Transit
Rail transit modes fall on a continuum
Despite wide variety in transit systems around the world, here in the United States we only have a few words to describe transit modes, such as commuter rail, heavy rail, and light rail.
In my recent counterpoint post on the Silver Line and the viability of a Dulles Express Line using the Washington and Old Dominion Trail, I touched briefly upon the differences between American and European transit concepts. Because we don't have many words to describe a variety of systems, it can be difficult to compare and constrast different types of transit, especially those in other countries. With so few words, things get lost in translation.
As far as the Federal Transit Administration is concerned we have only a few words to describe our modes. There are certainly more variations on modes than those found in the National Transit Database's glossary. Instead of a fixed set of modes, transit systems really fall on more of a continuum.
In this graphic, more local modes serving shorter trips, like guideways and streetcars, appear above more regional modes used for longer trips. Even within a group, some systems have a more regional flavor than others. For instance, BART, while still heavy rail, is closer to a regional rail system than is Washington's Metro. And Los Angeles' Red and Purple Lines with far less suburban-serving segments are closer to the subways of the pre-auto era.
Spencer Lepler correctly called Metro a hybrid between subways and commuter rail. Since it is a hybrid, it can't be considered purely a subway or a commuter rail system. But the FTA doesn't consider the Washington Metro any different (as a mode) from the Boston T or the New York City Subway.
Metro is a modern heavy rail system, while Boston and New York have pre-auto age heavy rail systems. Chicago's Green Line is quite different from Washington's Green Line. The root of these differences comes from the market for which each was built. Chicago's Green Line was built mainly with walk-to-transit riders in mind, while Washington sought to cater to drivers bound for the central business district from the suburbs.
In the comment threads which have sprouted from the various Silver Line posts from last week and this week, many comparisons have been drawn to European systems and systems in other American cities. These other systems are good places to look, because we can gain insight from the variation between modes. Europe is a particularly good place to look to add a little perspective.
Because the distinctions are important, let's look a little more deeply at nomenclature.
- Streetcars: Streetcars often operate as single units in traffic with curb-side stops. Sometimes they have a semi-exclusive right-of-way, like a median or private ROW. They act as feeders to the regional system and mainly serve closer in nieghborhoods. Streetcars are sometimes older systems dating back many years (as with New Orleans), while modern streetcars are quite in vogue today (as in Portland and Seattle). Other cities have (re)created vintage (San Francisco) or faux-vintage (Charlotte) streetcar lines.
- Light Rail: This mode began appearing in North America in the late 1970s. Light rail offered a cheaper method to create a regional system and have become popular as substitutes to heavy rail. They serve corridors where heavy rail investment is not practical. In some cities, light rail systems take on many of the attributes of heavy rail, including downtown subways (San Francisco, Seattle) and level boarding (Charlotte).
- Subways: Systems like the New York Subway and the Chicago L are a type of heavy rail. While these types of service are certainly regional in nature, they often serve only the central municipality and not surrounding jurisdictions. Stop spacing is fairly close and speeds are slower than more regional-type services. The defining characteristic of this subset of heavy rail is that ridership is based on walk-to riders rather than drive-to riders.
- Rapid Rail: The modern heavy rail systems constructed in the 1970s, 80s, and 90s were often known as Rapid Rail. In fact, the Rs in BART and MARTA both stand for 'Rapid'. However the distinction gets a little blurry here. Rapid Rail systems typically serve the metropolitan area as a whole. Like the Washington Metro, surrounding jurisdictions have high levels of service as well. Speeds are faster than on subways and the distance between stops is higher.
- Regional Rail: Regional rail is a subset of FTA's commuter rail mode. I think of regional rail as different from commuter rail mainly based on service patterns. Regional rail trains, as their name suggests, serve whole regions. They also offer bi-directional and off-peak service in the region. One of the best examples of regional rail is the SEPTA Regional Rail network in Philadelphia. With a center city tunnel based on the S-bahn tunnels in several German cities, the network serves the central city and the suburbs with quick trips. Service headways are lower than one would find on Rapid Rail.
- Commuter Rail: Commuter rail, as I pointed out above, typically only offers peak-period, peak-direction service. In the Washington region, VRE and MARC's Brunswick Line offer good examples. MARC's Penn Line, on the other hand, acts more like regional rail.
