Basics: Should Bus Rapid Transit be Open or Closed?

If you are involved in debates about Bus Rapid Transit, you need to stop and think about whether the project will be closed or open, because this will have a big effect on how useful the service is.  I’m always surprised at how few BRT projects clearly debate this issue.

A BRT system is open if the buses can continue off the end of the infrastructure and operate as conventional buses on local streets.  In situations where multiple bus operating companies run along the same path, open can also mean that the infrastructure can be available to multiple operators, although that almost always implies the first meaning as well.

A BRT system is closed if the buses must remain with the infrastructure, so that service must end at the end of the infrastructure, just as all rail services do.

In a given situation, a closed BRT option will require more transferring than open BRT for people to reach actual destinations that lie beyond the infrastructure.  As a result, it will tend to lead to longer overall travel times unless the speed advantages of the BRT compensate for that transfer delay.

There are two reasons this is a problem for your actual ability to go places:

  • A very single-centered urban form may logically need services to branch as they head out of the city, because as demand gets lower, you need less frequency but needed to cover more area.  Branching divides frequency, and in that case this can be OK.
  • But the bigger problem is that for non-transit reasons, the infrastructure may end where the demand doesn’t end, and closed BRT in this situation forces a lot of people to transfer just to keep going in the same direction.  In a high-frequency grid, for example, it’s important that service operate continuously all the way across the grid, so that while some people will have to transfer once few have to transfer twice.  Closed BRT can be an obstacle to this.

Despite this disadvantage, BRT systems are often closed for two major reasons:

  • In extremely crowded systems, closed BRT allows for tighter control of operations, for maximum capacity and minimum waiting time.  Capacity considerations may also dictate that all buses using the infrastructure be as large as possible.
  • In wealthy countries closed BRT more likely to be about trying to mimic the experience of rail transit, so as to be more attractive to a supposed discretionary or “choice” rider. If the goal is to make BRT appear special and different from regular buses, this goal is muddied if BRT buses run outside the infrastructure, or regular buses run inside of it.

The first of these reasons translates into measurable benefits in travel time, and thus access to opportunity, while the second does not.

Closed BRT is the more common kind of BRT in the United States, mostly for the second reason.  Where it appears in developing countries with very high public transit demand, it is mostly for the first reason.

BRT can be closed by any of several mechanisms:

  • Full separation. It can be made physically impossible for buses to enter or leave the infrastructure.    This is unusual, since buses may need to enter or leave in emergencies or to travel to and from the operating base.
  • Station and fleet incompatibility. Stations and buses may be designed so that they can only be used together.  For example, Eugene, Oregon’s BRT can run in regular lanes and even in mixed traffic, but its stations have high platforms that only match the floor height of the designated BRT buses, effectively requiring a closed system.  Fleet incompatibility can also be created through electrification, especially if end-of-line charging stations are required.  These stations become barriers to continuing service beyond the end of the line at that station, because the charging requires an amount of time that is practical only at the end of the line when no passengers are on board.
  • Operating plan. Service can be operated as closed even though the infrastructure doesn’t physically prevent open operations.

So should a new BRT system be open or closed?  In most cases, the advantages of open BRT are about people being able to go places so they can do things.  The advantages of closed BRT are mostly about branding and some limited kinds of amenity.

Brisbane, Australia’s BRT system uses ordinary buses that continue onto local streets, but it’s still really, really nice.

The concept of amenity is worth unpacking.  Many great amenities are possible on open BRT – see the beautiful busway stations of Brisbane, for example – but these generally do not include special buses with special features, unless you buy enough of these that they can continue to wherever those buses logically need to go to create the most liberating possible network.  There’s another reason to be cautious about special buses: Really, all buses should be nice, so creating a distinct brand of buses amounts to disparaging the rest of the bus system as much as it’s promoting the BRT.   We may be spending a lot of capital money to promote the idea that most buses are inferior.

But a few things, such as absolutely level boarding, benefit from buses that stop exactly at the platform level, and these buses tend not to be able to stop an ordinary bus stop.  Absolutely level boarding is great, and especially important to people using mobility devices.   But well-designed open BRT, with good operations and training, can still do reasonably level boarding where it’s easy to cross with a wheelchair or stroller.

