires analysis of each situation.
    If a given mixed-traffic section is a sufficient bottleneck as to make the whole route useless–speed and reliability aren’t much better than mixed-traffic running the entire length (or local bus service for that matter), then don’t build this way. OTOH, if it’s a low-traffic structure near the end of the line that experiences little congestion, this can be a very useful optimization.
    A concrete example: My concern with the proposed Powell/Division BRT line here in Portland is that the portions of the route that will be hardest to give an exclusive lane to–along Powell, between SE 12th and SE 39th (along with the stretch along SE 82nd where the line likely will switch over to SE Division)–are the sections that will be most useful to provide exclusive-lane operation in. Both Powell and 82nd are state highways (at least Powell is, 82nd might now be owned by the City despite being signed with a highway number), and the state is unlikely to permit taking lanes from cars. Both streets also have a narrow footprint and many buildings fronting the street–widening the ROW to provide bus lanes while maintaining existing auto travel lanes would require extensive condemnation and demolition, especially along inner Powell. East of I-205, SE Division has plenty of room for a potential bus lane or two, but that’s the least congested part of the proposed corridor.

  15. Simon December 5, 2014 at 6:04 pm #

    EngineerScotty, that is not a Portland-specific problem. It is always the case that the most difficult parts to get bus priority are the most congested parts. In these places road advocates are the most strident, for obvious reasons.

  16. Spike December 6, 2014 at 11:45 pm #

    No – BRT was absolutely the right choice to start with and maybe keep as is without conversion for the next several decades.
    Have a look at how it works and where the route goes. Routes servicing the Northern Busway (including the Northern Busway only NEX) departs from Britomart in dense centre city and travels by bus lanes to reach the motorway and travels over the Harbour Bridge. From there buses enters their own dedicated highway and stops at stations until they reaches their exits to spread out into low density suburbs.
    If it had been rail, you can guarantee it never would have happened, as a rail crossing would have cost several billions and a massively lower benefit cost ratio (BCR). Added to that it would still have required heaps of man hours (and other operating expenses) in running frequent local connecting routes and separate express bus routes to town on the motorway anyways.
    Double decker busses are being introduced and frequencies upped significantly to bring in capacity increases. And personally I’d prefer a bus coming every minute to a train coming every ten minutes.

  17. Kantor57 December 7, 2014 at 2:32 am #

    I am not (and I cannot be) an expert of the specific features of that BRT and so I have to bow to your judgement..
    However I was responding to a specific point of @Goosoid; and I maintain that leaving people behind with 2/3 minutes headways is not a measure of success, but rather of bad progemming…

  18. valar84 December 7, 2014 at 4:43 pm #

    I’d be curious about how a rail crossing can cost several billion dollars as you seem to claim. With light rail, you also don’t actually need rail crossings at every intersection it goes through, France does very well with tramways, some nearly 50-meter long (160 feet), going through intersections in Paris with only traffic lights to control their coming and going.
    As to the claim that you still need “heaps of man hours” to run local bus lines to feed the rail… that is just outright dishonest. Sure, you need feeder lines, but if buses run simple local service on a 15-minute route instead of going downtown on a 60,70-minute route, it means that you still need plenty less buses to offer the same local service and thus save heaps of man hours.
    As to headways, three points I find most relevant:
    1- If you have buses going through at more than 1 per minute, forget traffic signal priority at intersections. You can only guarantee traffic signal priority at intersections if headways are sufficiently long, else you will end up cutting the area the line crosses in two because the light will always be red for cars and pedestrians wanting to cross the (semi-)rapid transit line.
    2- Speed gains for headways below 6 minutes are not that significant. The average delay is headway divided by 2, so there is little gain to expect below headways of about 6 minutes, even at 10 minutes, the average delay is still just 5 minutes while waiting for the vehicle.
    3- Often, the headway of BRT lines is not comparable to that of rail lines. BRT can hope to achieve rail-like capacity only by superposing many different lines in one corridor. But if many of these lines do not take a given user where he wants to go, then these lines might as well not exist for him. For example, if there are 5 lines on one BRT corridor, with one bus every minute, but only one of these lines will take you where you want to go (typical of BRT corridors where buses leave the corridor to continue on local routes), then the effective headway for you is actually 5 minutes, even if the total headway is 1 minute between buses. Since all buses do not serve the same areas, they cannot be compared to rail lines where each vehicle can be taken and serve the exact same route.

