BL&OB Rail Trail

After years of circling the drain, the Somerset Generating Station will close on February 15, 2020. Since its railroad line will no longer be necessary, a 15 mile Buffalo, Lockport, and Olcott Beach (BL&OB) rail trail should take its place!

https://twitter.com/skyjacked793/status/419536663908347904

Opened in 1984, Somerset Generating Station was a landmark of my childhood. Its tall smokestack would billow puffy white exhaust, always drifting with the wind off of the lake and we would often see it from Olcott while getting popcorn. My dad told me that it was one of the more efficient plants worldwide and most of the white exhaust was steam. Unfortunately, the pollutants were still in the steam which will not be missed. Economics finally caught up and next year, it will be the last coal power plant to close in New York State.

In the 1960's and 70's, the lakefront site was eyed for a nuclear plant but only one was built further down the lake in Oswego. By the late 1970's though, nuclear plant construction stopped and instead, a coal plant was built. Coal power plants require too much coal to be delivered by truck so some old railroad track was reactivated. Buffalo, Lockport and Olcott Beach Railway (BL&OB) was built around the turn of the century as an interurban electrified line to connect Lockport and points south to Olcott Beach. A favorite summer destination in itself, people could also transfer to a steamship to Toronto. By 1937, passenger volumes had declined and the line was abandoned for the next 46 years.

In 1983, the former BL&OB line was rebuilt north of Lockport, renamed the Somerset Railroad, and connected to a short section of the New York Central's abandoned "Hojack" line. A large curving concrete bridge was built over Eighteen Mile Creek and several bridges grade separated the line from some of the busier roads. A spur connected the line to the power plant. The Someret Railroad is owned by CSX and uses their locomotives to pull the coal trains as Somerset has no locomotives or crews of their own. There are several small chemical plants and other industry served by spurs in Lowertown Lockport. A data center is proposed to replace the plant and will not need freight rail service. North of Old Niagara Rd, there are no other industries on the line which makes that 15-mile section perfect for conversion to rail trail since without the power plants or other industries along the line, CSX will likely abandon it after the power plant clean up.

Costs of constructing the rail trail are not exorbitant and would be eligible for grant funding. First, the County should acquire the line north of Lockport from CSX. Then, the rails can be removed and sold for a profit while the ties will have to be recycled for a net loss as recycling costs have climbed in recent years. The County could then chose a trail paving material, fine dust gravel being the cheapest but asphalt being more expensive and longer-lasting. The County could also construct several parking lots to increase access by leasing or buying land from adjacent owners or constructing narrow parking lots on the former railroad right of way.

While some parts of the County have trail access, especially communities on the canal, Newfane and Olcott do not. The trail terminus in Lockport would be a short ride away from the Erie Canalway Trail on low volume streets, expanding the number of potential users. Rail trails are popular community amenities nationwide and locally too. This trail's potential connection to the lake would attract lots of users and quickly become a popular community amenity. The County Parks Department, the County Legislature, and local jurisdictions should begin the planning process to acquire the railroad right of way from CSX and begin applying for grants. The BL&OB Rail Trail could be a gem of the community, connecting residents to Lake Ontario and beyond, and help the County recover from the loss of its largest taxpayer, the Somerset Generating Station.

Boosting Buffalo Metro Rail

While New York City has one of the largest subways in the world, Buffalo’s Metro Rail has never attracted the number of riders it expected. Running more service could attract more riders at a low, incremental cost.

For many years, US transit agencies have run a lot of service at rush hour and comparatively little outside of rush hour. The 9 am to 5 pm commuter has been king since the 1950’s when most private transit companies were purchased by local governments. And that approach has attracted generations of commuter riders. But in many countries, people going to and from work are only part of their ridership. Instead, they run frequent service at all hours to attract other trips such as people running errands, visiting friends and family, or going out on the town. Frequent service means that people traveling outside of rush hour do not have to wait for a train or even check schedules. In fact in Moscow, the Metro authorities do not even publish schedules and instead tell their drivers to follow the train in front of them by 1.5-3 minutes. Running more service is a key reason by other countries attract far more riders than their US counterparts.

