Sunday, October 31, 2010

Ogden "Streetcar"

Reading the comments section on newspaper articles never fail to enrage. Regarding the Ogden Streetcar, the comments are nothing short of slanderous.
UTA seems not to understand the difference and is pretending to be experts at streetcars, when neither they nor their consultants have ever planned a streetcar route. 
Unless you count the Sugar House Streetcar or the San Francisco F-line streetcar...

Some are merely ignorant:
..It brings revitalization within two blocks of the rails....
All the data on streetcar revitalization is based on a study done in Portland five years ago. Damn the souls of the consultants who did so, for using blocks as a metric instead of feet. Forever after, everyone else is condemned to making the distinction that Portland blocks are ONLY 200 feet long, which is about 1/3 the distance of an Ogden block.

Some are inane.
The benefits of fixed-rail street cars moving in an existing traffic lane are well-proven in other cities, and are a cost effective way to rejuvenate existing neighborhoods 
Note that he/she mentions no specific cities or street-care lines. Very few streetcar lines have actually been built, and even fewer studies done on their rejuvenation effects. To my knowledge, precisely three. All cite the presence of large parcels of underutilized industrial land and/or rundown low density residential as being critical. Having visited Portland prior to the streetcar renaissance, I can guarantee it had plenty of the former.

Admittedly, some people had thoughtful things to say:
The last I heard the proposed route is 36th Street. This is a two-lane but extremely busy route across town. To interfere with this WSU traffic artery would be foolish. House on both sides of the street would have to be moved, since the last time this street was widened the road was moved to to only a few feet from front doors. There is nowhere to expand without destroying houses. 24th, 25th and 26th street are already wide enough for a trolley line. Population downtown is more concentrated that at the south end of town.
What is missing in his calculation is the section of Harrison between 25th and 27th, which is too narrow to both keep UDOT's lanes for cars AND a streetcar, thus requiring the taking of historic properties.

Some are just blatantly wrong:
UTA is promoting a design that drastically over-engineers the guideway. The projections they are using are almost twice the cost of what they should be ($12-15 mil. per mile). 
That cost estimate comes from a redevelopment project in Kenosha, Wisconsin used to redevelop an automotive plant, and was built on donated right of way, around a public park, and is totally devoid of expensive intersection crossings. It is also notable that a) The Kenosha streetcar is no longer operating; and b) the total ANNUAL ridership of the streetcar totaled about 67,000---about the same as the WEEKLY ridership for the Portland Streetcar. Kenosha built the street-car on the cheap, and is paying the price for it. Estimates for other modern streetcar systems have come out at a very similar cost/mile of $22-30 million per mile.

This one made me think:
The guideway needs to be fixed, but according to the Obama admin. guidelines, it does not need to be exclusive to the streetcar- there is no need for a separate travel lane- but w/ the streetcars operating in the same lane as cars. This is how street cars work- in the flow of traffic- this is not light rail. 
On one level, he/she is right--there is no 'need' for a streetcar to have a separate travel lane. It works in Portland. But it works in Portland because it has 200' long blocks, so a parallel street is never more than 200' away. And as time goes by, Portland is increasingly moving toward dedicated guideway, using 'Bus-Only streets', and prohibiting turns across the light-rail/streetcar lines.

Saturday, October 30, 2010

Private Buses

The last few weeks at Cap'n Transit Rides Again and Human Transit have included an interesting discussion on private transit.

I've earlier written on the use of rail as a method for governments to experiment with different types of transit services. I do not mean to advocate the large-scale privatization of transit. For transit to function efficiently, it must function as a cooperative network. But that begs the question: Must the network paths be determined by some sort of central coordinating agency, or can a most-efficient solution emerge from the rough-and-tumble give and take of market competition?

One some level, the network effect is fundamentally valuable, and a private operator would recognize that connections to other transit services is the way to go. In unregulated developing countries, informal 'hubs' for bus, taxi, and jitney operations arise.

On the other hand, private operators don't go where there isn't any money. 'Coverage' services providing essential social services for seniors, the disabled, the young, and the poor... would almost certainly cease to exist. Providing para-transit costs UTA more, on a per-rider basis, then bus, TRAX, or FrontRunner. Largely because Para-transit functions as a sort of specialized taxi. While the price per ride is much higher than a bus ticket, ($4.00) it's still a superior alternative to taking a regular taxi.

Another problem is 'clustering'. Small private operators are bad at providing 'backbone' service, and thus tend to cluster around terminals for other high-volume services. Most taxi operators make their living giving rides to and from the airport, not running people around down-town Salt Lake City. Waiting an hour for the near-certainty of a pick-up at the airport beats out waiting an hour for the hope of a pick-up in downtown.

In New York city, small private jitney operators frequently compete with buses--following established bus routes. On one level, they compete with buses, taking passengers that would otherwise ride the buses. On another level, they act as an unofficial 'frequency boost', adding more vehicles per hour on the route.

For high frequency buses, where passengers arrive at random intervals and expect to be picked up within a reasonable period of time, jitneys would compete with buses. But for less frequent buses (on 1-hour, half-hour or even 20 minute headways), limited privitization of local bus routes would provide a mechanism for agencies to experiment with bus frequency without undertaking any long-term liability.

