Tuesday, July 26, 2011

126> A Few Good Destinations

In the early days of PRT, back when governments and multinational companies were first eyeing the idea, the whole concept was so futuristic that it was plenty enough to envision a standard vehicle, track and station and multiply them around into a grid, and call it a system. Back then it was a question of whether those new computers were capable of reliably and safely managing traffic flows.  As I was introduced to the concept, there would be a station every block or so, so that it would not be too long of a walk to get to one. The track would be one-way and you could get anywhere by circling around. Thinking back, maybe this just wasn’t good enough, and nobody realized it.  

Personally, I have had a very hard time trying to shoehorn the cities I know into such a scheme. It would be great if the funds were there to actually make such a comprehensive grid, but since they are not, it becomes a question of giving the most bang for the buck. But the systems that were designed for the grid model may not exactly fit the new roles. Moving away from a grid model has implications for the track, the vehicles, and the stations.  

Here are some typical situations that may exist outside of any downtown grid; The freeway commute - this calls for a relatively fast vehicle and outlying stations with lots of parking. The “strip”-  This is where eateries and retailers have reached a critical mass so that the  whole stretch of road has become like one long mall. It probably would call for two-way track that does not interfere with signs and driveways. Stations should be minimal footprint, perhaps designed specifically for private property, such as in the parking lot of a major retailer. Major destinations – Areas such as a museum district, a major hospital complex, or stadium need access to the system, although there may not be enough in the budget to put stations every half mile between them and other destinations. These call for large stations and a system with distributed automobile parking, since it is unknown where a visitor’s origin is, but it is likely that the first leg of the journey was from outside the system.

I think that each of these scenarios is extremely typical and each plays to the strengths or weaknesses of a given system design. True, these introduce design complexities that are much greater than what a PRT company would ideally want to tackle. But what is the choice? Try to interest cities in a “one-size-fits-all” system? 

In the end it is each city’s specific layout that must be addressed. Perhaps rather than a grid mentality, what is called for is a destination mentality. How can the most important destinations be served with the least amount of track and stations? That, after all, would seem be the best value proposition from the city’s standpoint. Yeah, I know... None of this really plays to PRT’s strengths. But PRT track is also cheaper and less disruptive to install, and being raised so as not to block crossing traffic is a huge bonus. So perhaps PRT can prove itself with less track and stations than the network we would like to see. 

A couple of points: First, parking. It seems pretty obvious that most people will have to park their cars to use the system. Maybe there are a few older cities out there that are teaming with pedestrians who live very close in. But for most of us, the construction of arterial highways has created a suburban landscape of car dependent homeowners and apartment dwellers. This calls for assessing each of the out-lying station’s potential to be a gateway to the system, and therefore a place to store the car in the meantime. Will merchants be willing to share parking in exchange for being convenient to the riders? Probably not in areas where people would want to park for the whole workday. Bottom line – Any system will need sufficient parking to support enough passengers to make the system viable. That potentially means thousands of parking spaces. It is likely that some stations will essentially be parking lots. Land costs are not inconsequential, so parking ends up becoming a factor in routing.
Another point is that in a landscape of very limited funding, shuttles (GRT) must be reconsidered. If the system is centered around serving the most important destinations, then it stands to reason that more people will be sharing a common itinerary. This has implications for track size, although we must avoid anything too big to be visually acceptable. The track I have shown in previous posts is about as big as I would want to risk. I think it is noteworthy, though, that technologically it is a simple matter to keep heavier vehicles spaced further apart than lighter ones to minimize weight concentrations on the track. I would keep it under six passengers anyway.  Such vehicles would simply share track with the PRT vehicles and move between high capacity stations. These would be “express” shuttles, so if your destination isn’t a main terminal, you would use PRT, which could service all destinations.

A last point about stations. I have opted for a suspended design mostly because such a system can drop to ground level and ascend with a minimum of station related hardware and track. Neither long ramps nor elevators are required, which is of paramount importance in a stripped down, budget starter system. A suspended system would also seem ideal for parking lots since PRT vehicles could go directly to your car yet there would be no track to cross.  
In a “destination oriented” design, the point would be to enable the rider to eliminate the lion’s share of driving from his/her day-to-day routine. The idea is to make all of one’s normal destinations available and convenient to the rider – shopping, dining, entertainment, etc. A well thought-out system could provide traffic relief that would ripple throughout the city’s side streets, not just the roads that parallel the track. This is because it would cut out what would otherwise be individual outings in the car. A few choice stations, (and some very lucky merchants!) would make most driving unneccessary.  Also, I cannot help but consider such a proposition from a tourism point of view. To some cities this is a very big deal. And, being elevated, it’s naturally scenic! 

