Posts about Amtrak

The plaza was once a wasteland of traffic lanes and hadn't been properly maintained for years. The reconstruction project, which included Amtrak and multiple local and federal agencies, can make it an attractive and welcoming gateway to DC. The design treated all forms of traffic well—pedestrians, bicycles, cars, and bus.

Unfortunately, the police parking helped cause damage to the curbs and sidewalks before, and will do so again if this practice continues. It's highly unlikely that the sidewalks and curbs were reinforced strongly enough to withstand the pressure from mutli-ton vehicles.

It's easy to see a parallel with other examples of infrastructure in DC, like this story about Woodson High School, where we spend big bucks on nice new things, then fail to maintain them right after the construction is done.

Police presence and patrols are necessary in highly visible, active areas like a major train station. There has to be a better way and some better locations for police to leave their vehicles, though.

Without long-distance trains, Amtrak is profitable? Not really

The report crunches the numbers on how many riders each route carries, how much it costs to run the trains, and how much revenue they generate. A number of people read the numbers to conclude that if Amtrak cut its 15 longest routes, Amtrak would operate in the black.

But this depends on how you define "in the black" or "operating profit." Cutting those trains would eliminate the need for a federal subsidy, but the states also contribute money for short-distance trains, and Brookings counts that as "revenue," not "subsidy."

The Adirondack, for example, runs between New York City and Montreal, via Albany. The report says that this train has a positive balance of $1.3 million. Amtrak makes a profit on it!

Not really. It costs $13.3 million to operate, and the train earns $7.0 million in revenue. That equals a loss of $6.3 million, but New York state pays Amtrak $7.6 million a year to operate the train. Add that in, and the "balance" ends up being a positive $1.3 million.

All told, the states fund Amtrak to the tune of $190.5 million a year. Counting that but no federal payment, Amtrak would have ended up with a surplus $30.9 million without the long-haul trains.

Even this doesn't mean the short-distance trains can be profitable. The report doesn't include the major cost of capital maintenance on the Amtrak-owned Northeast Corridor, which passenger revenues will never be sufficient to cover. Projects like the Gateway Tunnel, which will add capacity into Penn Station and new rolling stock to replace aging cars are also capital costs that can't be covered by the operating profit alone.

Amtrak's Adirondack in Westport, NY. Photo by The West End on Flickr.

Coverage or ridership?

One of the biggest problems with Amtrak from a political perspective is that people don't agree about the goal of the railroad. Is Amtrak supposed to make a profit, as conservatives tend to insist? Or is Amtrak supposed to provide a transportation service to much of the nation, as liberals tend to claim?

The Brookings report makes it clear that there really are two Amtraks.

One of the Amtraks is efficiently providing frequent short-haul service within or between metropolitan regions, such as on the Washington-Boston Northeast Corridor, Los Angeles-San Diego Surfliner, and Chicago-St. Louis Lincoln Service. A few services like these can run operating profits, but most still don't even if they're successful.

The other Amtrak provides basic service to other parts of the country. This service was never meant to make money. It was meant to include more states in the system and provide another (or in some cases the only inter-city) transportation alternative to rural communities.

Lawmakers did intend for the company to be profitable, despite evidence that it would not be, but that was changed in 1978 when it became clear that was unlikely to happen.

In this respect, Amtrak is like many transit systems. When Congress created Metro, they assumed it would run self-sufficiently without government support, too, but that didn't happen either. No transportation system, not roads, rails, or aviation, actually makes a profit when you incorporate all of the infrastructure.

Amtrak is a basic system plus extra short-distance services

The way Amtrak was set up is an important part of understanding the current situation.

In 1971, as Amtrak was coming together, a basic system was drawn on a map. This base would provide some minimum level of service across the nation, and be funded initially through federal subsidies and ticket revenue.

The Pacific Surfliner is a good example of a state partnership that grew out of the base system. In 1971, there were 3 roundtrips between Los Angeles and San Diego. These 3 roundtrips were part of the Amtrak basic system.

By 1976, California wanted more service, so it paid to add a 4th roundtrip. This train was specifically a California-subsidized train, while the other 3 were Amtrak-subsidized trains. Over the next 2 decades, California continued to add trains to the corridor, and Amtrak added 2 more to the basic system.

In 1995, California and Amtrak agreed to end having some individual trains be part of the basic system and others state-subsidized. Instead, California and Amtrak would split the cost of the San Diego-LA service proportionally. At the time, California covered 64% and Amtrak covered 26% 36%. It's now 70%/30%.