UTA could 'badge' private providers with licenses to operate along existing UTA routes, with the private providers being allowed to experiment with different schedules (and frequencies) to determine which ones are most effective.

Wednesday, October 27, 2010

Bad Routes and the Politics of Transit Funding

Clearly, many existing bus routes would not make a profit, and would be unable to make a profit--and yet they continue to exist because of an operating subsidy. In Utah, each city contributes to pay UTA, and on the basis of that contribution, they get a certain number of buses, or a certain amount of TRAX. UTA is pretty scrupulous about matching the amount of transit a city gets to financial contribution its tax assessment makes. More central cities (Murray) get more frequent 'through' transit service, while cities on the periphery (Riverton, Herriman) are forced to content themselves with thrice-daily commuter buses. 

In a sense, each city 'buys' transit service, and has some say about where that transit service goes in the city. But that 'say-so' is often contrary to technical considerations about alternatives offering the best ridership. If a city wants bus service to run to its 'historic downtown' as opposed to the new Wal-Mart on the edge of town, it is going to pay for that choice in lower ridership and lower fairbox recovery. 

Yes! Somebody gets it!

"There's strong emotions that more lanes means more traffic means more danger." - Taylorsville City Manager John Inch Morgan

It's to see somebody make the connection--wider roads don't just get you places faster--they also endanger the people along the way.

Sunday, October 24, 2010

University of Arizona

I was recently in Arizona, visiting the University of Arizona. It's an urban college campus, surrounded by the most shocking parking garages I've ever seen. Numerous six story parking garages. Divide the number of stalls by the total cost, and the average parking stall in a six story parking garage costs about $22,000. If each deck has 100 stalls, each garage costs about $22m dollars. And there were SIX parking garages. I remain aghast.

And this is just the 'cost of doing business', by providing an 'essential service' to students. I've got a better idea---BUILD A BRT. A nice one, a 'Rubber Tyred Light Rail' sort of BRT. Dedicated lanes, signal priority, electric buses with an overhead caternary (to eliminate that diesel-engine rumble), the works.

But therein resides one of the fundamental problems of financing transit systems. The parking garages were not built all at once-- U of A has doubtless been built one every five or ten years for several decades. You can't really build transit that way--transit relies on a network effect. The 'yield curve' for benefits is different. For parking garages, it's linear. For transit, it's exponential.  If you built $22 million of transit every 5 years, you would get:

  • 1/2 mile of light rail
  • 1 mile of streetcar
  • 2 miles of high quality BRT

But those are variable costs, ignoring the 'startup' costs of starting a new transit service--maintenance sheds, 'tail track' to the sheds for rail vehicles, signal control.... Rail is a transit sector with high barriers to entry. Lower marginal costs, but very high fixed costs.

Cost of land seems to be the straw that breaks the camels back. When a University looks at the cost of acquiring 5 contiguous acres for a new parking garage, which has to be within walking distance of campus, other alternatives are on the table. And more transit typically looks good, simply because a new transit system makes use of existing public right of way.

Friday, October 22, 2010

Dimishing Returns for Frequency Enhancements

I was reading this post on Cap'n Transit's blog.

One of the commenters posted:
a route that is by mistake operating at a profit is under-providing, and should have its schedule extended or frequency increased until it gets back down to break-even.
Typically, when a transit line increases either frequency or extended hours, that tends to increase ridership. It would be an open question if ridership was rising fast enough to cover the cost of increased operations. Beyond a certain frequency, I would tend to expect not. But I would anticipate that bumping frequency from 30 minutes to 15 minutes would more then double ridership.

The rule of thumb I've learned is that at 15 minute service, you start attracting more unscheduled trips, where someone will show up at the bus stop, and wait for a bus to come. It would be amazing to have some data to show under what conditions increases in frequency show diminishing returns. If ever.

*I've never seen any study data for this. (If anyone knows of any, please drop me a line.) It would be great to see a case comparison before-and-after study of a transit line's increased or decreased frequency. Sadly, changes to transit schedules or routes are rarely done in isolation for a single transit line. They tend to be done en-masse, making it difficult to disaggregate the effect. The Federal New Starts has made this even more difficult, as upgrades to frequency on a single bus line tend to occur as part of a general upgrade to BRT status, which includes a number of other improvements.

Wednesday, October 20, 2010


When the Medical Center Trax extension first came online, the WFRC model predicted heavy ridership. Within the model, it looks reasonable--there are a lot of jobs within a half-mile walk of the final TRAX station. But TRAX comes to the foot of the hill at University medical center, near research park. And it's a brutal hill, with something like a 30 degree slope. Needless to say, ridership was not quite as good as expected. (The extension met projections, but every other TRAX expansion has vastly exceeded ridership projections.) From Gallivan Center TRAX station, it takes an HOUR to make it the 20 blocks east to Huntsman Cancer institute. Riding TRAX from Main to 21st East takes a half hour of that time, meaning the 3 block slog from the foot of the hill to Huntsman Cancer institute takes a half an hour. And that's THROUGH parking garages, up elevators, along passageways. 