From David Wilson, 20150827 61 NFTA Light Rail @ Fountain Plaza

Buffalo Metro Rail runs just 4 trains per hour in off-peak hours and 6 during rush hour. Still, the system attracts about 13,700 riders per weekday and while many have criticized its short length and lack of development near many stations, it’s per-mile ridership is reasonably good compared to US other systems. The low level of frequency requires just 6 trainsets for service, at peak hours 18 railcars out of 27 in its fleet. Typically, three cars are undergoing maintenance and one serves as a spare so NFTA could run up to 23 cars but rarely does except for special events. By running less service, Metro Rail is in effect not using its most expensive assets to its full potential.

Adding more service would attract more riders because waiting times would be significantly reduced and the transfer penalty or time cost of transferring from a bus and waiting for the next Metro Rail train would also decline, both of which make the service more attractive for riders. Instead of waiting an average of 5 minutes at rush hour and 7.5 minutes midday, more frequent service would reduce that time to an average of 3 minutes.

I propose running 10 trains per hour throughout the entire day, using 5 three-car train sets and 4 two-car trainsets which retains the current spare and maintenance ratio of 4 to 27. Increased use of the cars will likely increase wear and tear costs but the cars were recently rehabbed and are getting close to the end of their useful lives, requiring replacement in the next decade or so. See sample schedule.

While Metro Rail would not need to immediately buy new trainsets to run additional service, the key cost increase would be labor. Working with the train driver's union to increase the number of segments each driver operates during their shift could lower the per hour operating cost while still requiring more drivers than today. In addition, driver productivity is increased because train drivers will have shorter times to turn around a train, in some cases as lower as 3 minutes compared to 15-20 minutes today. Funding this dramatic service increase will certainly increase operating costs but this will be offset by increased fare revenue from additional passengers.

With planning for an expansion to the suburbs underway, it will be helpful for authorities to demonstrate increased ridership through better use of existing assets. Why not run more Metro Rail trains?

Bring back the Clocker

While more people take the train (13%) than fly in the Northeast Corridor (2%), far more people drive (81%) instead. Bringing back the Clocker could boost rail’s modeshare.  

Background

Amtrak has two intercity services on the Northeast Corridor. The high-speed Acela Express is targeted at business and first-class passengers and is Amtrak’s most profitable train. The slower Northeast Regional targets economy passengers and carries more passengers than any other service. However, Amtrak has a profit mandate and as a result, fares are high for both services. Furthermore, the company makes more money carrying passengers to more distant points than New York to Philadelphia or Washington to Philadelphia.  

Amtrak uses locomotive-hauled trainsets with single-level cars for the Northeast Regional which typically hod 422 passengers (7 cars) to 638 passengers (10 cars). Due to high demand, the company is short on railcars which limits trainset length. Similarly, the Acela trainsets seat up to 304 people and are often full. Limited capacity is another reason for high fares.

Source Amtrak's Clocker in 1975

Conceived by the Pennsylvania Railroad, the original Clocker used to run between New York City and Philadelphia on the hour. Amtrak took over the service in 1971 but stopped the trains in 2005. Some of the service from New York City to Trenton has been replicated by New Jersey Transit. However, there is no dedicated service for the largest intercity rail market in the country and as a result, rail has just 18% of the 14.6 million passenger market, far less than auto's 80%.

Source

Solution

Why not boost modeshare by bringing back the Clocker with far more frequency, high capacity trainsets, and to more stations. Done right, the Clocker could double rail’s modeshare to 33% and attract an additional 12 million passengers. 

Amtrak runs 2-4 trains per hour between New York City and Philadelphia, about 51 daily roundtrips, with about two-thirds continuing to Washington while the other trains go to Harrisburg or Pittsburgh. New Jersey Transit runs 47 daily roundtrips between New York City and Trenton while SEPTA runs 30 daily roundtrips between Trenton and Philadelphia. Initially, the Clocker should run 2 trains per hour and eventually build to up 4 per hour which is light rail frequency. While Amtrak trains make 2-4 stops between New York City and Philadelphia, the Clocker will make more stops. A proposed schedule is forthcoming and works around existing constraints such as Penn Station at rush hour. 

Amtrak’s current services use Philadelphia's 30^th Street Station which requires a transfer to another train, trolley, or subway to get to the City Centre. The Clocker would use the City Centre Commuter Tunnel serve downtown Philadelphia without requiring a transfer. To return to the Northeast Corridor, the trains could initially reverse at Wayne Junction. Later, a six mile track from the West Trenton Line to New Jersey Transit’s Morrisville yard could be electrified to allow trains to loop back, saving time and providing direct connections to New York City from Philadelphia’s northern suburbs. 