  19. Simon December 7, 2014 at 8:15 pm #

    In the Auckland context the BRT used the existing Auckland Harbour Bridge. In effect, the BRT is only a queue jump onto that. I would question the achievability of putting LRT onto the Auckland Harbour Bridge. Even if it is technically achievable, it sure as hell isn’t politically achievable. It is similar where I live in Sydney wrt to the Gladesville Bridge although that is virtually a flat impossibility for LRT technically (steep grades).

  20. Spike December 7, 2014 at 8:20 pm #

    @Kantor57 – that is why they are bringing in higher capacity vehicles and better frequency. Suffice to say, no one ever thought it would anything as successful as it is – it now carries 50% of the Harbour Bridge peak hour traffic.
    @valar84 – I should have been more specific as I thought “Harbour Bridge” would have made it clear – the Northern Busway crosses the Waitemata Harbour using the Bridge, and follows the grade separated motorway to Albany.
    Any rail crossing of the harbour would have required it’s own tunnel or bridge and thus would have drastically bumped up the costs and made it unpalatable.
    If you are going to discuss the Northern Busway, please read up about it and the context of why it was constructed and the service it delivers. BRT has it’s applications that it works very well for and rail likewise – I would never argue for BRT to be a stand in for say, the NYC Subway or the London Tube as massive capacity is clearly required through dense areas.
    However, the suburbs of northern Auckland are vast tracts of low density suburban subdivisions with only two routes (Harbour Bridge/Motorway and Northwestern Motorway/Upper Harbour Motorway) into town and therefore needs effective PT to bypass those chokepoints. That is why a busway where busses fan out from stations in addition to a busway only service works so well here.

  21. JJJJ December 10, 2014 at 12:37 pm #

    One problem people have in the US is that they see it as BRT vs rail, and not both.
    Mexico City has 12 heavy rail subway lines, 1 light rail line, and 4 BRT lines. BRT wasnt seen as a half-witted “lets do this instead and be cheap” but as a way to improve transit to corridors that dont need heavy rail. They have no problem continuing to expand their heavy rail system in corridors that DO justify it.
    Sao Paolo is the same. Massive heavy rail AND massive BRT expansion.
    In the US, only LA did it right. It doesnt make economical sense to extend the heavy red line to where the orange line went. That area is just too suburban. But a bus works. You dont need to build for a 10 car rail train every 90 seconds when a bus every 3 minutes does the job.

  22. Eric Goodman December 19, 2014 at 12:12 am #

    The latest from Seattle (King County Metro)
    “RapidRide continues to exceed expectations and provides more frequent all-day service along King County’s most-popular corridors. It’s A, B, C, D, E, and F lines have seen a combined 44 percent increase in ridership compared to the routes they replaced. That represents 14 percent of Metro’s entire annual ridership.”
    We’ve seen a large increase on SwiftBRT in Snohomish County too.
    BRT is almost always much better than the service that existed before its implementation, even when that isn’t pure in every aspect desired.

  23. Marco December 24, 2014 at 2:13 am #

    In France, most new extensions are made with BRT rather than tramways (as it was between 1990 and 2010). And they have developped a new bus that looks more like a tram than like a bus:

  24. Leila August 5, 2015 at 10:15 am #

    Sorry but about Brazil: Bus Rapid Transit is the worst service possible, even worse than streetcar buses. Our BRT buses (in Rio de Janeiro city) have hit and run so much there’s already a community of families whose children died by being hit. The distances are too big, the traffic is too intense, the system is too complicated and the drivers spend their days stressing and panicking over some crazy schedule mantained by the private bus companies.
    The government doesn’t control any choices related to transportation in Brazil but big politicians’ rich families do. It’s absolutely not efficient and we don’t even have more than THREE LINES on our subway yet. THREE. It’s not even enough to cover one zone of the city, the smallest one, on the south.
    So please check your facts.
    (We have more than enough money to fill the country’s territory with rails, if politicians stop stealing all of it).

  25. Magnus Gao June 1, 2017 at 4:47 pm #

    China needs the alternative between Subway and Bus. Many planners chose BRT, while a few of them chose LRT. Tram or LRT is not widely used in China.