What to do? A brutal hill, several major hospitals. In sequence, there is Primary Children's Hospital, University Hospital, and the Huntsman Cancer Institute. Half a dozen parking garages. For doctors, nurses, janitors, food servers, and visitors. 

Utah is already home to a fair number of ski-resorts, who have resolved the issue of steep hills handily. The Canyons resort has a very nice one. And it covers quite a distance. 

Ogden studied using a Gondola pretty seriously, for a five mile corridor. The price tag for capital wound up being something like $10m per mile. Operations costs were higher that for alternate modes, a conclusion I find strange. At the Canyons, the gondola just has a couple of people at each end, where a bus or street-car needs a driver for every vehicle. 

The people-moving capacity used in the Ogden study is almost ludicrous. 10-person cars every 30 seconds? That would move 1200 people an hour at capacity. In comparison, a bus system running every 15 minutes provides 4 buses with 60 person capacity. Halving the frequency of gondola cars to every minute would reduce capital cost significantly. 

Tuesday, October 19, 2010

Average Density

As anyone who has read The Black Swan knows, averages are dangerous. Outside of natural phenomena, nothing follows the Bell Curve. But it's a frequently used metrix, especially in sustainability comparisons.

Pleasingly, someone has taken the time to actually depict the distribution of city density.

It's an impressive comparison, and it helps show why NYC does so much better on transit ridership than anywhere else--there is a LOT of high density there. If we were to cut the graph off at 10^4, and integrate the area under the curve past that point, NYC would beat out all the other cities combined.

Once the new census data is available in shapefile form, it's going to enable all sorts of new analysis. One of which might be doing a nation-wide analysis of the relationship between 'Transit Density' and actual transit networks.

Monday, October 18, 2010

Maximum Density

My brother sent me a fascinating link the other day.

While SimCity 3000 is a game (designed to be fun) rather then a model (designed to be useful) it has enough verisimilitude to be worth considering. Skip to the end, and consider the city with the highest density, and it's transportation network--the city relies entirely on subways, and devotes zero surface area to roads.

Cars take up an enormous amount of land area. Insuring sufficient parking means that there are four or five parking spaces for every registered driver. A standard parking spot is running about 200 square feet, once you count landscaping and space to enter and exit the parking space. About the size of a small studio apartment.

SimCity provides a very interesting perspective on development and scale. Unlike the real world, which is developed by multiple independent agents, a SimCity effectively has a 'single developer', who has the perfect cooperation of all municipal entities. Decisions that make sense in the microcosm don't always make sense in the macrocosm.

Saturday, October 16, 2010

Access Control

Access control is the phenomenon where traffic engineers close or limit the number of curb cuts providing access to a roadway, typically a high-capacity facility such as an arterial.

For automotive users, the benefits are less certain. In traffic movement, the ideal is 'laminar flow', where you have a steady stream of cars all moving at the same velocity, until you reach 'maximum saturation' of the lane, where adding an additional car causes the car behind it to slow down, in order to maintain a safe following distance. This in turn causes the second following car to slow down, and so on, rippling back down the line of cars.

Traffic engineers apply an equation to calculate delay based on a number of factors, most notably # of lanes and facility type. Facility type covers a multitude of geometric design factors--curvature, slope, turning radius, speed, etc.

At some point, someone realized that roads with a large number of access points did not move traffic as well as those without. On this basis, it seems logical to reduce access points, because each access point represents a disruption in laminar flow, slowing the speed of traffic. And so traffic engineers began to close access points, to improve traffic flow.

But this is treating the symptom, not the cause. Why does the road have so many access points? Because there are that many business accessing the road?

But closing access points is treating the symptom, not the cause. Acme Widget Co. has 300 employees, and closes at 5 PM, five days a week. Their parking lot has three access points onto the street. 150 cars leave using each exit. Cut it down to one access point, and you have 450 cars leaving via that one way. You have fewer disturbance points to laminar flow, but bigger disturbances.

Thursday, October 14, 2010

The Bad Guys

Randal O'Toole
Joel Kotkin
John Stossel

I'm not saying they are bad people, or even that they are wrong. But I am saying that they have an agenda, and it is an anti-transit agenda. The are radicals and extremists (which is why newspapers like to quote them).

Also suspect for their anti-transit agenda:

Thoreau Institute
Cato Institute
Sutherland Institute

Tuesday, October 12, 2010

Capacity Equilibrium

In California, in San Francisco, they replaced a 3-person carpool lane with a 2-person carpool lane, because it was vacant most of the time. Makes sense, right? Lower the limit, more people will carpool, and the lane will be filled up with carpoolers!

This is exactly what happened on the first week. One the second week, all the new carpoolers realized that the carpool lane was just as congested as a regular traffic lane, so there was really no point in carpooling anyway. And fewer people carpooled until the lane reached some kind of equilibrium, where it made sense for people to carpool again.

So, no problem, right?

Wrong. One lane of 2-person carpoolers is worth as much lane capacity as two lanes for regular cars. One lane of three person carpools is worth three lanes of single-rider cars. Switching from a 3-person to a 2-person carpool effectively sacrifices a lane.