While the Clocker could just serve New York City's Penn Station, extending it to New Haven and Jamaica would reach far more potential riders, especially where current services do not. Jamaica has connections to almost all Long Island Railroad trains, three subway lines, and JFK Airport. The New Haven Line is the busiest commuter line in the country but has no through service south of New York except for about 21 Amtrak roundtrips. The Clocker would almost double that to 36 roundtrips. 

The Clocker would use a bilevel, electric multiple unit trainset like NJT just ordered from Bombardier. While Amtrak’s trainsets are locomotive-hauled, the electric multiple units' better acceleration would help the Clocker make more stops without losing significant travel time. While the NJT coaches carry as many 142 passengers, Clocker coaches would seat up to 121 passengers (60 on lower level, 59 on upper level, and 2 on midlevel plus 1 accessible seat). The midlevels would have large luggage racks and two bathrooms. While Amtrak trainsets have a dedicated food service car, Clocker’s short trips and fast turnarounds will not likely warrant food service. Similarly, business class would not be offered as demand would likely be low for a short trip. A six-car Clocker trainset would carry 736 passengers, more than double a Northeast Regional trainset, and have cabs at both ends, allowing for quick turns in congested stations. About 25 trainsets of 6 cars each would be sufficient for a four train per hour schedule (175 railcars). 

One of the reasons for rail's low modeshare is high fares. Combining a high capacity trainset without offering expensive services such as food and business class should result in low operating costs as just two staff would be required, a conductor and an engineer. Seats could be reserved ahead of time or be bought onboard for a surcharge. Commuter fares could be offered too. While some trains may be profitable, the goal of the service is to shift people out of cars so a subsidy may be required. 

The largest markets would see immediate increases in rail's modeshare. Even some of the smaller markets where the Clocker would serve indirectly such as Connecticut to Long Island would still provide new, one transfer connections at Penn Station instead of a multi-station, multi-subway transfer today. Instead proposing billions in infrastructure upgrades for the Northeast Corridor, why not try running more service? Bringing back the Clocker would dramatically boost rail's modeshare and better connect our country's largest travel markets. 

Metro Express

Merging our two commuter railways, Virginia Railway Express (VRE) and Maryland Area Regional Commuter Train (MARC), would create a region-wide Metro system, Metro Express. Similar Metro, the system would transform our transportation options, reduce pressure on housing, and connect people across our vast region far faster than driving.

Background

Parisian commuter trains used to stop at one of 12 stations and turn around, requiring commuters to transfer to the crowded Paris Metro. Over the past 50 years, the city has carefully connected its commuter rail lines to allow them to run through the city center to the other side, creating the Réseau Express Régional (RER). The system dramatically cut travel times for long trips across the region, in most cases beating travel times by driving and as a result, carries more people than our entire Washington Metro.

Similar to Paris in the 1970’s, both of our commuter rail systems terminate at Union Station. Their primary ridership market is people who work in downtown Washington, DC. Each system is funded by their respective states and while passengers may transfer from one train to another free of charge for some trains, each railroad operates largely independent of the other.

One key constraint on both systems neither own the track upon which they operate trains. Both MARC and VRE have operating agreements with CSX and Norfolk Southern freight railroads. MARC also has an agreement with Amtrak to use the Penn Line. Any expansion requires cooperation of all of those entities.

Final Metro Express Network

While there many plans for expanding commuter rail service in each state and DC, none specify the level of passenger service which would result from a combined system.

I merged VRE's Manassas Line with MARC's Camden Line to create the Cyan Line as both of their current schedules align more than the other lines and allows trains to run straight through Washington, DC. Similarly, I merged the VRE's Fredericksburg Line with MARC's Brunswick Line to create the Black Line as they are geographically in a relatively straight line. Both services would use diesel locomotives and run express past most of the inner suburb stations which would reduce travel times for people living in far away stations.

I extended MARC's Penn Line to create the Pink Line which would terminate in Alexandria instead of Union Station which puts more jobs within a one-seat ride. Pink Line trains would skip every other stop to reduce travel times. The Brown Line would connect Germantown with Broad Run and Woodbridge. Electrification of the passenger tracks would extend from Union Station to Alexandria and beyond, allowing both trains would use electric multiple units (EMUs). EMU's have faster acceleration times than diesel trains which allow stops at all inner suburb stations which the diesel trains would skip.