This seems like the best approach. Amtrak, in consultation with the US Department of Transportation and the states, and input from stakeholders, including organizations representing passengers, should determine what the basic federal system should look like. If the states want additional service beyond that amount, they can pay for it, perhaps with federal assistance in the form of competitive, merit-based grants.

Several successful services have grown out of arrangements like that, including the Surfliner, the Charlotte-Raleigh Piedmont Service, and the Cascades between Eugene, Oregon and Vancouver, British Columbia.

Congress is shifting costs to the states

However, the funding environment is changing.

Under Section 5 of the Passenger Rail Investment and Improvement (PRIIA) Act of 2008, starting in October, California will have to cover 100% of the cost operating deficit of running the LA-San Diego service. That's because all Amtrak routes less than 750 miles in length must become state-supported or be terminated.

The Brookings report supports shifting almost all costs deficits to the states for long-distance Amtrak trains as well.

That's the wrong approach, and would likely mean the end of most of the long-distance trains.

It is difficult to expect all the states served by each of the long-distance trains to support their own routes consistently, much less to agree on schedules, service amenities, and cost allocations. Many states are already facing a huge challenge in coming up with the funding to keep existing short-distance service running under the PRIIA mandate.

Just as governors in Ohio and Wisconsin blocked high-speed rail funding, some states will refuse to pay. Many have no history of supporting train travel at all in the modern era, and have conservative or divided legislatures. Long-distance routes would either have to stop at the state line or run through without stopping within it, neither of which makes for a useful transportation service.

Other states have no history of supporting Amtrak routes financially, and conservative or divided legislatures would make it unlikely that those states would step up to the plate to fund what has so far been a federal commitment.

Right now, several states are having the discussion about whether to keep their short-haul trains. While New York, California and several other states have budgeted the required funds to keep their short-distance trains running in fiscal 2014 (starting October 1, 2013), Pennsylvania could cut its Pittsburgh-Philadelphia/New York train and Indiana could lose the Indianapolis-Chicago Hoosier State unless their citizens convince state lawmakers to appropriate the needed funds

We need a national network

It's important to keep Amtrak's long-distance trains, even though they're not profitable. A recent white paper from the National Association of Railroad Passengers elaborates on many points.

One major reason is that the long-distance trains and shorter ones fit together into a system. They're not completely isolated.

The Southwest Chief might run over 2200 miles across the nation. But many riders are not going all the way from Los Angeles to Chicago. Some are only going between Los Angeles and Flagstaff. Others ride between Albuquerque and Trinidad. And Kansas City to Chicago is a very popular pair of stations on the route.

Brookings seems to think that 400 miles is where passenger rail stops being competitive. And that may be the case. But just because the train goes more than 400 miles doesn't mean that the passenger has to.

From the whitepaper by the National Association of Railroad Passengers and the Midwest High-Speed Rail Association.

For example, the report lists the Bay Area to Sacramento Capitol Corridor as one of the examples of a good corridor. The same corridor is also covered by the Coast Starlight, which continues north of Sacramento and south of the Bay Area.

Further north, the Coast Starlight overlaps with another success story from the report: the Cascades, which runs between Eugene, OR and Vancouver, BC. It provides an additional frequency on both corridors, and connects them with each other and points in between.

Having long-distance trains also creates a market and proves the demand for short-haul trains. One of the 3 profitable routes in the system, informally dubbed the "Lynchburger," only exists because of the longer New Orleans-New York Crescent. People wanting to take that train between Lynchburg and Charlottesville and Washington were having trouble getting a ticket because the Crescent would sell out frequently.

As a result, Virginia decided to pay for a new train to run between Lynchburg and New York/Massachusetts. It's proven to be so popular that it actually covers its costs with ticket sales. And Governor McDonnell has proposed funds to extend the train to Roanoke.

But the Lynchburger probably wouldn't exist if the Crescent hadn't demonstrated the demand. It's also very difficult (though not impossible) to get the host railroads to agree to passenger trains where they don't already run.

There are many other reasons as well. Having a national network also makes possible many operating efficiencies, such as the ability to move equipment to other parts of the system to meet demand, which would otherwise be lost.

Besides, ridership on state-supported short-distance routes has only grown so much because state investment has translated into increased capacity. If a similar investment were made in long-distance trains—meaning additional frequencies or longer trains—their ridership would soar as well.