- Other Modes: The FTA also considers Cable Cars, Inclines, Monorails, and Automated Guideways as rail modes. These tend to serve specialized and smaller markets. Cable cars, for instance, can easily be grouped with streetcars, as they serve similar markets and have similar attributes. Inclines have very limited use, but often serve as vital links within transit systems. Monorails and Automated Guideway systems also tend to serve very localized markets or act as distribution systems. It is possible to build entire systems with these technologies, as is the case with Vancouver's SkyTrain, but in the United States, they tend to be limited to moving people around central business districts, as is the case in Miami and Detroit.
Looking across the pond can create confusion as well. In London, for instance, the Docklands Light Railway has dramatically improved service to East London. But the system is not light rail in the sense that most American's think of it. It would be more accurately be described as Automated Guideway Transit, which is a separate mode according to the FTA. The Docklands system, unlike most American light rail systems, is entirely grade separated.
And contributing to the idea that America and Britian are two countries separated by a common language, heavy rail in Britian denotes regional and inter-city trains - not urban subway and rapid rail systems, which is what it means in the United States. Even Dr. Gridlock managed to get that one confused, telling a reader that MARC and VRE are heavy rail. In fact, according to FTA, MARC and VRE are both commuter rail.
But looking across the globe for good examples of transit generally yields greater understanding of the continuum, even if it makes it harder to quantify. Karlsruhe in Germany has a unique system known as the Stadtbahn. The Stadtbahn name in Germany usually denotes light rail-type trains, but in Karlsruhe the transit system is a unique hybrid of light rail and regional rail. In the city center, trains run in street. However, some services merge onto the conventional rail network for direct, rapid services to suburban destinations.
Munich and Stuttgart both have large regional rail (S-bahn) networks which feed into a central tunnel in the urban core. These tunnels have high platforms and essentially act as a central subway bolstered by the high frequencies resulting from the combined lines.
Expanding on this continuum concept, I added short-haul inter-city services, although perhaps a better name for the concept is needed. Amtrak services like the Capitol Corridor in California or Keystone Services in the Mid-Atlantic cater to longer-distance commuters as well as inter-city and inter-region travelers. Additionally, even high-speed services like the Acela offer a quick trip from far-flung suburbs. Joe Biden, for instance, commuted from Wilmington, Delaware daily on the Acela as a Senator.
We need a less rigid structure for naming rail modes. Individual systems possess a richer set attributes beyond just their FTA mode. Thinking of transit systems as lying somewhere on a continuum may help us find better ways of comparing systems across regions and nations.
Comments
- Community stories show the shift to a walkable lifestyle
- Focus transportation on downtown or neighborhoods?
- Young kids try to assault me while biking
- Some are pushing to limit sidewalk cycling
- Where is downtown Prince George's County?
- Endless zoning update delay hurts homeowners
- Metro bag searches aren't always optional
















And Alaska Railroad is represented as a mode too. Wonder why that is? Oh, right. Ted Stevens So it can get a total exemption from accountability measures that the rest of the country's systems are subject to: (From the CFR)
http://www.law.cornell.edu/uscode/49/usc_sec_49_00005336----000-.html#
(a) Based on Urbanized Area Population.— Of the amount apportioned under subsection (i)(2) to carry out section 5307—
(1) 9.32 percent shall be apportioned each fiscal year only in urbanized areas with a population of less than 200,000 so that each of those areas is entitled to receive an amount equal to—
(A) 50 percent of the total amount apportioned multiplied by a ratio equal to the population of the area divided by the total population of all urbanized areas with populations of less than 200,000 as shown in the latest United States Government census; and
(B) 50 percent of the total amount apportioned multiplied by a ratio for the area based on population weighted by a factor, established by the Secretary of Transportation, of the number of inhabitants in each square mile; and
(2) 90.68 percent shall be apportioned each fiscal year only in urbanized areas with populations of at least 200,000 as provided in subsections (b) and (c) of this section, except that the amount apportioned to the Anchorage urbanized area under subsection (b) shall be available to the Alaska Railroad for any costs related to its passenger operations.
by AA on Oct 21, 2009 12:59 pm • link • report
The light rail category has systems that carry a few hundred riders a day (Kenosha, Little Rock) to ones that carry several hundred thousand (Boston Green Line)
by AA on Oct 21, 2009 1:02 pm • link • report
http://en.wikipedia.org/wiki/Wuppertal_Schwebebahn
by Boots on Oct 21, 2009 1:40 pm • link • report
Again, this is a continuum, and all of these systems have elements of both. But certainly San Francisco's LRT is most like a streetcar out of these, while Seattle Central LINK is probably the closest to heavy-rail. And while I think there is ultimately a significant distinction between pre-metro and LRT, I don't know where you'd draw the line with some in-between systems.
by J.D. Hammond on Oct 21, 2009 1:42 pm • link • report
by J.D. Hammond on Oct 21, 2009 1:44 pm • link • report
I think measuring these would give you more of an x-y scatterplot than a linear continuum.