Proposed frequency of trains
Trains per hour	Line name	Origin	Destination	Trainset type
1	Cyan Line	Broad Run	Baltimore Camden	Diesel
1	Cyan Line	Haymarket	Baltimore Camden	Diesel
1	Black Line	Spotsylvania	Frederick	Diesel
1	Black Line	Spotsylvania	Martinsburg	Diesel
2	Pink Line	Alexandria	Elkton	EMU
2	Pink Line	Alexandria	Baltimore Penn	EMU
2	Pink Line	Alexandria	Baltimore Camden (Via Penn Line)	EMU
2	Brown Line	Broad Run	Germantown	EMU
2	Brown Line	Woodbridge	Germantown	EMU
2	Clocker	Alexandria	Philadelphia/Wayne Jct	EMU
2	Amtrak south of DC
18	trains per hour

Projects

Creating Metro Express would require building new infrastructure projects such as additional track, a new Long Bridge, and electrifying existing lines. Unlike Paris RER, Metro Express would not require lots of expensive tunnels and would use existing track right of ways. I detailed the projects in this Google Map.  

Long Bridge

In order to run substantially more service, a new bridge is required and would be used exclusively for passenger trains. While current plans call for a new span independent of the current bridge, mine uses some of the existing spans over Ohio Dr and Washington Channel to reduce construction impacts and save money. This bridge would also include a new bike path across the Potomac but unlike current plans, mine extends the path across East Potomac Island directly to L'Enfant Plaza, using new spans which would have to be built anyways. Electrification should be included in order to eliminate disruption later on. 

Station rebuilds

VRE was designed cheaply to carry commuters into Washington and its platforms in Crystal City and L'Enfant Plaza only serve one train at a time. To run trains in both directions at the same time, new platforms are necessary. VRE has plans to rebuild both stations. Unlike VRE's plan, my L'Enfant Plaza station does not require new train bridge spans across 6th and 7th streets to save money. Instead, I use the existing 5 track bridge over 6th St and move the current platform off of the 7th St bridge to save money. The only new bridge span would be one for a platform which could be built after the project is complete (initially trains would not open doors for the car over that span). Electrification or provisions for easier electrification should be included in both projects. Notably, these new platforms would be four feet above the rail (high platforms) instead VRE's low platforms which would allow level boarding. This allows people to get on and off much faster and makes loading people with mobility impairments much easier. Alexandria's low platforms would also be raised as would most of Union Station's lower platforms. 

Track work

Adding more tracks is necessary to expand service. Some plans are already in the work as a 4th track from Long Bridge to Alexandria, funded by the Atlantic Gateway project, while other track work is planned but not funded. Key is negotiating with CSX to allow for additional service while maintaining freight service.

For Black Line service (ex-Fredericksburg Line), one section of 3rd track was recently completed with stimulus funding and a 3rd track from Springfield to Lorton is funded by the Atlantic Gateway project. Connecting those tracks would require new bridges over the Occoquan River, Neabsco Creek and Powells Creek. Extending the 3rd track through Fredericksburg would require three more bridges over Aquia Creek, Potomac Creek, and the Rappahannock River. Since the track is owned by CSX, negotiations over these improvements should include at least 2 commuter trains and 1-2 intercity trains per hour as the third track would substantially boost capacity. Brown Line  service to Woodbridge with electrification may require some passing tracks. Brown Line  service in Maryland would also require an electrified 3rd track. Running the Black Line through to Frederick and Brunswick would likely require a 3rd track also and double track along the entire Frederick branch.

For Cyan Line service, Norfolk Southern only uses the line to serve some local industries and rarely runs freight trains through Washington, DC. Metro Express should offer to buy the Manassas Line in exchange for allowing offpeak freight service. Owning the track would speed up improvements, allow for easy electrification, and ensure passenger trains run on time. In addition, CSX would likely require improvements for their Camden Line such as a 3rd track or passing sidings to allow more passenger trains. 

For Pink Line service, Amtrak is better at allowing commuter trains but additional improvements will likely be necessary. For example, a 4th track is planned for BWI Station and to allow frequent service, a fourth track along the entire line from Baltimore to Washington, DC is likely necessary. I propose constructing a spur line from the Penn Line into Camden Station to connect downtown Baltimore directly to Washington, DC.