So how stripped down could a system be? I guess I can imagine a single convoluted loop as a starter, but every city is different. Finally, I would add that there MUST be a way to branch track without a long shutdown. Any prospect of skipping over areas can only be temporary. Success will mean a demand for stations all along the track, so adding stations must be easy to do.  I posted a design for a branchable box-beam track in post 71.


Asko K. said...

Good write, thanks Dan.

Robotic parking solutions can help, i.e. from Skyline Parking (Switzerland).

Andrew F said...

Dan, the problem with relying on drive-up business to justify a system is that you can never build enough parking to fill the system. You'd spend as much again on land acquisition and construction for parking lots as the system itself would cost. I could understand supplemental parking (run on a cost-recovery basis), but a system would have to rely on either walk-up traffic or transfers from other forms of transit for the bulk of its ridership.

It doesn't have to be sophisticated transit. It could be dial-a-ride jitneys that pick up people around a neighbourhood and bring them to the nearest PRT station. This would be more cost effective than building even unsophisticated surface lots. In my necks of the woods, I found an example of a 200 spot park and ride that cost $2.5 million. At 10% discounting (should probably be higher), 250 days per year, that puts capital costs at $5/spot/day. This does not include land cost, which runs a few hundred thousand dollars per acre, minimum. Operating costs would be higher. Park and rides are not an efficient use of infrastructure funds, especially if they are given away for free. Charge $10-20/day for them? Fine. I just don't see how you could get the ridership you need relying on people paying those rates to park, in addition to the fare.

A nascent PRT system needs to have walk-up customers. Try linking the local university with a mall, a hospital, a bus terminal, etc. Surely even the sprawliest cities have malls, universities and bus terminals? If they don't, they're probably not good places to try to prove PRT. Once PRT is proven, I think the need for the system to operate on a cost-recovery basis from a small network size will be less pressing. A PRT network creates such substantial positive externalities that local governments ought to be willing to at least bootstrap the system. They can include foregone road investment and transit operating costs in their business cases.

The most important thing is to prove it works in urban environments. We might as well pick the easier applications to prove the system, and leave the hardest (sprawlier) applications for once the technology is mature and accepted.

Maybe not what you want to hear, Dan!

Dan said...

Hi Asko, You got a new look!
Andrew, I think you've got to appreciate that Toronto is not a typical city for the lion's share of my readers, although what you say might well apply to many cities in Europe. In the US, in the 50's, then president Dwight D Eisenhower thought it was in the nation's security interest to have high speed highways. It was all part of the cold war. Local politicians soon realized that they could bring home the pork, and now we have many cities with virtually no pedestrians at all. In Houston, I can think of just one "strip" where pedestrian traffic MIGHT support PRT. But virtually everybody there has parked their cars somewhere already. Nobody actually LIVES there...

Parking? We have lots of that. Freeways make for very uneven development. Many cities would gladly spring for parking in an area in need of renewal.

Park&Rides are a proven winner as far as most people are concerned in the US.

Luckily, PRT would tend to create pedestrians and more mixed use development, even in cities with no zoning or planning whatsoever, like many US cities. Until you have people living close-in, you have to deal with parking...

Andrew F said...

The other problem, of course, is that a system that only serves park-and-ride lots will have the last mile problem. If every station is surrounded by acres of parking, riders will still need to walk long distances to their destination.

I guess I'm just not seeing the business case. How could you recover the cost for all that free parking? Or are you suggesting that the system just rely on the parking that already exists? Whenever I've looked at the numbers, PRT seems to need decent usage to get to cost recovery, and that is without having the pay for any parking facilities at all.

There are plenty of cities in the USA that are just as walkable as Toronto or moreso. San Francisco, Boston, Seattle, NYC and I'm sure a dozen or so smaller cities that I am not familiar with. I think nascent PRT systems shouldn't even try to penetrate the most walkable parts of a city initially, as these are likely already well-served by transit. I've looked at suburban applications that are not-well served by transit and not especially walkable.

Andrew F said...

I'll also point out that Toronto also has the second busiest highway in North America. As an accident of history and geography, there is one main east-west highway running through the city. In places it is 20 or 24 lanes across. The city itself is not well-served by highways, but the suburbs have a pretty extensive highway network.

Dan said...

Problem is, Andrew, that many cities don't have any place you can put a station and get more than a dozen walk-up customers per hour. In the US there has been a real flight from the inner cities as a place to live. This is not the case everywhere, mind you. Everybody in NY loves "The Village", and Boston has a great mix of residences and commercial properties (They also have rent control, zoning etc., unlike the sun-belt cities) I'm thinking you would need several thousand urban residences including some high-rise to get enough foot traffic to make a station pay off. Your point about Universities is well taken: Places where a big school is right in the city do appear pretty prime, though often somewhat self-contained.