I said “in most cases,” open BRT offers the best travel times and thus the most access to opportunity.  So what are the exceptions?  Closed BRT can be more efficient at very, very high levels of ridership – such as we see in big cities in less wealthy countries.  Here, an entire corridor may be continuously very busy, and in this case, the most efficient operations, and hence highest capacity, arise from being able to use every bit of the infrastructure and keep buses evenly spaced.  This is harder to do with open BRT, because buses may be entering or leaving the infrastructure part way, thus leaving a part of the infrastructure with fewer buses.  Buses may also be entering unpredictably, because they are coming from route segments where they are running in mixed traffic and thus subject to delay.  Where such huge volumes of people are traveling, these problems can cause pass-ups that do measurably reduce travel.

But this important exception arises only where massive capacity is critical, and this case rarely arises in the moderate-density wealthy countries of North America, Australia/New Zealand, or even most of Europe.  So in those countries tends not to offer any advantage to people’s ability to go places so they can do things.  In these contexts, closed BRT can deliver a better “brand” or “look and feel”, but open BRT is more likely to get you to our destination as soon as possible.  You decide which matters more.

18 Responses to Basics: Should Bus Rapid Transit be Open or Closed?

  1. calwatch April 21, 2021 at 9:31 am #

    Reliability is a key concern left unmentioned. One of the reasons Foothill Transit and LA Metro moved from an open BRT on the El Monte Busway to the Silver Streak and Silver Line, respectively, was because downtown congestion – even with a contraflow bus lane (Spring Street) – tanked reliability for riders on the suburban ends. It was felt that running buses every five or six minutes on the busway, and forcing a transfer, was better than trying to synchronize a bunch of 20-30 minute headway routes during peak hour to use the busway.

    Also, relatively few riders rode through El Monte, with something like 60%-70% of the trips solely within the suburban ends. In the off peak, many buses which would run to downtown from the San Gabriel Valley suburbs would end at El Monte anyway, and it does create some added travel time because of a transfer from a 15-20 minute headway bus to a 40-60 minute, whereas before it would likely be a direct trip – but given the weak Downtown LA center and that more non-work trips are likely going to be local errands, appointments, or school type trips instead of 20 mile trips downtown to see a ball game or go to an event in the central city.

  2. Chris April 21, 2021 at 2:31 pm #

    Boston brings us more than 100 years of experience with an “open” rapid transit system, with the famous Green Line downtown subway dispersing into 4 different streetcar routes in the suburbs. Any visitor or resident of Boston will immediately be familiar with the constant and never-ending bunching of routes in the tunnel, with five or more streetcars routinely back to back with a large gap in between. The Muni Metro in downtown San Francisco experienced a similar phenomenon for years until they made it more of a “closed” system, with some of the more lightly routes being cutback to end before they enter the subway portion of the alignment.

    In Los Angeles, ridership has increased on the 110 HOV lanes since the “open” BRT arrangement of south suburban express buses was replaced with the branded “closed” Silver Line. I suspect similar things hve happened on the El Monte Busway, referenced above. Also in Los Angeles, the Metro Orange Line originally branched at the western end of the valley, with half the buses going to Warner Center and half to the new Chatsworth extension. Dissatisfied with this “open” arrangement, Metro planning officials now send all buses to Chatsworth with a local circulator operating from the nearest station to Warner Center. Originally there was also a though to have some Orange Line buses leave the right-of-way and travel up Van Nuys and Reseda, but that was never implemented, perhaps due to concerns about bunching and also to the significant headway degradation caused by branching. In theory an open arrangement would be beneficial because it would reduce transfers but actual implementation, at least in the United States, seems to have always been problematic.

    • Henry O Mulvey April 22, 2021 at 5:18 am #

      Thanks for pointing out the Green Line Chris! Although I understand Jarrett saying all rail systems are closed, I think Boston’s historical rail system proves that it is not rule.

      • Jarrett April 26, 2021 at 9:23 am #

        All rail systems are closed in the sense that the vehicle can’t go beyond the infrastructure. Of course, it’s possible to run train lines that branch and/or run in mixed traffic. But I’m after a simpler concept of closure that’s more analogous to the BRT tradeoff.

    • calwatch April 22, 2021 at 11:41 am #

      You can do trunk merging on a grade separated heavy rail system like BART, WMATA, or the New York subway, but doing so in a mixed traffic system when the trunk has a headway of eight minutes or less is going to be problematic operationally. And through running during the off peak but forcing a transfer during the peak may be better operationally and for the rider, who no longer has to budget 10-20 minutes mid trip to make a transfer from the trunk to the branch, but looks backward to the average rider.