Great Circle Tunnel

The Baltimore & Potomac Tunnel was completed in 1873 and is in need of total rehabilitation. The two track tunnel has a 30mph speed limit and replacing it with a new tunnel would save two minutes for all DC bound trains. The stimulus funded preliminary engineering for a new tunnel, called the Great Circle Tunnel, which would create up to four new passenger tunnels. I propose building two track tunnels first. That way, the old B&P Tunnel could be rehabilitated as a one-track tunnel or traded to CSX for its Howard St Tunnel which would then be rebuilt for Baltimore's light rail service. 

The Washington Metro took 25 years to build. Metro Express could be built in a similar phased fashion. Dramatically reducing travel times would connect more communities to more jobs, particularly Baltimore, which would reduce pressure on housing in the inner suburbs and intercity while boosting demand in the outlying communities. Merging MARC and VRE into Metro Express should be the first step towards creating this transformative system. 

Amtrak in 1971

Whenever someone laments about the state of passenger trains in the US, they often recall that before Amtrak, there were many more passenger trains. Check out where these trains went on the day before Amtrak began service.

Green represents Northeast Corridor trains. Red represents daytime, short-haul trains while blue shows the routes of long-distance, overnight trains.

On April 30, 1971, there were approximately 191 daily passenger trains, operated by railroads which elected to join Amtrak. Five railroads declined to join, Denver & Rio Grande Western, the Georgia Railroad, the Reading Railroad, Rock Island Railroad, and the Southern Railway, but all eventually discontinued their trains by the 1980’s.

The Pennsylvania Railroad had extensive passenger service while the New York Central had significantly trimmed theirs. The Central had about 17 daily trains including just two long-distance ones. Some of their long-distance trains were converted into Chicago-Buffalo or New York-Buffalo day trains which cost significantly less to operate. In contrast, the Pennsy had about 50 daily trains including six long-distance ones, three of which competed with the Central on the New York-Chicago route.

Some corridors had multiple railroads serving the same market. Chicago-Milwaukee, Chicago-Minneapolis, and Chicago-Detroit. It was sensible to combine competing routes into a single line.

There were many more long-distance overnight trains than operate today. About 44 daily round trips operated overnight and some markets had multiple trains to choose from. New York-Chicago travelers had five trains, Chicago-Los Angeles passengers had three, and there were four New York-Miami trains. About 14 daily long-distance trains remain with no route receiving multiple trains.

The Northeast Corridor was a bit of a mess as it had many short-haul trains such as New Haven to Hartford or New York to Philadelphia. Combined with ancient ex-PR GG1 electric locomotives and unreliable Metroliners, plus decades of deferred maintenance, years would pass before operational profitability was achieved. At the time, planners thought that Northeast Corridor profits would subsidize the rest of the system but significant operational “profits “only materialized in the mid-2000’s and even then, some question whether Amtrak’s accounting system overstates Northeast Corridor “profits” at the expense of long-distance trains.

By 1971, most passenger trains had already disappeared, especially the small towns. However, most large cities retained passenger service. The largest city to lack service was Dallas although Fort Worth had service nearby. Of the 48 continental states, Maine, New Hampshire, Vermont, and South Dakota had no service at all. Today, South Dakota still lacks service and Wyoming last saw service in the early 1990’s. Looking at major cities, Dallas and Austin now have train service but many more large cities have no service today.

Ten largest cities without passenger trains in 1971

• Dallas, TX 844,401
• Austin, TX 253,539
• Des Moines, IA 201,404 in 1970 Rock Island canceled service
• Shreveport, LA 182,064
• Knoxville, TN 174,587
• Huntsville 139,282
• Springfield, MO 120,096
• Columbia, SC 112,542
• Allentown, PA 109,871
• Manchester, NH, 87,754

Ten largest cities without passenger trains in 2019

• Phoenix, AZ 1,626,078
• Columbus, OH 879,170
• Nashville, TN 667,560
• Las Vegas, NV 648,224
• Louisville, KY 621,349
• Colorado Springs, CO 465,101
• Tulsa, OK 403,505
• Wichita, KS 390,591
•  Lexington, KY 321,959
• Madison, WS 255,214

Some service was added shortly after Amtrak began. Vermont subsidized the Montrealer from Washington to Montreal. New York State paid to resume its own Montreal service. Massachusetts paid to extend two trains from Washington-Springfield to Boston, restoring the Inland route. Illinois reactivated train service to Quincy almost immediately. Only in 2001 did Maine begin subsiding the Downeaster train, restoring passenger service which had been discontinued in 1965.