Report is useful for its data, but reaches the wrong conclusions

The Brookings report provides a wealth of insight into Amtrak's operating costs and revenues. But the report is misguided in its suggestion to turn the primary responsibility for the basic national system over to the states.

Passenger rail is an essential component of our transportation network. The 55% increase in ridership since 1997 is an indication that more federal and state investment is needed, not less.

Improving service on the long-distance trains will lead to ridership increases just as improvements to the short-haul trains did. Now is not the time for the federal government to waver in its commitment to passenger rail.

In the map, each circle represents one Amtrak station. The larger the circle, the more riders there are at that station.

Cross-posted at BeyondDC.

Union Station's tracks don't line up VRE and the Brunswick MARC

At Union Station, VRE trains coming from the south enter the station from the First Street Tunnel, at the far eastern end of the station. But the Brunswick Line's tracks come into the station at its far western end.

A Brunswick train crossing from the CSX Metropolitan Subdivision tracks on the west side of the station to the through tracks on the east side would have to cross all the tracks at Union Station, blocking trains on the other MARC, Amtrak and VRE lines.

That doesn't mean crossing over is impossible. It just makes scheduling more difficult. If in the future, for example, Acelas leave (and arrive) every 30 minutes and Regionals leave/arrive every 30 minutes, and MARC Penn trains leave and arrive every 20, there aren't many gaps for trains to cross over, and doing so can cause delays.

The Penn Line trains wouldn't face this problem, and as the highest-ridership line, that is the most logical one for through-running. But this is also the line with the most high platforms and thus the greatest incompatibility between MARC and VRE.

There isn't enough reverse-direction capacity

The Penn Line tracks do line up with the First Street tunnel; that's what Amtrak trains use to get to Virginia. Those trains could continue south to Alexandria, or even farther south. This could be a great asset for Penn Line riders, who could stay on the train to L'Enfant Plaza if they want the Metro Orange, Blue, Yellow or Green Lines, or ride to Crystal City or Alexandria if they work at jobs in Virginia.

Unfortunately, this would run into a second problem. There isn't enough track capacity south of Union Station. Right now, there are 2 tracks across the Potomac on the Long Bridge, and 3 through the L'Enfant Plaza area. CSX controls these tracks, and doesn't let VRE use all of the tracks.

Right now, VRE trains run almost entirely one-way. The trains head into DC on one track in the morning, and back out to Virginia on one track in the afternoon; in between, the trains sit at a yard near Ivy City in DC. The VRE schedule lists just one reverse-direction train, on the Manassas Line each morning and evening. Amtrak's trains go both ways all day, but there are only a few of those and mainly not at rush hours.

The bottom line is, if Virginia and Maryland wanted to have all or even some Penn Line trains continue past Union Station at least to Alexandria, there wouldn't be enough track space.

A similar problem applies to letting VRE trains continue north of Union Station. CSX has resisted letting Maryland's MTA add more MARC trains on the Brunswick and Camden Lines without also demanding Maryland invest some money into improvements along the route.

Stations are limited

Another issue with through-running is the design of the VRE stations at L'Enfant and Crystal City. These stations each have just one platform on one side of the tracks. That means trains can only serve the stations in one direction at a time.

This means the reverse peak direction trains on the Manassas Line each morning and afternoon can't stop at Crystal City or L'Enfant, because the trains running in the peak direction are using the platform track.

If Brunswick (or other MARC) trains could run south of Union Station today, they couldn't stop at L'Enfant or Crystal City, which is where most of the MARC riders would likely want to go.

Most of the other stops on the Manassas Line have the same configuration, with platforms only on one side as well. This will prove to be an obstacle for additional reverse peak direction trains, whether they're MARC or VRE.

Maryland Ave plan and Long Bridge study could fix this

There is hope on the horizon. The recent Maryland Avenue study recommended building a fourth track at L'Enfant Plaza. CSX could then let passenger trains travel in both directions at high frequency in that area. It would also transform the station from just one platform to 3, with a combination of high and low platforms for both types of trains.

Proposed L'Enfant commuter rail station layout. Image from the DC Office of Planning.

The next bottleneck would be the 2-track Long Bridge across the Potomac, and there's another study going on for that. VRE's 2004 strategic plan recommended adding a 3rd track from the Long Bridge to Crystal City, where the line widens to 3 tracks, and giving the Crystal City station an island platform to serve trains in both directions.