  3. Sailor Boy April 21, 2021 at 7:36 pm #

    It’s probably important to mention that your system isn’t open or closed, it is somewhere along a scale from fully open to fully closed. It may make sense for a system to be partly open. For example, buses can continue beyond the end of the busway or branch off at a key fork in demand, but minor stations are closed. This is the Northern Busway in Auckland, where both ends are open, but the existing Sunnynook and future Rosedale stations are closed. So the question isn’t ‘should the busway be open or closed’ it’s ‘how open or closed should it be’.

  4. Barbarian April 22, 2021 at 3:23 am #

    The under-construction TransBrasil BRT in Rio de Janeiro (which fits your criteria for benefiting from closed-BRT) currently has a lot of concerns about potential transfers, especially from people living in suburbs north and west of it. This is exacerbated by the BRT getting cut back from Candelária (in the city centre proper) to Central (on the outskirts of the city centre) to the long-distance bus station (which requires a transfer to the light rail in practically every case except for going to the bus station itself).

    Also, of the in-operation BRT lines (TransCarioca, TransOeste and TransOlímpica), two of them (TransOeste and especially TransCarioca) don’t just have too much demand for open-BRT, they have too much demand for BRT at all! Hugely overcrowded lines in which it can take hours to get on a bus at times. Other big-and-poor cities like Bogotá have similar problems, even Curitiba is facing calls to build a metro now.

  5. Henry O Mulvey April 22, 2021 at 5:26 am #

    I say open in most circumstances, especially for small city BRT systems where the infrastructure can serve many different routes. The debate about rail systems being open/closed aside, I do think having a BRT system open is a clear aspect that BRT has over rail, as rail transit is today in the United States.

  6. John L April 22, 2021 at 5:45 am #

    I agree with the above comments on the benefits of a closed system in terms of reliability. This is magnified in periods of disruption. On a simple, point to point line, it is much easier to skip trips and deadhead buses around to restore service. But if the open part of a system has branches, the problem of managing both the trunk regularity/capacity and the branch on-time performance become far more complex.

    Another complication is with fare collection. If there is off-vehicle fare collection on the guideway and on-vehicle fare collection on the local service, it adds to the complexity for customers and for operators.

    Yet another downside of a branched open system is the frequency of the branches. Two branches have 1/2 the frequency, three have 1/3, etc. You quickly get to 30 minute or worse frequency on the local portion. I’d honestly rather see high frequencies on BOTH a closed BRT and on connecting local services, with convenient transfers at a quality facility.

  7. Will April 22, 2021 at 6:30 am #

    Winnipeg’s BRT, the Southwest Transitway, used an open model when service started in 2012, but switched to a mostly closed model when the line was extended in 2020. The trunk service has been branded the “Blue Line” and the suburban routes that used to run through to downtown now terminate at Blue Line stations.

    The change has led to better reliability on the trunk line, a more regular service frequency on the trunk line, increased frequencies on the suburban feeders, fewer peak-only services, and a far more legible network design.

    I’m having a hard time reconciling this with the blog post, which makes it sound like the open model is undeniably superior. That certainly wasn’t the experience here.

  8. Barry Watkins April 22, 2021 at 4:38 pm #

    The comment that “The advantages of closed BRT are mostly about branding and some limited kinds of amenity” takes a narrow, western view of BRT. Yet most BRT systems are outside of the western world, and most are Closed. Having been fortunate enough to work on BRT design in not only Australia and New Zealand, but also Asia and Africa, the reason why most BRT systems are closed, is because that’s the only form of BRT that is going to operate successfully in those environments.