Still, none of that addresses network density or integration with land use.
by Alex B. on Oct 21, 2009 1:46 pm • link • report
by ah on Oct 21, 2009 1:46 pm • link • report
An excellent point about separating LRT even more so.
I'm not sure about the "Pre-Metro" name, though. In Europe, many cities built LRT systems which they intended to upgrade to full Metro later. Many of these were not converted due to expense.
Calling it "Pre-Metro" indicates that there is intent to upgrade later. That's not the case with most North American LRT systems.
by Matt Johnson on Oct 21, 2009 1:47 pm • link • report
You mentioned the slight difference with MARC. Yes, its three lines radiate out from DC, which would imply it's meant to bring people from the outer suburbs in to work. But the Penn and Camden lines also bring commuters in to Baltimore and back out at the end of the day. All lines only have weekday service, though you can't really say that the service is only in during morning peak and out during evening peak.
Some commuter rail systems run seven days a week, like New York's LIRR and Metro-North. Of course, it's clear there that people take it inbound in the morning rush hour and outbound in the evening.
SEPTA regional rail is definitely more of a commuter rail, as I see it, in that it brings commuters in from the suburbs and back out in the evening. I don't know, though. The distinctions are fuzzy at best.
So what about PATH? It's almost like what BART and WMATA are, but even more in the direction of commuter rail. Then again, it's a much shorter and more simple system than commuter rail, and obviously, it's mostly subterranian. But it's definitely not a subway, because it has so few stops in Manhattan that it'd be foolish to try to use it for travel within the borough.
by Tim on Oct 21, 2009 1:49 pm • link • report
The Green Line is street running LRT at points - and uses overhead power and smaller cars. Its operation is generally reflective of LRT systems and not heavy rail systems like the Red, Orange, Blue, etc. lines in Boston.
LA Metro has the same duality with the Red/Purple and the Gold/Blue/Green lines. Same with Cleveland RTA.
by AA on Oct 21, 2009 1:52 pm • link • report
by Jasper on Oct 21, 2009 2:35 pm • link • report
by цarьchitect on Oct 21, 2009 2:47 pm • link • report
- grade separated vs. at-grade
- dedicated ROW vs. street running
by Froggie on Oct 21, 2009 2:56 pm • link • report
by Reid on Oct 21, 2009 3:05 pm • link • report
http://bit.ly/aQT1v
by Steve O on Oct 21, 2009 3:08 pm • link • report
by J.D. Hammond on Oct 21, 2009 3:19 pm • link • report
Let's never speak of it again. Here is a good discussion of why the PRT concept is unfeasible. Now, let's get back to rail system nomenclature because it actually exists and works.
by Cavan on Oct 21, 2009 3:33 pm • link • report
glad to see an American with an open mind- I have seen the Schwebbebahn in person and it is most definitely a rapid rail type of thing.
It goes quite a ways and connects a number of hill towns including Wuppertal and Solingen- the legendary knife and cutlery capitol of the world.
It is fascinating to watch it in action.
My understanding is that it dates back to the late 1800's and is the oldest system of it's kind anywhere. A friend of mine who lived there said it was not damaged too bad in WW2 and was service was quickly restored . The landscape of this area looks like parts of West Virginia.
I have yet to actually ride the thing- but I can tell you- it zips along pretty darned fast ...
by w on Oct 21, 2009 3:59 pm • link • report
Boston is a good example of "whats the point". The green line. Underground subway which was developed in heavy rail style...because the plan was to convert it. Median running light rail. Above ground fully grade separated rail with 55mph speeds. Street running in mixed traffic. All with the same cars, all the same line. Thats 3 names in your continuum.
Then the blue line, which by your standards is subway/heavy rail....except half the line runs with overhead power (the rest is 3rd rail) and the cars are as short as street cars because the line used to be a street car line.
Then the red line was developed as a subway/heavy rail, but in the 70s expanded to giant park and rides with stations miles apart, aka rapid rail.
Then you have the commuter rail system (you call it regional rail). The franklin line is getting 4 new stops, which will make every stop less than a mile apart, and all within the city of Boston. Service is to run every 20 minutes....but using the same trains that continue on to rhode island! Yet today the line only runs on weekdays. What would this be called?
Now this is just one city.
There are so many hybrids available that do so many different things, theyre impossible to categorize. Like the train that turns into a bus when the tracks end (japan).
Perhaps train capacity would be a better way to separate them.
Low, including people movers, street cars and light rail.
Medium, including subways, and rapid.
High, including commuter and intercity.
While a commuter train can hold 1,000+ passengers, a trolley can only fit 120. Thats a more significant distinction.
by J on Oct 21, 2009 8:24 pm • link • report
For example Baltimore within downtown it pretty much acts like a streetcar for some portions so if there are portions of systems that have both designs that classify something as a lightrail or streetcar which is it.
Most modern streetcars and lightrail cars look exactly alike except for a few minor changes such as the track gage so what makes them different.
Btw what the hell is a metro it seems like that is up to the choice of the builder some are
light rail
streetcar
subway
rapid rail
regional rail
commuter rail
by Kk on Oct 21, 2009 10:27 pm • link • report
by ah on Oct 21, 2009 10:28 pm • link • report
Typically, Light Rail has its own right-of-way or its own dedicated lane, while streetcars typically share the lane with other vehicles.
by Froggie on Oct 22, 2009 9:57 am • link • report
This distinction does not exist in Europe, hence why modes are blurred and mixed a bit more. But Europe also has much smaller, less-frequent freight trains that allow such a system.
by Steve on Oct 22, 2009 10:35 am • link • report
It's also worth noting that one system/technology can (and often does) fall into several categories.
by Alex B. on Oct 22, 2009 11:05 am • link • report
For example - San Diego's SPRINTER. I, from an engineering background, call it Light Rail. It does not meet Crash Criteria to mix with freight, and therefore needs all attendant separations and platform designs. Someone from a planning background would most likely call it Regional Rail because of the frequent service and features of the corridor. Both names are correct within their context, but incorrect to the opposite profession.
by Steve on Oct 22, 2009 11:51 am • link • report
On that note, aerial tramways (distinct from "gondola lifts" used as people movers on ski slopes) are certainly more significant than cable cars, and deserve to be on this spectrum. They use the same gripping-cable means of propulsion.
by Squalish on Oct 22, 2009 2:04 pm • link • report
PRT - HA, there I mentioned it! Imagine that - someone bringing up transit alternatives in a discussion about transit! :-)
The LightRailNow PRT paper (which you link to) is complete bunk. Their entire thesis is based on the assumption that PRT cannot operate at closer than 15-second headways. In reality, three separate systems have received regulatory approval at less than 3 seconds headway, and one is soon to open at Heathrow.
That's a 400% error by the folks at lightrailnow, and it's never been corrected. How can anything else in that "analysis" be trusted?
In any case, in the next decade at least a half dozen PRT systems will be constructed on 4 continents, so any moratorium on PRT discussion will be difficult to enforce. PRT will be part of the discussion, and that's a good thing, because it may fill a significant transit niche and make all of transit better.
(disclaimer: I am a PRT proponent and frequently comment on discussions involving PRT)
by Mike C on Oct 23, 2009 12:37 am • link • report
Then let's sit tight until the Heathrow system opens. No discussion of PRT until there are results we can judge.
by цarьchitect on Oct 23, 2009 2:35 am • link • report
Conceptually, the only way that makes sense is in contained, campus-like environments, such as airports.
by Alex B. on Oct 23, 2009 9:12 am • link • report
That's fine with me, but when someone posts that absurd lightrailnow article as evidence of PRT "infeasibility", they're going to get a discussion, because that article is nothing but an attack piece full of disinformation, authored by people with a financial stake in seeing PRT fail.
For those who are interested in a more balanced treatment of PRT, see this recent Boston Globe article.
by Mike C on Oct 23, 2009 9:20 am • link • report
That's a fallacy. PRT is no more expensive per unit capacity than light rail. A 200 passenger train running every 7.5 minutes is 1600 passengers per hour maximum capacity. PRT vehicles spaced at 3 seconds and carrying an average of 1.4 passengers per vehicle is 1680 passengers per hour.
