14 February 2014

Buses v. Rail In Colorado

FasTracks To Longmont

The Denver Post editorial board makes a convincing argument today that Bus Rapid Transit (BRT) from Union Station to Longmont makes more sense that the commuter rail services originally planned as a part of the FasTracks proposal. I quote it in full in order to present readers of this blog with their opinion on this political issue, mostly summarizing a study of the issue done by RTD:
Like many supporters of FasTracks, we argued for years that folks in the northwest suburbs deserved the commuter rail they were promised and pay taxes for. However, recent cost and ridership projections are another sobering call to reassess that stance.

Total daily ridership on bus and rail would increase by only 300 riders by 2035 if the $1.15 billion line were built to Longmont, according to a recent analysis by the Southwest Energy Efficiency Project of RTD's Northwest Area Mobility Study. Not only that, travel time actually would be longer on the train than in rapid transit bus — 71 minutes on the rail from Longmont station to Union Station compared to 53 minutes by rapid bus.

Eighteen miles of U.S. 36 is already being rebuilt to handle bus rapid transit, or BRT, between Boulder and Union Station, a $219 million project expected to be completed by next year.

Meanwhile, last fall's mobility study projected significant improvements in ridership if BRT were expanded to include six main arterials in the northwest. The cost of building those routes would be an estimated $350 million, providing exclusive lanes where feasible, dedicated boarding areas and buses coming every 10 to 15 minutes. By contrast, rail would have to share the lines with BNSF Railway trains, meaning commuter trains would be spaced every 30 minutes. With BRT on the arterials, daily ridership is predicted to increase by as many as 18,900 people.

Those BRT routes would be along Colorado 119 from Longmont to Boulder; along U.S. 287 from Longmont to Broomfield; on 120th Avenue through Broomfield; on South Boulder Road from Boulder to Lafayette; on Colorado 7 from Boulder to Interstate 25; and Louisville to U.S. 36.

We understand northwest voters overwhelmingly supported FasTracks, but it's time to acknowledge the ridiculous costs for such little benefit and start looking at BRT as a workable solution.
The issue of bus v. rail is as much an issue of how culture interacts with transportation technology as it is one of the technology itself.

The stereotype (which is often true), is that city buses appear at infrequent intervals making it critical to know when it is supposed to stop, are often early or late, are slow because the take on and less off riders at stops that aren't very far from each other, and that bus systems are hard to understand.  Equally important, the stereotype s that city buses, in practice, are used predominantly by people so extremely poor that they can't afford a car, people so reckless that they have lost their driver's licenses, drunks, disabled people, and large groups of unruly high school students.  This ridership can make riding a bus, especially outside of rush hour, seem unsafe. City buses also have a reputation for being dirty, bad smelling, uncomfortable, ill maintained and short on privacy.

Rail service, in contrast, tends to appear at stops frequently (making it unimportant to know the exact time that it makes particular stops), tend to be faster because they are immune to road traffic and make fewer stops than ordinary city buses, have more acceptance from middle class riders, and tend to be cleaner, more comfortable and in better shape.  Rail still means a sacrifice in privacy, but public support for rail v. buses mostly flows from these often true stereotypes about how the service is provided.

There is an environmental issue.  Traditionally city buses run on smelly diesel fuel which is a scarce resource that will run out someday and which we import.  Meanwhile, urban rail systems are often electrically powered which is perfectly clean at the train's location and ultimately as clean as the sources of the power grid (which in Colorado is growing much more renewable and greener and doesn't rely to any meaningful extent on foreign fossil fuel imports).

The bus, however, gets a bad rap in this regard.  Buses can and do operate base on a variety of alternative fuels, and even when they are running on diesel, even a mostly empty bus uses less diesel per passenger mile traveled than any car or truck you can buy.  Running full, buses get better mileage per passenger mile traveled than the most fuel efficient cars ever made (well over 100 mpg for a single occupancy vehicle), and better even than the MPG equivalent of existing mass production plug in electric cars.