    Many of the factors as to why BRT systems in Latin America, Asian and Africa are designed as Closed BRT largely relate to the Station and Fleet compatibility issue and the reasons why specific design decisions are made. Most BRT systems in these parts of the world continue to use high-floor, high-entry buses, with a door positioned 90cm (3 feet) off the ground, which inherently requires stations to be designed accordingly. There’s a number of factors that relate to this, that when considered together, show us that the high-floor design is a defining feature of this type of BRT, rather than a quirk:
    – High floor buses are cheaper to build than low-floor vehicles, which is important in cities with limited budgets, and often little to no public subsidies.
    – High floor buses are more durable and cheaper to maintain than low-floor vehicles, which is an issue as vehicle maintenance is potentially undertaken with less rigour than we see in the west. These buses may also drive many more miles per annum than we are used to, and sometimes in harsher environments.
    – The high-floor design results in a significantly improved interior layout, as the wheel arches are then below the floor resulting in greater space for passengers and improved movement within the vehicle. (The impact of the wheel arches on interior space in low-floor buses is something we often overlook, because we are so used to seeing it now).
    – The high-floor design also makes it simpler for buses to be built with doors on both sides, as the structural rigidity of low-floor buses can be compromised when additional doors are added. Two-sided buses add flexibility in BRT design, allowing both median island platforms and traditional kerbside platforms in the same system.
    – The high-floor stations need to be gated at all entry points. The ticket gates obviously, but also the vehicle boarding points due to the danger of people falling between the platform and the bus. These gated high entry points then act as a barrier to fare-evaders, which is important in cities where severe poverty and/or lawlessness is an issue.
    – The high-floor stations also help prevent unregulated operators from illegally using the BRT corridor, as their vehicles won’t have the high entry doors needed. Nor will they be on the necessary side of the bus at median island stations.

    I’ve come to the conclusion that Open and Closed BRT systems are so inherently different from each other in the way they are planned, designed and operated, they almost represent different modes. Closed BRT must be viewed as akin to a rail solution (light or heavy), but implemented in cities where the cost and complexity of rail mean it simply isn’t viable. But it can also be used in environments where a greater level of operational control is needed. Open BRT is a substantially simpler concept, where the focus is on providing an express corridor for fast movement, and (hopefully) high-standard station facilities – I say ‘hopefully’ because systems like Brisbane’s are the exception, and not the norm. And because Open BRT systems are rather lenient on what sort of vehicle can use them, the customer experience in cities like Brisbane will often include a random assortment of buses including 20-year old diesel clunkers that many cities would have retired years ago.

    Both types of BRT have a role to play. But as with all transit planning, it’s about choosing the right technology for the city you are in, responding to the needs of the residents you have but acknowledging the economic, social and political environment that encompasses you.

    • Jarrett April 26, 2021 at 9:30 am #

      Barry. As I noted, closed BRT is much easier to justify in the developing world for many of the reasons you mention, but thanks for filling me in on the benefits of all-high-floor. In developing countries the sheer volumes of passengers, combined with the looser enforcement of traffic laws, certainly argues for systems that are as separate from traffic (and the buses stuck in it) as possible. Hope you’re doing well!

    • BindingExport April 27, 2021 at 9:36 am #

      Yay sandbagging a bus line to make it as expensive as a rail line – by “high quality” stations. In my city (Frankfurt -modal share for work trips 40% public transit) – a kassel kerb is enough for all street-bound transit (buses and trams) – it’s the universal platform and station type for the entire transit system.

      I don’t know what your work in Africa was – I hope it was not Myciti – a BRT system build at the cost of almost a metro system with a ridership of an European town bus network. The actual most successful system in terms of continous roll-out and ridership is Lagbus, an open system, with inexpensive low-floor side platforms and it is simple and cheap enough so that the local transit authority (LAMATA) has no problem to expand it on its own not needing outside “experts” overdesigning and complicating their system. It took Dar Essalam almost 15 years to design, built and operate one single BRT corridor and they had to spend $100 million on top of that on the ubungo interchange because running transit every 30 seconds over a extremely busy at-grade intersection is not a good idea.

      The problem of BRT in Africa is that it’s so over-engineered with the help of consultants that local authorities can’t implement on their own but are locked into the tedious and time-consuming bureaucracy of multilateral funding source to get their replica of the Transmilenio snake oil. Instead of being an adaptable and easy to implement solution it’s essentially as expensive and complicated as a rail-based one with the only advantage that’s popular with World Bank.

  9. Christof Spieler April 22, 2021 at 5:17 pm #

    Three additions:
    (1) I can think of two kinds of in between that the US has: branded BRT routes that extend beyond the busway but don’t branch (though they sometimes have lower frequency due to turn backs) which means that the BRT can extend from one end of the city to the other. Indianapolis is a great example — every other “Red Line” bus keeps going past 66th and runs in regular traffic and stops at regular stops to 96th. It’s not really open, since it is a dedicated fleet of buses, but it’s not fully closed either, and it does not have the forced end of line transfer to keep going the same way issue. The other in between is branded closed BRT providing local service within the busway section sharing the busway with non-branded longer distance routes that extend beyond it. You get the dedicated fleet, reliability, and branding of closed BRT and the one seat rides of open BRT, but at the cost of more service hours. Hartford is a good example, but Pittsburgh also operates like this to some extent.
    (2) One reason why many US cities go to closed systems is that they treat ticketing completely differently — board front door and pay the operator on local bus, Proof of Payment with ticket machines and all door boarding on BRT. That means to go to a good open BRT system (not delayed by slow boarding), they have two actually redo the fare system for the entire network. That’s effectively a fourth closure mechanism. Fares are an invisible but very influential barrier here, so this discussion has to take on fares.
    (3) Another characteristic that suits closed BRT is having lots of destinations on the busway. That’s part of what works in Cleveland, for example — a lot of trips happen entirely on the line. But where the major destinations are off the busway (like Pittsburgh) open has a huge advantage.
    I’ll also echo the reliability advantage of closed BRT — good closed BRT requires the ability to operate fairly reliably outside the guideway.

    • Christof Spieler April 23, 2021 at 7:01 am #

      Last sentence should read “good open BRT requires the ability to operate fairly reliably outside the guideway.”

    • Jarrett April 26, 2021 at 9:26 am #

      Great comments, Christof. Fares are indeed an important issue, and there are all kinds of hybrids, like Indy, that involve picking the best of both and accepting compromises to all desired outcomes.

      I was hoping this post would provoke you to talk about Post Oak BRT in Houston. I remember working on your plan and pointing out that it would be way more useful as open BRT!

  10. Jonathon April 25, 2021 at 5:26 pm #

    I would look to Ottawa as the best North American example of the benefits of open BRT. From the 1970s through the late 2010s (when the central transitway was replaced with light rail), Ottawa’s Transitway BRT used grade-separated bus roadways in the central city to reach terminals at the edge of the greenbelt (the undeveloped land between the city and suburbs). During peak hours, ‘express’ routes would use the central transitway and then continue out to each individual suburb. Literally- almost every suburban resident was a one-seat ride away from the downtown core at rush hours. Outside peak hours, branded transitway routes connected each of the suburban bus stations to the transitway infrastructure, using the exact same buses as the rest of the system.

    These suburban ‘express’ routes provided a massive incentive to drivers not to drive to work (a one-seat ride direct to downtown), and charged an extra fare for the privilege. Meanwhile, the combination of express and transitway routes meant that riders waiting at stations inside the city could catch a bus every few seconds that would take them downtown at 80-90 km/h.

    There’s actually another benefit that wasn’t noted in the article- because the Transitway was an open BRT model, it meant that the city could extend the transitway bus routes further and further into the suburbs, without having to build new infrastructure. There was a consistent pattern from the early 2000s onwards- a typical transitway bus route would terminate at the end of the transitway (at a major bus station on the city side of the greenbelt), then it would get extended to terminate at a suburban bus station (on the suburb side of the green belt). After a few years of ridership and increased service, the city would build new transitway infrastructure, extending the transitway to the suburban bus station. Rinse and repeat- the bus route would then extend from the edge of that suburb into its center, and a new bus station would open. Then after another few years, more service, then new transitway infrastructure. With this model, communities got to try out what it would be like to be on the Transitway, without needing new infrastructure, while the city evaluated what to do next.

    Now, this is not without its downsides- the open BRT model meant that many transitway segments got delayed indefinitely, never built, or substituted with highway HOV lanes intead. It meant that the downtown bus tunnel was never built, because buses could just use street lanes. But in practice it meant that the transitway could go absolutely anywhere, as soon as the city decided to expand service, without waiting years for new stations and roadways, and it also meant that Ottawans associate rapid transit with speed and frequency, more than infrastructure or branding.

  11. Chris May 5, 2021 at 12:56 pm #

    “It meant that the downtown bus tunnel was never built, because buses could just use street lanes.” The gridlock caused by the massive amount of Transitway buses on Albert and Slater Streets in downtown Ottawa significantly erased a lot of the speed advantages the buses enjoyed once out of downtown, and was a major reason why the Confederation Line is built. In any case, the Ottawa Transitway has strong elements of closed BRT as well, with identified routes – 95, 96, 97 – operating frequent all-day service only within the Transitway. With the exception of a few hours in the morning and afternoon peak periods, the Transitway is closed BRT.

Leave a Reply