« berlin: a new use for the brandenburg gate | Main | would you let this man design your city? »


Feed You can follow this conversation by subscribing to the comment feed for this post.


It might be worthwhile for you to take a look at the Toronto system maps, we do it more or less the same way. Things are less than perfect in terms of the full service designation being meaningful, but surface routes are marked as surface routes period.

There is now a separate category for local service with it's own ROW, but right now there's so little of it as to be little more than an oddity. Things might change with the new light rail lines coming in the next few years, but for now it is pretty much what you describe.


In your last streetcar article, you said bus technology is improving, and that anyone who doubted that should ride the oldest bus in their city's fleet, followed by the newest. I did just that, and the difference was enormous. The older buses are loud and have little suspension. When they go over a pothole or bump, the whole bus shakes and it is physically painful for the passenger. By contrast, the newer buses are a dream. The suspension reminded me of what you would find on a luxury car.

In terms of physical comfort and overall user experience, I would say the streetcar is superior to the oldest bus in our fleet, and inferior to the newest. (On an unrelated note, it is irritating to have to listen to the robotic woman's voice on the streetcar announcing every single stop and repeatedly instructing passengers at EVERY STOP on how to request a stop. Am I the only one who finds this unbearable?).

So the newest buses are more comfortable, and quieter (in more ways than one) than the streetcar. And they go 2-3 times as fast. They must be cheaper, too, as you don't have to lay tracks. And they can easily adapt to construction, etc. if they have to temporarily alter their route.

Given these facts, I find it incredibly perplexing why this perception exists in America that streetcars are better than buses. I can only conclude that perception lags reality by a couple of decades. But it is the responsibility of transit professionals to build the best system based on their superior knowledge, not to cater to the ignorance of the masses. Streetcars have a certain romance to them, but streetcar networks in America existed before the invention of the internal combustion engine so at the time there was no technological alternative. Now that we have choices, we owe it to ourselves to examine them carefully rather than simply rebuild a streetcar network out of a combination of nostalgia and regret at having ripped them out.

The length of the Portland streetcar line is 3.9 miles. It takes 34-36 minutes (depending on time of day) to run this route. This means the average speed of the streetcar is 6.7 miles per hour. To put this in perspective, the average human walking speed is 3 miles per hour.

In contrast to Berlin, Portland clearly brands their transit service according to the technology. Anyone who wants a good example of technology-centric transit branding should go to their website (www.portlandstreetcar.org). Not only have they set up a separate website just for the streetcar, their logo and color scheme is completely distinct from TriMet's. The name "TriMet" is not found anywhere on the website, except for in tiny print when you click on contact information. There is no information as to bus or MAX routes you might connect to from the streetcar, or any sense that the streetcar is part of a larger transit network with multiple options. Furthermore, they proudly advertise an annual pass for the streetcar alone. It is as if the streetcar exists in a universe unto itself.

This puts the lie to the claim that Portland's streetcar expansion is based on anything other than technology-centric thinking.

Just like with MAX, Portland can lay claim to being the first to implement a "new" technology. But the long-term risk is an inferior transit network.

I wonder what the hoi polloi will think 10 years from now when they are whizzing along at their 7 miles an hour to the Hawthorne district, and they are passed by a brand-new, gleaming, clean and quiet, Hydrodgen-powered bus. Perhaps then it will finally occur to them to question why they are so attached to a 19th century technology (that is if they can even think at all with the droning robo-woman's voice in their ears). It is at this time that the whole laughable streetcar fantasy finally falls apart. But by then it is too late.

Joseph E

There is no reason that a bus should ever be faster than a streetcar on the same route, Pantheon. Many city buses only manage 6 or 7 miles per hour in busy districts, just like the Portland streetcar. However, some European "trams" are more like light rail, with large sections of separate travel lanes down grassy medians, separate from car traffic. This, along with a greater distance betweens stops than most local buses, leads to quicker travel; often 10 to 12 miles per hour including stops. You might call it "street-running light rail" (LRT).
The only way buses will be faster on the same route than streetcars is: 1)they have fewer stops OR 2) fewer passengers per vehicle OR 3) the buses have a way to avoid traffic, by getting on a freeway or changing lanes. I suppose that some buses also accelerate faster than streetcars, but this is really a service flaw; most trams are designed to accelerate and brake at the maximum comfortable speed of 1.0 meters per second, while some bus drivers have a lead foot.