Bus rapid transit tries to address these issues with frequent service that makes learning the bus schedule irrelevant, having a few simple routes like a light rail system, having a nicer passenger space than an ordinary city bus, and having dedicated bus lanes and few stops that make the service much faster.  The up front infrastructure costs are much lower for BRT than rail, the operating costs aren't vastly higher for BRT than rail (although probably a little bit higher in the long run since BRT takes more drivers per passenger), and one hopes, for those reasons and perhaps due to a higher bus fare, would have a less seedy clientele than the stereotypical city bus.

The examples of working BRT systems are mostly from Latin America where it was developed and I'm not aware of great urban U.S. examples of these systems, although they may be out there.  But, it has the potential to be just as good or better at a lower cost and so it deserves a look.

Proposed State Funded Intercity Bus Service

[Note:  Some particulars of the post below regarding this program are inaccurate because I did not have access to the December 13, 2013 Power Point presentation in the first link below until after most of this was written and instead had to rely on secondary sources, but the general gist of the conclusions remain sound.  I will try to correct this post, but please click on the link before citing this post as authoritative.]

Meanwhile, the Colorado General Assembly is considering a bill to have the state operate six round trip express bus runs each weekday from Denver to Colorado Springs, five each weekday from Denver to Fort Collins, and one each weekday from Denver to Glenwood Springs (a total of about 6,240 one way trips per year).  This commuter service would be at least as fast as Amtrak in the mountains (at least when the roads are not closed or in gridlock), and would be faster than comparable Greyhound routes.  The CDOT proposal as of December 13, 2013 is here.  Fares would be $10 to Fort Collins, $12 to Colorado Springs, $17 to Vail and $28 to Glenwood Springs (with multi-ride subsidies available).

The proposed initial appropriation for this service would be $10 million over four years.  This would pay for the purchase of the buses and other capital expenditures, with fare revenue exceeding operating costs within four years.  This would be a temporary subsidy of about $8 per passenger trip.  This is about 25% of the subsidy for Amtrak service.  The per trip operating subsidy on regional bus routes within RTD is about $7.

Assuming that capacity and 60 passenger buses, the service would provide 318,240 one way passenger trips per year.  Assuming that the service was used mostly by people commuting between these cities, it would serve 306 regulars between Denver and Colorado Springs (about 68 miles), 255 regulars between Denver and Fort Collins (about 65 miles), and 51 regulars between Denver and Glenwood Springs (about 157 miles), for a total of 612 regular commuters.  Pueblo is 44 miles beyond Colorado Springs.

Predecessor Service

This would replace the FREX and FLEX bus systems that had ridership at 85% of capacity.   The central issue is money.  FREX was discontinued on August 31, 2012 due to a lack of funding support from the transit system that serves Colorado Springs.  FLEX is still in business, however, and goes from Fort Collins to Longmont and back.  The CDOT plan would provide substitute service with substitute funding.

Alternatives: Amtrak and Greyhound

Amtrak currently has one trip per day from Denver to Glenwood Springs on the Colorado Zephyr that takes 5 hours and 48 minutes (leaving a little after 8 a.m. and arriving a little before 2 p.m. in the unlikely event that the train is on time) with no intermediate stops, and there is a corresponding trip each day back to Denver at a cost of about $54 for a pre-reserved coach seat and $162 for a roomette.

As of 2012, 113,393 passengers used the California Zephyr service through Denver (down 20% from the 2011).  The marginal cost of simply running the California Zepher train without considering infrastructure or station and booking costs is $41.73 per train mile as of fiscal year 2013.  The California Zephyr line, as a whole, generates more losses than any other Amtrak line at about $30 million of operating cost losses per year (this source puts the figure at $62 million a year).  Approximately 55% of its operating expenses are paid for with federal subsidies (as of 2011) implying a subsidy of about $27 per passenger trip from Denver to Glenwood Springs.

Ending California Zephyr service would only impact a modest number of highly subsidized, very slow trips, and ownership of the rail right of ways, if it could be obtained (it might not as Amtrak shares rail lines with freight in the West) would be a valuable asset for future high speed rail projects in Colorado.

A Greyhound bus trip on the same route takes just under seven hours on a gridlocked Friday night, and has a standard fare of $54.  The proposed CDOT service might very well entirely kill Greyhound's business in this corridor (and possibly much of Amtrak's business as well).