Car dependency is a self-fulfilling situation. Once everyone uses them, the whole city layout changes to accommodate them. Soon nothing is within walking distance, so every store tries to be about a 5-minute drive from as many people as possible. This landscape favors very big stores with lots of parking.

The traffic is out in the suburbs, not in real close, and so are the jobs. Downtown is more of crossroads than a destination since it is at the center of the whole mess.

We may be trying to solve two different problems here, where what each of us is talking about is not a problem in the others city...

Andrew F said...

Dan, if the problem is that no one will walk to a PRT station, won't the problem also be that no one will walk from a PRT station? If the land use is such that the density is too low for a critical mass of destination points to be within walking distance of the stations, people still won't use it, even if they can park and ride where they enter the system.

How are you envisioning that that problem be addressed?

5 minute walk at 5 kph gives you 400m radius. 10 minute walk is 800 m radius. Those would give a station catchment area of 0.5 km2 and 2 km2 respectively. There are ZIPs in Houston with nearly 5,000/km2 population density (77002, 77006). That should be the core of the initial network in a city like Houston. For it to cover the whole metro area, I think, would require some infill development or a pretty loose network.

I don't think it's realistic to make PRT viable in very low density areas, especially as the initial network.

Dan said...

The destinations where the most pedestrians can be found in a car-dependent city tend to be in commercial areas that are driven to. Sure, there are usually some residences nearby, but not really enough to contribute to the foot traffic. Such an area would probably have at least one large grocery store, a department store or two, and a half dozen fast food places, and a few strip centers with places like a hardware, shoe, pet stores, and maybe a professional building or two, all within walking distance of each other. Such places are fed by nearby neighborhoods that are within a couple minutes drive. It’s kind of like a little town, except there the situation repeats itself again and again (every few miles) in a sprawling city. It is why most people can generally get most things done within a five minute drive. I say “most things”, certainly not ALL things. One’s dentist, or library, or favorite restaurant might be in a similar but different commercial cluster in a different direction.
As to the first question…Shopping malls, strip centers, etc. are often out of walking range to more than a handful of residences. If PRT passengers have left their cars somewhere, they can go to the various businesses that are on route. The problem is on other end. If they have a house with a yard that pretty much means that there is just one dwelling every 50 to 150 ft. (Sorry for my lack of metric conversion, I’m not at home and have to get this out before the place closes.) That means chances are there is a long walk involved.
Now about those zip codes… One was the entire downtown, so that would be a destination … The other is one of two close in neighborhoods that are very densely populated. (You picked the Montrose area, the other is the “Heights” just north of I-10. ) Both are very old neighborhoods, both once had trolleys, as a matter of fact, and they generally have very small houses and yards, hence the high population density. Neither have much in the way of shopping, (nor does downtown) but the Montrose area has a lot of quirky little eateries and boutiques.
Both neighborhoods are within a 4 minute drive to downtown, and there really isn’t much traffic. That’s why houses there are so expensive. So the question becomes, “Do you want to build a system for the 3% of the population who doesn’t have a commuting problem when the other 97% sits in traffic for on average for twenty minutes or more?” (that’s not counting the greater metropolitan area, which reduces those zip codes to about 1%.)
Don’t get me wrong, serving those communities would be a good thing, and Montrose Blvd. intersects with Westhiemer, a “strip” with every type of commercial activity which just gets denser as you go further from downtown, and stays that way for 20 miles… One the way it goes through “uptown” Houston, (otherwise known as the Galleria Area) which actually sports Houston’s tallest building, which is about 5 or 6 miles west. It is that area which is a real mess. Everything is bigger there, the yards, the houses, the stores… That would be ideal for PRT but people would have to park somewhere, except for the hotel patrons, (actually there are quite a few…) Anyway, the place is closing… Hope this gives you some insight, Andrew.

Andrew F said...

I hear what you're saying Dan. I'm just approaching this from the perspective of how do you design a small system (low capital costs) that can generate a return on investment that will be leveraged into expanding the system further. Clearly, covering the whole metro area might cost billions, even with a lightweight network. It would also likely lose money for several years as you wait for demand creation.

If you focus on those neighbourhoods that are dense, and close to downtown, you can help your network reach critical mass earlier. It sounds like you would need to serve downtown anyway, and while you're there, you should serve the two closest neighbourhoods. Even though these neighbourhoods are close to downtown, I imagine that making car-less mobility much easier would make it quite tempting to just get rid of the car for many people who live and work close to the system. That is where the value-creation occurs for PRT. Just saving kms driven on the car is only worth 10-20 cents/km (8-15 cents/mile). I'm not sure how expensive parking is in downtown Houston. But avoiding the capital and insurance costs of a car is worth 30 cents and up / km, at least where I am. Those are largely binary costs, too, so once you sink that cost each year, your incentive to take transit drops considerably.