Joseph E

Los Angeles actually has a "frequent service" map: http://www.metro.net/riding_metro/12minute/12minute.htm
However, only the limited-stop lines (which all qualify for this map, I think) have special branding. I think emphasizing the local lines with frequent service would be beneficial. On the other hand, Los Angeles's local buses stop too frequently; often every 200 meters (1 to 2 blocks). 400 meters would be a much better average stop spacing for local buses. I already walk 550 meters to my bus stop; another 100 would be a small price to pay for faster trip times and a smoother ride.


The reason that the Portland Streetcar is branded differently is that it is actually run by a separate agency from Tri-Met, though Tri-Met handles most of the operations (and you can transfer seamlessly from one system to the other). But http://www.portlandstreetcar.org is a different organization than Tri-Met.

Now I've often wondered what would happen were Tri-Met to give a bus (a newer one) a nice red paintjob, brand it as a Streetcar, install a ticket vending machine (as is found in the streetcar system), and drive it along the streetcar route--most of which runs in mixed traffic. Would people still get on and ride, or not?

People who refer to streetcars (or rail in general) as "19th century technology" are sorely missing the point. The automobile is also 19th century technology, and the airplane came only a few years too late to also qualify for the label. Rail has proven to be a durable technology, and it certainly has changed quite a bit from the days when men shoveled coal into furnaces to produce steam.


Actually, the ability to maneuver around obstructions is a major advantage for busses, even trolleybusses. But most case, local service in either vehicle will average the speeds indicated.

One other important factor in ride quality is the quality of the running surface. Poorly maintained tracks, or rutted pavement, will significantly deteriorate ride quality. One advantage of the streetcar is that rails hold their condition much longer than pavement--and indeed, the worst vehicles for pavement wear are city busses, which distribute heavy loads across only two axles in a typical configuration. Rail-running vehicles also don't encounter speed bumps, pavement sloped for drainage purposes, reflectors and "botts dots", and other intentional deviations from a flat level road surface.

Jarrett at HumanTransit.org

Sorry, but Joseph E's comment is just wrong. If the signal priorities, exclusive lane arrangements, fare collection etc were all the same for the bus and the streetcar on the same route, the only difference in speed would be that when an obstacle occurs in front of the vehicle, the streetcar system stops and waits while the bus just goes around it. You've probably never seen a true bus-streetcar comparison -- I certainly haven't -- but that would be the outcome given the intrinsic capabilities of each vehicle. The point is that in their ability to get you somewhere, the bus and streetcar technologies are far more similar than different.

Jarrett at HumanTransit.org

Agreed. European and Australian stop spacing for local service is closer to 300-400m, while North American is rarely over 250m and usually closer. That's purely a difference of habit as near as I can tell.


I disagree with your statement that there is no reason a bus would run faster than a streetcar on the same route. There are some obvious reasons for this which have been alluded to previously - ability to maneuver around obstacles and faster acceleration speeds of buses. However, I want to mention one less obvious reason.

In riding the Portland streetcar, one thing I have noticed is that the doors take a long time to open and close. You know exactly how long it takes because a loud beeping sound occurs throught this process to warn passengers that the doors are opening or closing. I have never timed it, but I think it takes approximately 10 seconds.

Now you may say to yourself "big deal? It's only 10 seconds". Indeed, 10 seconds isn't that big of a deal. But when you are designing large transit systems, you have to recognize that the effect of every small technical advantage or disadvantage will be compounded through repetition throughout the system.

Let us assume the streetcar takes 10 seconds for its doors to open and close, and a bus takes 2 seconds. I think these are fair estimates. A northbound Portland streetcar has 26 stops to make along the line. Because the doors need to both open and close at every stop (and yes, I have almost never witnessed a situation where the streetcar could skip a stop) that means that at each stop, 20 seconds are spent waiting for the doors, as opposed to 4 seconds on a bus. Thus the time differential at each stop between bus and streetcar is 16 seconds.

16 seconds might not seem like much, but a funny thing happens when you compound this delay over the entire system: it starts adding up.