High Speed Rail

These are the same lines that proposed high speed rail would serve in Colorado, but at a high cost of about $5.4 billion would be for the Fort Collins to Denver to Colorado Springs portion and about $13 billion from Denver to Eagle (which isn't all of the way to Glenwood Springs).

Even at a modest 2.5% long term interest rate with no repayment of principle, this is a $369,863 per day commitment of infrastructure cost for the Fort Collins to Colorado Springs portion.  Assuming that an infrastructure subsidy in the range of $10 to $100 per trip would be acceptable for this kind of service, a ridership of 3,700 to 37,000 trips per day would be required to make this a sensible investment.

On the same basis, this is an $325 million a year commitment of infrastructure cost for the Denver to Eagle rail portion (which doesn't even get you to Glenwood Springs), which would have a much more seasonal pattern of ridership and a much higher weekend than weekday ridership.  This would require ridership of 3.25 million to 32.5 million trips per year, with peak ski season weekend ridership of 50,000 to 600,000 or so trips per day.

A June 12, 2013 story in the Summit Daily Newspaper reports on current high speed rail ridership estimates that were lower than in an earlier study:
Officials close to an ongoing Interregional Connectivity Study (ICS) — commissioned to investigate the plausibility and cost of a rail servicing central Colorado — are realizing earlier expectations of the number of people who would use such a system were over estimated by more than 100 percent.

“We’re a little bit disappointed with the ridership numbers,” said Mike Riggs, a consultant on the ICS study, who gave a presentation on the progress of the project in Silverthorne Tuesday evening. “They’re not as big as we’d hoped.”

A study completed in 2010 by the Rocky Mountain Rail Authority (RMRA) indicated a high-speed train providing service along the Front Range and out to Eagle County could attract as many as 7 or 8 million passengers per year, whose fares would help fund the operation and management of the system. But using what Riggs called a more open, transparent and ultimately accurate model to gather data, officials conducting the ICS study are finding actual ridership will likely be closer to 3 million per year. . . . Models vary slightly based on the type of train — faster trains can cover a commute in less time and would attract more riders — but with a presumed fare of just under $24 per person, riders would generate . . . only enough money to cover roughly 85 percent of . . . . [what] it is expected to cost to run the system. The earlier RMRA study anticipated fares would cover 150 percent of operation and management costs, charging less than $20 per person for a train ticket.
With 3 million trips a year, the infrastructure subsidy only would exceed $100 per trip assuming a very modest 2.5% interest rate per annum and no repayment of principle for this very long term infrastructure item, in addition to any operating cost subsidy.  This is a very substantial government handout to benefit travelers who would presumably mostly be affluent skiers on vacation.

On the other hand, a presumed fare of $20-$24 per person sounds too low when people are currently willing to pay $54 to take much slower trips on Amtrak and Greyhound from Denver to Glenwood Springs.  High speed rail would make the trip from Golden to Eagle in an hour in a more comfortable and modern cabin, when it takes about 75 minutes to make the trip by car or shuttle on a day with almost no traffic, and two to three hours to make that trip when I-70 is clogged with ski traffic.  An hour is tough to match even if you fly.  So, I suspect that the market would still easily bear a high speed rail fare from Golden to Eagle of $60, which would easily cover all operating costs and make a dent in the infrastructure expense as well, without seriously reducing ridership.

Another recent study puts a 120 mile per hour mag lev high speed rail from Golden to Eagle (118 miles) at $13.5 billion, which is very similar to the previous study.  But, it also imagines that a comprehensive I-25 corridor 180 mile per hour high speed rail system from Fort Collins to Pueblo (177 miles) at $14 billion.  I'm skeptical that it would be that expensive to build a high speed rails system in the I-25 corridor, although ridership levels would probably be higher in that high population density corridor (at least up to Colorado Springs).

Interstate express bus service on I-25 and from Denver to Glenwood Springs is a baby step that is definitely worth a try, because it can be done now, involves only a modest financial commitment, and would provides a benchmark against which additional ridership and trip speed on high speed rail systems could be tested.

Footnote Re: The Ski Train

There is also a ski train from Denver to Winter Park during ski season that carries 30,000 passengers a year.  This is a separate service from the Amtrak California Zephyr service, is not part of the Amtrak system, and operates only seasonally.

No comments: