Friday, November 5, 2010

108> Putting a Foot Down for PRT

Here is omething to think about…  As I have previously pointed out, because of the boom in highway projects started in the 50’s, cities in the U.S. tend to be spread out.  Instead of one central business district, there are usually many little ones.  To get people to leave the car at home is to get people to travel many miles without it, and there is no predictable direction of travel, just the likelihood that where they are going isn’t far from a freeway.  Therefore I have tried to accommodate this need with my designs.

Building PRT along freeways follows a different logic than many of the systems that are out there today.  It is like creating “worm-holes” between the more traditional routing schemes.  Naturally the structural requirements are different for both track and vehicles.
The track can be much cheaper as I will show.

Urban track must not be too visually imposing and must be routed to minimize distance between stations.  This generally means one-way.  There is also a lot of interference of all kinds in terms of track support placement.  This includes driveways, corners, underground utilities, telephone poles, signs, and traffic issues during construction.  Freeway following track, on the other hand, can have more frequently placed posts, which means shorter (less expensive) spans.  It can have two-way traffic supported by each post. There is less of the double track associated with stations and “Y” interchanges.  (With two-way traffic this necessitates up to four tracks, something that would create an unwanted umbrella effect, hence the general adoption of one-way urban designs.) 

Freeway following track represents a real chance to highlight one of the advantages of PRT, which is the limited weight of the vehicles.  Freeways must support 18 wheel trucks, and this sturdy construction is overkill for commuting purposes.   When a highway becomes congested, the call goes out to estimate the cost of another lane.  Each new lane is expected to absorb a certain number of cars for a certain number of dollars. PRT vehicles can be substituted for cars in this equation.  I recently read a Houston Metro study, which was an evaluation of a HOV lane running parallel to a section of I-10 (the main East-West interstate across the southern U.S.) The project was evaluated as a success primarily on the number of vehicles diverted from the main road.  As most of us know, one lane of PRT equals many lanes of ordinary highway.  It would be hard to make such a clear-cut case in a typical PRT route of convoluted loops.  Of course PRT vehicles are not free to the city, like cars, which are supplied by the commuters.  But autos are not revenue producers either, unless it is a toll road.  With PRT there is also no increase in traffic congestion on either end of the highway leg.  It may be a complicated cost-benefit analysis, but I think it’s less so than most of the other route proposals I have seen.  They generally look like nightmares of compromise and contentious litigation.

And so I have devoted some time to the problem and have come up with this.  In the picture above, the need for excavating has been eliminated by creating a support structure with a “foot.”  It occurred to me that the weight of concrete traffic dividers could be utilized along with a wide base to create a system that could be placed in the breakdown lanes.  Such a structure could not be made with just concrete and rebar because the edges must taper to nothing, so that vehicles’ tires can roll onto it freely.  Therefore there must be some clever integration of steel plate into the design. In addition to its shear weight, (about a ton per running foot) the base can be further immobilized by pinning it into holes drilled into the concrete.

Such track could be installed cheaply and rapidly by being pre-fabricated and trucked to the job site.  Sections of the concrete lane-divider would be interlocking or connected by steel splines, so that individual sections cannot be tipped or moved by impact.

Hopefully I will be able to figure out a realistic cost estimate in the near future, since estimating labor and materials appears to be a pretty straightforward job. I do not think I would go with the previously shown truss design, though, because it is for long spans, not cheap and easy construction. Anyway, my guess is that we will be pleasantly surprised.


Bengt Gustafsson said...

I like your thinking about a on-surface PRT post. This is an option that could be viable in many locations.

I wonder how you can claim that one PRT guideway can replace multiple freeway lanes. Are you thinking about subsecond headways at high speeds? - A freeway lane congests at about 2s/vehicle...

Andrew F said...


It sounds like you're suggesting this as a way to create 'backbone' capacity of fast, cheap and wide pipes to move people over longer distances at lower cost. Backbones don't do you any good if the last mile isn't there, however. I get the understanding that you're envisioning a system rolling out as a commuter system, bringing people long distances along highways from the suburbs using park-and-ride lots. I think this would be a tough market, as it's usually competing with commuter rail which is actually a pretty efficient form of transit. And if the circulator isn't there at the destination it is doomed to pretty low usage. So the circulation problem is the key.

Where I think this idea might have more merit would be for longer-distance automated freight transport. The USA, for instance, has an extensive network of interstate highways that would provide ideal rights of way for such a system. And the US logistics market is enormous. If you could get a system that covered dense corridor with poor freight rail coverage and high levels of congestion on the highways, you'd have a pretty strong market. Many firms already locate distribution centres along these highways, so extending spurs to customers should be pretty reasonable. This technology is ideally suited to just-in-time replenishment as it should be possible to deliver to very specific windows. Deliveries could be scheduled with some built in dwell-time near their destination to further improve reliability without significantly increasing costs. I could see FedEx/UPS and car manufacturers being big clients.

Dan said...

Bengt, I certainly do anticipate sub-second headways, and I am trying to gather my thoughts on this (and platooning) for a future post, so I’ll hold off for now.

Andrew, when you say, “commuting” it conjures a vision of a spoked layout bringing people to a centralized downtown. In the U.S., at least, there are often many small “downtowns” with freeways that connect them. If you look at a map of Phoenix, AZ., you’ll see that the cities of Glendale and Scottsdale are both less than 15 kilometers from the central business district (CBD). In Houston, an area about 10 kilometers from downtown sports a 67 story building and is one of the top twenty metropolitan areas in the U.S., being as large as Denver or Pittsburgh. Such spotty development calls for a larger PRT layout with faster vehicles than many have proposed thus far. It still benefits from a web-like configuration with many stations, however, which is ill-suited for transporting large groups at once. I wouldn’t advocate a single station in such “mini-downtowns,” but a loop at the very least. There will always be a problem with cost-effectiveness of the “last mile” regardless of the scale of the system overall. I must say though, that the drawback of buses (time spent loading/unloading) is greatly reduced if the trip is short. So I’m more worried about getting within 2 miles of more non-downtown locations than perfect “last-mile” coverage of the CBD. I wrote about this in Post 58. As for freight, I’ve always advocated that, and I don’t have anything to add at the moment.

Andrew F said...

Fair enough, Dan. I agree that in building a network, connecting two or more major trip generators is a good way to reach critical mass. This is pretty standard fare in mass transit planning. But to really leverage the ability of PRT to put people close to their destinations quickly, cheaply, without transfers, etc. requires good last-mile coverage. I see your point that providing good last mile coverage in a small initial network isn't that useful if the longest trip is only a mile or two, and that connecting two small local circulators with a fast cheap connection via a highway right of way. I'm visualizing in terms of Toronto, and connecting the main downtown train station/business district with the biggest suburban centre that has no subway service, Mississauga. Given that Mississauga's transit system does a decent job of funneling people to this centre, it could create a large market for a connection to downtown Toronto.

What I'm struggling with is whether this service is compelling enough to pull people away from commuter bus/rail. It might be a bit faster, but it could also be more expensive. A one-way trip from Mississauga's city centre (a big regional mall with some surrounding condos) to downtown is about $5 by limited stop coach bus. If PRT doesn't do a better job of getting people close to their destination than this bus does, I'm not sure its market is all that assured. The main station is on the south end of the downtown core, necessitating walks of up to 20 minutes through temperature controlled underground concourses (not too unpleasant) or short subway trips of a couple stations (for an additional $3 fare). If you could get people within a 5 minute walk of their destination, I think it could be a slam dunk. If you could capture a $5 or $6 fare for a 20 km trip, that's not too shabby for cost recovery either.

I presume you'd expect this guideway to work out to about $5 - 7 million/mile, otherwise the case for the more circuitous routing doesn't work.

Dan said...

Sorry I took so long getting back to you Andrew, sometimes I write responses in Word so I can spell-check. I just forgot to post this one…

I took a look at Toronto in Google Maps and I have to say that I would caution you about making any generalizations based on it. It appears extremely unique, having no south side, (lake) nor north side, (parks) and very few close-in highways. Out toward the airport or North/Northeast looks more typical. Also Toronto sounds like it has much better public transit in general than what I am used to. The Houston metropolitan area encompasses 579 square miles and only has 7.5 miles of light rail. I’m not saying that that is typical, but it is certainly not the only city where car travel is an absolute necessity. I really can’t make any recommendations for cities with decent, established transit. I can only design for what I know. Your comment makes me wonder something though, and that is how much better a single downtown loop with, say, 6 stations would be than the one station that you refer to. Also, how is the travel from Mississauga to, say Brampton or Vaughn? And would a single loop with a half dozen stations be a reasonable layout for these towns if they were connected with each other and downtown?

Andrew F said...

Hey Dan, I was just thinking about this a bit before responding.

I see what you mean about Toronto's geography, but a city on a large body of water is by no means unique. Most of the world's population lives very close to the oceans. In the US alone, I could rattle off a long list: NYC, Boston, Washington, Baltimore, Hartford, Miami, Chicago, Cincinatti, Buffalo, Seattle, San Francisco, LA, San Diego... And that's just based on my limited knowledge of the geography of American cities.

Toronto does have good transit by North American standards but it is pretty woeful outside of the downtown and inner suburbs. Any area that relies on buses, even if they are frequent, tends to have rather abysmal transit. Bus Rapid Transit can help can average speeds up, but otherwise you're trundling along at ten to fifteen miles an hour on average.

Andrew F said...

Mississauga and Brampton are largish suburbs in their own rights, with around 700k and 300k residents respectively. The roads are wide but can often get clogged, and most of the transit is a matter of funneling people by bus to the regional mall in 'downtown' Mississauga. Interestingly, there is a serious proposal at the moment to build a light rail line along Hurontario from Mississauga to Brampton, about 18 km (11 miles) for $1.2 billion (with likely cost overruns). Given that the peak pphpd is around 2000, I could see PRT being much more suitable to this corridor, and it could deliver faster travel times, more capacity, and more coverage for an equivalent investment.

I've been involved in some other web forums that take an interest in public transit in Toronto. It might help you to know how the average transit geek feels about PRT. Most of them see it as an evil plot by conservatives to block transit investment because these conservatives don't want to share vehicles with the hoi polloi. They also are totally convinced that PRT could not offer sufficient capacity to handle the kinds of loads we see here. I think that's a failure of imagination, as PRT is different than most forms of transit in that it tries to spread load out rather than concentrate it into large batches. So, the fact that a PRT station could not handle subway-station level volumes is a deal-breaker to them, despite the fact that subway stations cost upwards of 100 times what a PRT station would cost, and 100 PRT stations would have much greater throughput than a subway station, not to mention get people closer to their destinations. And that is where the idea of local circulators/loops comes in. To be able to handle the load, there would need to be an abundance of stations. Just one would probably be totally inadequate in an area that has sufficient demand to justify PRT service.

Andrew F said...

Let me just add that the crux of my interest in PRT comes down to my experience of how transit is developed in Toronto.

It is crushingly expensive, with subway running upwards of $450 million per mile due to overengineering and thoroughly unimaginative management. Even light rail costs upwards of $100 million per mile, and only manages average speed increases of a couple of miles per hour over buses. We're in the midst of a $4 billion LRT investment that will, at the end of the day, not generate any appreciable improvement in service. It's pure fetishism.

When transit is so expensive, it just means that it will never happen on the scale that is needed.

Anonymous said...

Hey, Andrew F, who are you? :)

Give me a line at Would love to discuss the Toronto area more closely.

This is why I envision giving local city-zens the power to plan their transport. We are the *best* specialists for our own daily needs, and putting enough of us together to plan for how the community should run cannot be such a bad idea. Right? :)

We've reserved a domain for this and will launch a public PRT track editor by the time BM One comes out.