26 stops x 16 seconds per stop = 416 seconds

416 seconds = 6 minutes 56 seconds.

The shocking fact is that a barely detectable inconvenience, when compounded systemwide, has added seven full minutes to a 3.9 mile trip. Thus the 34-36 minute trip on the streetcar would take a MAXIMUM of 27-29 minutes on a bus, assuming no other techinical disadvantages to the streetcar (i.e. acceleration speed and obstacles).

Now just think about what happens when Portland extends its streetcar system further out, incorporating greater distances and more stops.

When you are evaluating these technologies as part of a systemwide approach, you must remember that everything counts. Even a small, seemingly insignificant technical detail can have measurable consequences throughout the system.


Portland just has shitty streetcars. The doors are certainly a part of it, and there's also the fact that the vehicles themselves are just SLOW. They're designed to go 20-25 mph, and top out at 30. Their acceleration isn't really that great. I think a PCC in good condition can outperform a Portland Streetcar. Conclusion: bad streetcars can be worse than the typical bus. Which isn't really surprising.


The VTA in the San Jose area actually has pretty wide stop spacing (don't know actual numbers, but 400m sounds about right). It does help keep things moving. One bad habit they do have though is placing stops relatively far away from the intersection, which is especially annoying at transfers.

Christopher Parker

Jarrett I really like this idea and would like to spread. I wonder if this is something google transit could somehow be involved in creating and spreading.

Joseph alludes to something I want emphasize: there are two kinds of service to highlight: frequency and speed.

Boston (which has traditionally used this system on it's bus maps, showing thick lines for frequent service and thin for other routes) also integrates the silver line BRT into the light rail limited stop lines. So the branding achieves your goal of being technology neutral. However, I don't think people are fooled. The silver line is lampooned and resented because it's a bus. Even the green line is regularly the target of grumbles because it's not a subway.

Alon Levy

I don't agree with the statement that "Let us assume the streetcar takes 10 seconds for its doors to open and close, and a bus takes 2 seconds. I think these are fair estimates." In New York, door opening takes about 5 seconds both on the subway and on the bus.


The Silver line is lampooned and resented because it's a bus presented as a rapid transit line. The Washington street branch of the Silver Line does not have a dedicated right-of-way, or a direct transfer to the rapid transit system. Yet the Silver Line has the same fare as the rapid transit system, more than the bus fare.


Is it resented because it is indeed slower or less reliable than either the Green Line (a LRT, for those unfamiliar with Boston) or the rest of the MTA trains (generally full-fledged underground metros)?

Or does the mere presence of tires on the thing cause grumbling?

Here's my silly speculation for the day: The sound of gears changing, on pretty much any direct-drive combustion powered vehicle, contributes to the perception of slowness. Not to slowness, but the sound of a diesel running through the gears as the vehicle pulls out of a stop, reminds people of 18-wheelers and dump trucks, and other vehicle classes that they routinely get stuck behind. :)

When you have a diesel-electric or direct-drive electric vehicle, neither of which typically have transmissions, that sound ain't just there. Instead you got the rising "whoosh" of the motor spinning up from zero RPMs, a monotonic increase in pitch that suggests performance--even though the faster the thing goes, the lower on the torque curve the motor gets.

I know, this sounds utterly ridiculous. But rational, intelligent people can be subliminally "informed" by sensory cues such as this, and they think the bus they're riding on is slow, even if trip times and other objective performance metrics say otherwise.


I've been on quite a few modern buses that also have a "droning robo-(wo)man's voice", but it actually doesn't bother me that much. In fact, I usually find quoting favorite transit recordings to be quite amusing, and I certainly prefer it to inaudible driver announcements that help no one (although they too can be amusing!).

I think that Portland is also a bad case example of a modern streetcar. They decided to make very few improvements to allow for faster travel speeds. I've been on many European streetcars that zip along quite nicely.

I agree with Jared that there is a time and place for both buses and streetcars, and we should be spending time figuring out when to apply each technology rather than putting one or the other of them down. But it's good to hear that you are a fan of some of the great new buses they've developed in recent years!


"Given these facts, I find it incredibly perplexing why this perception exists in America that streetcars are better than buses. "

(1) Buses churn out horrible diesel fumes and even the newest ones are very loud.
(2) The ride quality is terrible.