You can have larger trains and higher frequency, but that significantly adds to rail capital costs, and for those added costs you can simply add more PRT guidway coverage. So PRT is competitive, on a cost per capacity basis, with light rail systems.
In my opinion, PRT is (currently) NOT a substitute for heavy rail, which is much higher capacity than light rail. With operational experience, PRT technology may evolve into higher capacity systems, but for now there are plenty of moderate-capacity applications which PRT can handle well.
The idea that PRT is too expensive for the capacity it provides came from Professor Vukan Vuchic, who is otherwise recognized as a transit expert. I've read Vuchic's analysis of PRT and he just plain got it wrong. He may be an expert on trains, but he doesn't understand PRT.
(I can elaborate on any of these points if you like - e.g. how Vuchic got it wrong - but this forum is not the place for such extended discussion. Feel free to email me at transenth@hotmail.com if you want to discuss further).
by Mike C on Oct 23, 2009 9:42 am • link • report
The other problem with the concept of PRT is that it gives you point to point transit. This, of course, requires a full network of guideways - all of which are quite expensive. Again, you shouldn't be comparing PRT guideways to transit guideways, you should be comparing them to a road - since that's the more interconnected local network. Both cars and transit use local roads as their end distributor, either for vehicles or pedestrians. They're quite cheap in the grand scheme of things.
Again, since you've got these relatively expensive guideways, and PRT travel will be limited within that network, you're left with the logical conclusion that the best application for such a technology would be in a contained, campus-like setting.
For actual, urban use, I suspect we'd get far more value in attempting to get automatically controlled cars - that way, you've got expensive vehicles on a cheap network (local roads) rather than expensive vehicles on an expensive network - not to mention that you could then take your robo-car off the network and drive it under manual control, which of course you can't do with a PRT vehicle.
Denying the fundamental mis-match between guideway costs and network density is awfully bold, since that's the single most obvious reason such a system has never been built. You need to build all of it to get the benefit, and doing so would require a huge initial investment in a dubious and completely unproven technology.
But I'll bet it can be the cutting edge in airport circulators.
by Alex B. on Oct 23, 2009 9:51 am • link • report
Furthermore, regardless of what the people who hope to sell you a PRT system say, it is going to be extremely expensive to install a whole second level of infrastructure on every street. Anyone who says otherwise is using fuzzy math, or they are talking about something much less substantial than "real" point-to-point PRT.
Also, for the record, we already have a perfectly good PRT network. It is called the taxicab, and provided you have adequate density in your neighborhood it works very well, without requiring one cent on new infrastructure.
by BeyondDC on Oct 23, 2009 10:05 am • link • report
All PRT stats are theoretical. I WANT REAL NUMBERS BEFORE YOU HIJACK THIS THREAD.
by цarьchitect on Oct 23, 2009 10:06 am • link • report
You can get benefit out of PRT even if it's on a straight line. The benefits will not be as significant, but there will still be some benefit due to the efficiencies of driverless, on-demand operation. And a line can always evolve into a network.
But I'm not arguing "You should have PRT", I'm only saying "You should not dismiss it based on flawed assumptions".
@BeyondDC:
You got it right, but backwards. It combines (most of) the convenience of cars with the efficiency of transit. In other words, the best of both worlds.
For example, ULTra has done years of testing on its test track, so energy usage per vehicle mile is known very accurately. When you combine that known energy usage with the worst case real-world ridership scenario, PRT would use significantly less energy per passenger mile than the best US light rail, and 1/4th that of average US light rail? Since these numbers are based on measured energy usage and worst-case ridership, there is very little to dispute there. But even if they are off by a factor of 4, it's still no worse than average US light rail energy usage.
@architect:
Real numbers are coming: Korea, Sweden (Google translated link), and UAE are planning city-based systems to be deployed in the next 5 years, all different PRT vendors. And there are half a dozen US cities actively investigating it as an option. Two of these effort involve billionaires with a lot of financial and political will, so there's a good chance they will get built.
Real numbers are coming, sooner than you think. :-)
Again, email me if you want more detail: transenth@hotmail.com
by Mike C on Oct 23, 2009 10:46 am • link • report
Add a Comment