Getting people to the PRT system (if their destination is downtown) could be as simple as putting in some bus/HOV lanes along the highways and running commuter buses into the city right up to a high capacity station. That would require less upfront capital cost. Those buses could pick people up from several locations and run non-stop down the highway to the station. (I just skimmed through the wikipedia entry, and it sounds like Houston already has such a commuter bus system--pretty sophisticated, too).

What I'm struggling with is that if you need a few thousand rides a day to justify the station, you need a few thousand parking spots. That is a heck of a lot of real estate, and quite a bit of expense to build. I guess Houston already has some park-and-rides with commuter bus service. In that case, using a PRT system as a downtown circulator would add value to that bus system, as I imagine the last mile problem is the biggest disincentive to using the bus. Once the system takes off, I can definitely see it making sense to run PRT down the highways, especially since those park-and-rides are already there, and there are bus services bringing people to at least some of the park-and-rides from farther afield (not relying strictly on drive-up traffic).

You demonstrated in one of your posts that building PRT down a highway ROW could be done quite cost-effectively. Nevertheless, using precious seed capital to build guideway that replaces what sounds like a pretty effective commuter bus sounds suboptimal to me.

Does that sound more reasonable?

Dan said...

Cities that were largely built “post highway boom” tend to develop unevenly into many mini-cities that often out-compete downtown in most respects. In Houston, the odds of getting people to live without a car is nil. Downtown is currently not a smart location to open the kind of businesses needed to meet the domestic needs of those close-in communities. Urban decay is something many cities face, but that is a whole other subject, as is the role of transportation in fixing it. Anyway, mobility between these suburban mini-cities can become highly compromised, and is impossible to fix economically. Think of a spider web layout, with radial and circumferential routing. Each direction competes with the other for right-of-way. Some of the worst traffic, (and therefore the greatest need for a solution) is on the routes that circle the city, not the routes going downtown. No 2D solution will ever work because these routes have been cut by freeways, and the underpasses have traffic lights that must be timed poorly for those traveling perpendicular to downtown. Often, finding the most advantageous circumferential route involves jogging closer or further from downtown, so even cross-grain travel adds to what one might assume is downtown related traffic. This forces thousands of us run their cars at walking speed for a 10 or 15 minute portion of our daily commute.

For a system to make financial sense it needs to solve a problem worth paying for and must do it better than the alternatives. In many cities this “cross-grain” traffic problem is severe and basically impossible to solve, since the other forms of land transit (buses, shuttles) face the same roadblocks. Yes, it is more expensive to go further. But fares are based on distance traveled and potential time savings. It’s as if you have to cross a river by ferry just to a go few miles because there is no bridge. The business model is very straight forward. It’s like creating the only bridge between the two cities. Take PRT or sit in traffic. The only alternative is that the city can buy millions of dollars’ worth of right-of-way, build a cloverleaf intersection, and plow the highway right through what is often highly prized and productive commercial property. (The arterial route to the “mini-city.”) This is not the only compelling model for PRT, but it is the one whose challenges I was writing about in this post.

I agree that needing thousands of parking spaces is a struggle logistically. I did not say thousands per station. But that is why I have tried to design stations that could be as minimal as possible, including simply “landing” in an area. The more stations, the more places to coax scarce pedestrian traffic from.

We can put a station at the zoo, but how many of the thousands of people visiting that zoo on any given day will be able to take an embryonic PRT system all of the way home? A few dozen? Connecting to existing transit might give a few hundred more practical door-to-door service, but designating some parking at the zoo for the PRT customers would also be extremely useful, allowing it to become a gateway into the system. This would amount to seeding the other stations with pedestrian traffic. So it’s a combination approach. Something like a stadium comes to mind as very useful, because it could be both destination (during games) and a gateway with ample parking the rest of the time.

I do not argue with your premises, Andrew. What you suggest would make perfect sense in many cities. We, in Houston, have unfortunately solved our downtown transportation needs by eliminating places to go. Our tunnel system that has lunch places. It closes at 6pm. There are plenty of other cities that need more downtown mobility though, even in the US.

Juho said...

Even in the highway based cities of the USA there are places where people don't have cars. I mean airports. If we take a city that has a relatively compact centre, then travelers might be willing to change the tedious process of renting a car to a personal rapid transit to the hotel and city centre.

There are already now various innovative transport systems to/from and within the airports. PRT could survive there as well (and also will in one form at Heathrow).

My point is simply that one should start from the easiest targets. Airports seem to be a good place to start. Next one should convince the decision makers that in the other end of the airport PRT (city) there might be not just one station but multiple stops (and possibility to expand the system).