Combined with needed improvements at Union Station, we might one day see a truly regional rail system, at least from Baltimore and maybe Frederick to Alexandria, alongside more frequent service from Fredericksburg and Manassas to DC. To make this happen, however, Maryland and Virginia will have to make it a priority. With an 8-year-old VRE plan and a 5-year-old MARC plan mostly collecting dust, riders will need to push their leaders to put resources into commuter rail.

Why does VRE use low platforms?

VRE trains operate on two lines south of Washington that are owned by freight railroads. Because the freight railroads (CSX for the Fredericksburg Line and Norfolk Southern (NS) for the Manassas Line) own the tracks, they get to have a say about what types of platforms can be built. And that means low platforms.

Low platforms are typically placed at about the height of the top of the rail. The reason freight railroads want these types of platforms is because freight cars are wider than passenger cars, and high platforms could intrude into the dynamic envelope of a freight train.

As long as VRE operates on rail lines that are predominately freight railroad corridors, it will be stuck with using low platforms. There's not really anything wrong with that, except that in this case, it's an obstacle to through-running.

Railcar design matters

For passenger railroads using low platforms, there are 3 basic types of double-decker cars in use. Two of these offer access to the lower level with just one or two steps up from the platform, and can therefore be fairly efficient in boarding.

They're also much easier for mobility-impaired riders to board. Platforms can be designed to have a small area at the height of the floor, set back from the tracks. When necessary, a bridge plate is used for wheelchairs.

Left: A bi-level commuter coach with low-level boarding in Minneapolis. Photo by the author.
Right: An Amtrak California double-decker car. Photo by Wayan Vota.

Many cities outside of the Northeast use these cars, with the Amtrak California cars operating in Southern California and the Bay Area. The bi-levels are used in places like Seattle, San Francisco, and Miami.

VRE uses gallery cars, the third type. These cars have 2 levels, but passengers must climb 4 steps to board even just to the lower level. But the stairs are fairly wide and are not as steep as the stepwells on high-platform equipment. Gallery cars cannot use high platforms at all.

VRE gallery car. Photo by Elvert Barnes on Flickr.

In order to speed boarding, VRE could easily move toward either the bi-level or Amtrak model. In fact, for a while, VRE was using a set of bi-levels leased from Seattle's Sound Transit.

What's wrong with using high-platform equipment on VRE?

If MARC can use high-platform equipment, why can't VRE? Well, VRE could use high-platform equipment, and in fact, they have done so in the past. But using high-platform equipment at low platform stops is inherently less efficient.

Using high-platform equipment means that conductors have to manually open each door (instead of automatic doors in the whole train) at low-platform stops. That generally means that only one or two doors aboard the train can open, as is the case with the MARC stops at places like College Park and West Baltimore.

Additionally, boarding a high-platform train from a low platform means ascending a narrow, steep staircase. That means it takes longer for passengers to board and alight.

Similar stairs on an Amtrak train. Photo by the author.

What about MARC?

Like VRE, MARC runs 2 of its lines on tracks that are primarily used for freight. For that reason, most of the stations on the Camden Line and all of the stations on the Brunswick Line have low platforms.

But for MTA, the agency that operates the MARC commuter trains, it makes a lot of sense to have a fairly standard fleet because it can then move its cars between lines as necessary.

MARC's busiest line, the Penn Line, operates on Amtrak's Northeast Corridor, and most of those stops have high platforms. This is due to the fact that Amtrak (and the previous operators) have long used single-level railcars for their services. For single-level cars, high platforms are the only way to have step-free access to trains.

On the Penn Line, south of Baltimore, only West Baltimore, Halethorpe, and some of the platforms at Union Station are still low. Halethorpe is currently undergoing conversion to high platforms, and Amtrak recently announced plans to convert most of the platforms at Union Station to high platforms.

On the Camden Line, Greenbelt and Camden stations have high platforms. The other stations all have low platforms. On the Brunswick Line, all stations have low platforms.

What does this mean for through-running?

The practical effect of this arrangement is that through-running trains from either MARC or VRE will be difficult.

The current fleet of VRE trains are not able to use high-platform stations, which means they can't operate on the Penn Line to Baltimore. VRE trains also can't operate on the Camden Line, because they would be unable to serve the terminal station in Baltimore, since it only has high platforms. And that leaves the Brunswick Line as the only viable line they could operate on.

But running VRE trains on the Brunswick Line presents some other challenges: the way tracks are laid out in DC. Tomorrow, we'll look at those issues.