Human Transit

sorting out rail-bus differences

Here's a crucial passage from the book I'm working on, though it may will end up in the next book rather than this one [Human Transit].  The topic is emotive, so I'm trying to be very carefully factual here.  I welcome your critiques in comments.  If you disagree on a matter of fact, please provide a reference to a source. 

 

Claims for the intrinsic superiority of rail over buses often arise from people’s actual experience of using the rail and bus systems in a particular city.  In these situations, you’re not comparing the intrinsic benefits of rail technology with the intrinsic features of bus technology.   You’re comparing a particular rail system against a particular bus system.  Obviously, those two systems are different for many reasons other than the rail-bus difference.  But it’s easy to assume that the rail-bus difference necessarily implies all the differences that you experience between your own rail and bus options where you live.

In 2009, the popular American weblog the Infrastructurist asked its readers whether streetcars are better than buses, and why.  Readers came up with 36 reasons, which formed a good summary of popular perceptions about the rail-bus distinction.

Of the 36 reasons, only six refer to an intrinsic difference between bus and rail technologies.  All the others fall into two categories, which I’ll call misidentified differences and cultural feedback effects. 

Misindentified Differences

In your city, the rail system has lots of differences from the buses, including technological differences. But that doesn’t mean that all these distinctions are true rail-bus distinctions.  For example:

• Propulsion: electric vs internal combustion.   In most North American cities that have both bus and rail, the rail is electric but the buses use internal combustion (diesel, “clean diesel,” or various forms of natural gas).  Electric motors have obvious advantages – in emissions, noise, acceleration, and comfort – but none of these are true rail vs. bus differences.  Rail can be run by internal combustion, and buses can be electric.  If you want to compare your electric rail option with a bus option, compare it to electric trolleybuses.  If you want to compare your internal-combustion buses with a rail option, compare them to internal-combustion rail options such as the Diesel Multiple Unit (DMU).
  
  
• Mixed-flow vs exclusive-lane operation.  Transit speed and reliability are mostly a result of how much you stop and what can get in the way.  Rail is more often run in exclusive rights of way, but some streetcars run in mixed traffic and some buses run in exclusive lanes.  Monorails never get stuck in traffic, but neither do buses in Brisbane, Australia’s busway system.  Most city buses can get stuck in traffic, but so can any streetcar, tram, or light rail vehicle that runs in a mixed traffic lane.
  
  
• Off-board “proof of payment” fare collection vs. “pay the driver” fare collection.  Fare-collection style has big psychological effects.  “Pay the driver” slows down boarding and is a greater hassle for all concerned.  Some rapid transit sytstems (rail and bus) provide paid areas with faregates, eliminating this delay.  The other solution is “proof of payment,” which means that you buy a ticket on the platform (or already have a valid ticket) but you only show it if a roving “fare inspector” asks to see it.  If you don’t have one, you pay a fine.  Rail is more likely to use “proof of payment” than buses, but there are exceptions both ways, and there’s no necessary link between the rail-vs-bus choice and the fare collection system.  High-capacity bus systems are beginning to shift to “proof of payment” fare collection to eliminate fare-related boarding delay. 
  
  
• Frequency and Span.  Your whole rail transit system may be frequent, while some your buses aren’t, and in that case, you’ll naturally associate frequency with rail.  As we saw here, a good Frequent Network map, which shows both frequent rail and frequent buses, will clear up that confusion.  Buses can be very frequent, while some rail services can run infrequently or peak-only.  (We usually call those commuter rail.)

Cultural Feedback Effects

A community’s attitudes toward rail and bus technologies can easily affect they way they are operated and presented.  In short, people who believe that rail is better than buses will tend to act in ways that make that belief true.  For example;

• Differences in investment or care.  A community that believes that buses are only for poor people, or that rail is the mode of the future, will under-invest in buses as opposed to rail, producing a difference in quality that will reinforce that belief.  It may also hold bus operations staff to lower standards than rail staff, and encourage other cultural differences between bus and rail operations that become real for the customer, but are not intrinsic to the bus-rail distinction.
  
  
• Perceptions of permanence.  If you don’t stop to think about it, rails in the street will make a service feel permanent, especially if you’re used to hearing people tell you that rails imply permanence.  History clearly shows that rail systems do stop running if their market disappears.  True permanence lies in the permanence of the market, and that lies in the pattern of development [See Human Transit Chapter 14].
  
  
• Perceptions of legibility.  The notion that a bus might do something unpredictable and a railcar won’t is also a cultural feedback effect, typically the result of insufficiently clear and compelling information about the bus network.  It is quite possible to build bus services with such a high level of investment in infrastructure, such as stops and stations, that the routing is as obvious as a rail line’s would be; the Los Angeles Orange Line bus rapid transit system is a good example. 

• Regulatory differences.  Government regulation often enforces different rules for road transport as opposed to rail transport.  These regulations are themselves a kind of cultural feedback, differences in habit and history between agencies that regulate roads and those that regulate rail.  By enforcing different standards and safety requirements, these regulations can cause outcomes that amplify the apparent difference between road-based and rail-based transit. 
  
  
• Different potential for mission-creep.  If you build a stretch of road for a busway, there’s always a danger that somebody might try to open it to cars.  If you don’t trust your government to protect the stated purpose of a facility, this can be a major decision factor.  This issue applies, however, to the narrow range of cases in which a road or lane is being built that could be useful to cars but is closed to them.  It is not an issue where the proposal is to reallocate existing roadspace from cars to transit, nor when building a higher-end busway whose design makes it useless to cars even if they were allowed on it.

Intrinsic Bus-Rail Differences

When we set aside those two categories and look at the differences that really follow, intrinsically, from the rail-bus distinction, there appear to be seven, and only the first three of them are always to rail’s advantage:

• Capacity.  Where demand is high, rail can serve that demand at a higher ratio of passengers to on-board staff, which means that once you absorb the (often large) construction cost, you will be able to offer greater capacity for a given operating cost.  A transit vehicle that’s too crowded to board doesn’t meet any of our seven desires for useful service, so this point is often decisive in favor of rail.
  
  
• Ride quality.   Ride quality in buses is improving, and guided busways may give buses an even more rail-like feel, but new rail systems will probably always have an advantage with their smoother running surface.   Is the smooth ride of rail indispensible to a useful network?  This can be a tough question whose answer may vary from one community to another.
  
  
• Limited energy-efficiency and emissions consequences tied to the difference between tires and steel wheels.  Again, the primary factor governing energy-efficiency and emissions is propulsion (electric vs internal combustion), which is not intrinsic to the rail-bus difference.  However, there is a small range of differences that arise from the physics of steel-on-steel vs tire-on-road operation, and that favor the former.
  
  
• Noise from wheel friction.  Most noise impacts are due to internal combustion, which either rail or buses may use, so that’s a misidentified difference.  Rail transit lines that intersect streets may be required to install noisy crossing signals -- a valid response to the extreme weight of commuter rail trains but more controversial as applied to light rail.  These regulatory requirements may be cultural feedback effects.  But rail has a further noise disadvantage that really is intrinsic: the tight fit between steel wheel and rail causes noisy friction when going around curves, especially when going fast. 
  
  
• Some variable cost differences.  Broadly speaking, bus-based projects that use portions of existing roadway will be much cheaper than building rail for those same segments would be.  Beyond that, costs for bus vs. rail projects can be hard to compare.  Capital costs for rail include vehicles, while a busway is sometimes run with an existing bus fleet.  Certain bus-rail comparisons in certain corridors may turn up significant differences in operating cost that may be valid in that situation, but need to be checked carefully to ensure that they assume the same factors on both sides.

• Maneuverability around obstacles is a specific issue for rail in mixed traffic, usually light rail or streetcars.  In mixed traffic, minor obstructions routinely occur in a lane, especially if the lane is adjacent to on-street parking.  People stop in the lane to make deliveries, get into and out of taxis, and parallel-park.  Accidents and breakdowns happen.  If these events block a streetcar, the streetcar is stuck.  A bus, in the same situation, can often go around the obstruction and continue.
  
  
• Ability to extend existing infrastructure.  If you’ve already built rail on a large portion the length of a travel corridor, it may be logical to build rail on the rest, so as not to create a technologically required connection.  On the other hand, busways can often eliminate extra connections because buses can run through the busway but then flow out onto ordinary streets.  In each case, an advantage goes to the technology that makes better use of the infrastructure that already exists, whether road or rail.

Of course, in a particular transit debate, you may not have all of the choices that I’ve articulated here.   Still, it’s important to remember that most of the things you hear about why rail is better than buses are not true in the abstract, as facts of geometry or physics that follow from intrinsic differences between roads and rails. 

It may very well be that rail is culturally better than buses in your city, in which case all you’re really saying is that people in your city think rail is better than buses and will therefore tend to act in ways that make that true.  If you’re interested in appealing to your current population, and motivating them to make investment decisions based on their current perceptions about the benefits of rail, that may even be a good reason to build rail even if you don’t need its intrinsic benefits. 

But if you’re thinking in longer-range terms, don’t forget:  Attitudes, assumptions and perceptions will change over time.  Physics and geometry won’t.

UPDATE!  See endnotes for this post here!

dissent of the week: my alleged "bias" against rail

I'm relieved to report that commenters who actually saw me give the presentation "A Field Guide to Transit Quarrels" seem to agree that I wasn't displaying a bias toward or against particular projects, except perhaps for projects that were based on misunderstanding or ignoring some basic geometry. 

However, finally I have a comment that attacks me full-on, which gives me yet another opportunity to think about whether I do have a "modal bias."  It's from commenter Carl, who I believe saw the presentation in Seattle:

My perception is that you have a mode bias toward bus and BRT, and that this comes out in your talk and writings. I don't know if there is a professional reason for this (perhaps this is your expertise and source of consulting engagements.) These are the reasons that I feel you have a mode bias towards bus and BRT:

1. You say that bus/BRT can be made "just like" rail - in vehicle appearance, fare payment, stations, exclusive right of way, etc. Embedded in that assertion is the fact that the typical features of rail are superior to the passenger comfort, which you don't fully acknowledge. But even if all these improvements are made, it may close the gap but it is still not rail.

If I've every used such a vague term as "just like" in discussing rail-bus distinctions, then I was just being lazy.  I do believe, however, that there's such a thing as mobility, which consists of the ability of a network to get you to a wide range of places with a certain travel time and reliability, and which defines a transit system's ability to provide a sensation of freedom to those who choose to depend on it. 

And I do contend that most of the features of a transit service that determine travel time and reliability are simply not about the rail-bus distinction.  They are about frequency, stop spacing, boarding/alighting time, and the exclusivity of the right-of-way (what can get in your way).  I've also pointed out that in mixed traffic, buses have a reliability advantage over streetcars because they can go around many minor obstructions that would trap the streetcar.

I've never contended that bus and rail are equivalent in matters of comfort, but they are certainly converging, becuase rail cars are defining the comfort standard that bus design aspires to.  I agree that there will always be a ride quality difference.  How much that matters, in the long run, will depend on all the other factors that influence our decisions about how to travel.

There are a few cases where mobility does arise directly from a technology choice:

• Capacity is one of the best technical reasons to build rail.  If you routinely need to move more customers per driver than buses can do, you need rail so that you can run much longer vehicles.  Capacity is the reason that the Los Angeles Orange Line probably will need to be rebuilt as light rail at some point.
• Frequency benefits hugely from driverless operation.  This is currently an option only with rail (well, technically, fixed guideway) services.  

I have been consistent in emphasizing the importance of both of these benefits.  Carl continues:

2. You cite the lower cost of bus/BRT. The reason that bus/BRT is lower cost is that the investments aren't being made to create the ride quality and reliability. The primary advantage of bus/BRT is to use existing roads and require less investment -> lower experience. It's a valid trade-off for lower demand routes or if the resources can't be gotten, but the lower cost comes with lower quality, not equivalence.

I agree that if you compare rail to a completely closed busway, you can get similar costs if you design to similar standards.  Some factors push one way and some the other:  A busway tunnel has to be a little wider than a rail tunnel, for example, but rail has a power supply system and most busways don't.

But busways still have certain kinds of versatility that rail lacks.  In the high-end BRT system in Brisbane, the busway itself was very expensive but the buses serving it flow through the end of the busway and onto various routes.  This means that a large area has the benefit of the busway's speed without requiring a connection, and without requiring dedicated transit infrastructure on all those outer routes.  This is a very specific and powerful feature of BRT that rail simply can't do.

3. You dismiss evidence that the riding public has a preference for rail and that rail on a given corridor (with enough latent demand) will attract 50%-100% more riders than bus service at the same frequency. (See http://de.wikipedia.org/wiki/Schienenbonus or the comments by the Munich transit planner on the Munich thread). You call this cultural or emotional, but if even riders make an emotional decision, the evidence is there that riders prefer rail and will ride rail in greater numbers.

I don't dismiss it at all.  But to use the terms of my field guide, the intrinsic preference for rail is the reuslt of a combination of factors that are mostly cultural -- though ride quality has an element of the biological.  The preference for rail observed in ridership is partly about ride quality, and that will always be better on rail.  But it is also partly something that has been trained into the culture by the way rail and bus are often presented to the public, such as the message that rail systems are simple and buses are complicated.  In other words, the preference for rail is partly about an intrinsic difference but partly an echo of the history of how these modes have been used. 

What I see in fairly modern German and Swiss systems (Heidelberg and Bern come to mind) is a deliberate effort to make bus and tram feel as similar as possible.  Other cities, notably Karlsruhe and Munich, seem more invested in maintaining a feeling of difference between bus and tram.  Berlin displays impulses in both directions.

4. You dismiss the mode debate as political or emotional. If a corridor has traffic demand to support rail, the debate is often about whether to be cheap and under-invest in transit capacity (bus/BRT) vs. make the big investment in exclusive right of way and grade separation. The USA's most successful BRT right of way, LA's Orange Line, clearly should have been built as rail, which would give it higher capacity and lower operating costs.

Again, capacity needs, such as are coming up on the Orange Line, are a very solid reason to build rail rather than bus.  Rail will always be better at carrying more passengers per driver, and in some cases doing without drivers entirely. 

But I do not "dismiss the mode debate as political or emotional."  I simply observe that emotional factors play strongly in technology debates, and that while these factors have their place, it's risky to let them get out of control when you're building long-term infrastructure. 

Why?  Look around your city and I bet you can find some long-term infrastructure that's not at all what you would build today, and that presents obvious practical problems for the life of the city now.  Those facilities were designed to meet the emotional needs of a past generation, and some of these were built in spite of obvious mathematical or geometric absurdity because of the passion of the moment. 

The US Interstate Highway System is full of examples.  Why did the US build grade-separated interchanges at every farming road in North Dakota, so that I-94 could be built to Interstate standard instead of being just a really fast highway with some very minor intersections?  Well, one answer is that the Interstate system was conceived as the Intersate and Defense Highway System, and defense is an emotive topic.  After all, we might need to move tanks from Seattle to Miami without stopping, or even being delayed by a truck slowing down to turn onto a farming road.  The Interstate system was driven by an emotional obsession with a single, consistent, national network, and this caused huge sums to be spent on things that no longer seem to have much value. 

You can tell the same story about the rise of car-dependence in cities in the 1950s and 1960s, a period where we fell in love with cars to the point that we didn't foresee obvious geometric limitations on how many of them we could fit into a city.  This simple geometry mistake led to a range of urban woes that many people, including myself, will spend their whole careers trying to undo. 

Both of these highway examples are the same story:  Things were built a certain way to meet the emotional needs of a moment in history.  Today, the emotions have changed, but the geometry hasn't.  So we're still stuck with the geometric consequences of those emotional decisions.  

Emotions change, culture changes, priorities change, but geometry will never change.  A project becomes risky when it starts trying to use the emotions it arouses to overwhelm geometry-based objections.  Emotions will always play a role in technology choice.  But it is in the nature of emotions to sweep us away, so sensible people just notice when that's happening and keep an eye on reality.  They don't ignore emotional factors or oppose their influence, but they recognise them as emotional and keep them in perspective.  It comes down to that age-old advice about teenage dating:  "Keep one foot on the floor."

ride quality: the driver's role

In my series on streetcars, I've been groping toward constructing a coherent view about technology choice, a hugely expensive and political issue in transit development.  Since this is a blog rather than a book, I'm thinking out loud, engaging with comments, and revising without erasing.  The effect has probably been jerky and lumbering, with lots of small lateral motions that evoke the feel of riding a bus.  

Speaking of ride quality, a reader asks:

Do you know if there are any cities that make a point of ensuring their bus drivers provide a smooth ride? In my experience, even with the same model bus on the same route, some bus drivers manage a vastly more pleasant and less jerky ride. So I'm just thinking that this aspect of the bus experience should be technically feasible to improve...

Good training covers this, but my own hunch is that drivers are good at ride quality based not on training but on how sensitive they are as people.  A sensitive driver will constantly make unconscious choices that produce a smoother ride for her, regardless of whether she's just driving her own car or driving a bus.  A person who's just not sensitive to quality of ride is unlikely to be made more sensitive by the kind of training that bus drivers get.  

But I'd be interested in other perspectives, especially from bus drivers and people who know them.