Now, if you're on a well-maintained European street in a trolleybus, these distinctions largely go away.

There's a third tricky point:
(3) In the US it is a lot easier to get fully dedicated lanes for rail-based technologies than for asphalt-based technologies.... and it's easier to get safety and high-speed enhancements like crossing gates. The Orange Line in LA is a classic example here. The "ability" to go into ordinary street traffic turns out to be a disadvantage when it comes to designing a reliable fast system. The ability to go off-street in a fairly narrow space is an advantage. Busways tend to be *WIDE*. (Australia, apparently, is one of the few exceptions, with guided busways, and I have no idea why they haven't used them anywhere else.)

By the way, there is never going to be a useful commercial hydrogen-powered bus, because hydrogen is not a fuel. Sorry to burst your fantasy bubble. *Battery powered* buses on cobblestone or brick roads (asphalt prices are going to skyrocket, folks), now that's a realistic possibility.


This is a rather good summary of advantages of rail even in mixed traffic.

An extraordinary number of city streets in Europe are still paved with very heavy-duty *permeable* surfaces like brick or cobblestone (rather than asphalt), and are maintained to a high standard of geometry (which we in the US are not used to with brick or cobblestone roads). This makes a significant improvement in the bus "ride" and in the wear-and-tear on the road, and significantly reduces the advantages of rail.

I'm not seeing that proposed in the US.


Bus doors around here take much longer than 2 seconds to open or close, and the Portland streetcar is exceptional in how long its doors take to open and close.

In fact, buses around here have to pull up, "kneel", open the door, and then have people file in single-file. The streetcars I've seen had multiple doors and didn't have to kneel so they boarded significantly faster. That *could* be done with buses, but it would probably be significantly more expensive.

FYI streetcars have better acceleration than buses because they're electric and have multiple powered axles. (Yes, you can design trolleybuses with multiple powered axles, but nobody actually *DOES* that.)


That difference actually arises from truly antique habits. In the 'old days', buses and streetcars would stop *anywhere* if flagged down. That was too slow, so it was reduced to "any corner".

Many systems in the US *STILL* operate with the "stop at every corner" rule, which dates back to the 19th century. Forget "stop spacing"; there aren't really designated stops. This is very much worth changing.


The Silver line is slower (partly due to more frequent stops, mixed traffic running, and many tight curves *even in tunnels*), less reliable, and crucially *more uncomfortable*.

The ride quality is *dreadful*. The ride quality on the Mattapan Trolley, which uses refurbished PCCs on barely-maintained track, is far far better. The ride quality on the Green Line is much better than that, and the ride quality on the Red/Orange/Blue lines is another order of magnitude.


Oh, and it also doesn't have effective level boarding. The drivers do not line up the buses with the platforms very well. This is one thing tracks are *very good at*; the train is exactly the correct distance from the platform every time (South Ferry station in NYC notwithstanding).


You are underestimating the size of the Bern tram a bit. A seven section Combino will be around 40 m long. At 2.3 m wide it is a bit narrower than a bus though, but still equivalent to at least two articulated 18 m buses. Your argument regarding capacity is still valid of course. By way of comparison Portland's streetcars are 20 m long.

Robert Wightman

I have just found you site and am enjoying it thoroughly. It is refreshing to find someone who is dispassionate about the different modes and technologies and is willing to present the good points and the limitations for each.

Your comment on how streetcars and buses should have similar speeds if both run in mixed traffic reminds me of a problem that happened in Toronto in 1963 when the TTC opened the University Avenue subway and replaced the streetcars on parallel Bay St with buses. The Toronto Roads department found that the speed of traffic in general dropped about 12% even though all the streetcars stops were in a 4 lane road in the left hand lane and required that the cars also stop. The Buses ran in the curb lane and cars could pass while they were loading.

A number of theories for this were advanced including the fact that the buses blocked the curb lane while left turning cars blocked the left hand lane. When there were streetcars, a streetcar travelling the opposite direction would and allow the left hand turns to get out of the road.

Please excuse my Canadian spellings, especially double l's and -our but I am too old to learn American.

The comments to this entry are closed.

the firm

Jarrett is now in ...

Related Posts Plugin for WordPress, Blogger...
Related Posts Plugin for WordPress, Blogger...