While many famous trains had disappeared long before Amtrak, about half disappeared with the creation of Amtrak. While a few routes have more trains than they did then, service along most of these routes is likely gone for good. 

Replacing Amtrak long-distance trains with shorter ones?

Amtrak is considering breaking up some long-distance trains into shorter corridor ones as part of its Congressional reauthorization. While in some cases it may make sense, the financial advantages may not pan out.

Amtrak has proposed to replace an unspecified number of overnight, long-distance trains with daytime corridor service, presumably sponsored by states. Of Amtrak’s 15 long-distance routes, the Palmetto is already a daytime train. Other trains are likely too long or lack substantially sized cities along their routes to make this transition worthwhile, think the Auto Train, California Zephyr, Empire Builder, Southwest Chief, and Sunset Limited. But the remaining nine trains could theoretically be replaced by daytime corridor service.

Background 

Amtrak’s key rationale for cutting long-distance trains is cost savings, primarily labor costs. Day trains only require some café service, coach seating and sometimes business class, with an engineer, conductor, and a café attendant as labor. Overnight trains require a dining car, coach seating and sleeper cars, with at least two engineers, two conductors, several dining car crew, and sleeping car attendants. This would also reduce the number of expensive sleeper and dining cars it needs to replace its aging long-distance fleet of Superliner railcars.

Amtrak’s proposal has been tried before. In the late 1960’s, the New York Central Railroad replaced several of its overnight services with day trains to save money. One of Amtrak’s largest costs is labor. Its New York City-Buffalo-Detroit-Chicago train was split into two day trains, New York City-Buffalo and Chicago-Detroit-Buffalo. One of its New York City-Cleveland-Chicago trains became two trains, New York City-Buffalo and Chicago-Cleveland-Buffalo. One of the trains referenced by the WSJ, the City of New Orleans, was originally a fast day train and would leave Chicago at 8 AM and arrive in New Orleans at 11:55 PM, taking just 15h55m and averaging 58 mph despite 22 stops. Two of those stops requiring connecting other trains from St. Louis and Louisville in ten minutes, a move that takes an hour on Amtrak’s Lakeshore Limited. Amtrak’s Palmetto train runs on a similar schedule, leaving New York City at 6:02 AM and arriving in Savannah at 9:04 PM. It leaves Savannah at 8:20 AM and arrives at 11:56 PM back in New York City.

Other countries have also seen their overnight service disappear, particularly in Europe with the proliferation of low-cost airlines and expanded high-speed train service. After Germany’s Deutsche Bahn dropped all overnight service, half of the service was taken over by Austrian Railways. After high-speed service began between Madrid and Barcelona, overnight service from Madrid to Paris was dropped even though high-speed trains take a little over 9h and require a transfer. In contrast, the UK is investing its two overnight services with its London-Scotland service receiving new cars.

Short-haul trains in the US benefit from subsidies from states. In 2008, the Passenger Rail Improvement and Investment Act (PRIIA) required all states to subsidize trains running less than 750 miles within their borders. When Amtrak began service in 1971, it inherited some state services and subsidized those itself. But over time, states began subsidizing additional service and by 2008, the situation had become unequal with some states receiving “free” trains from Amtrak while others had to pay the entire amount. PRIIA equalized that situation and states such as New York found themselves subsidizing train service it had previously received from “free”. Long distance trains were explicitly omitted from those rules and would continue to be subsidized by Congress. But Amtrak’s reauthorization provides an opportunity to rework those rules and could allow Amtrak to obtain additional subsidies from states.

Amtrak’s proposal would also allow better departure and arrival times for key markets along corridors. For example, the Lakeshore Limited’s current schedule has poor timings for the Chicago-Cleveland market with arrivals in the middle of the night or early morning. It also requires an hour stop in Albany to split the train for Boston and New York City sections. Retiming the train would allow for more reasonably arrival times in Cleveland and provide another day train on the New York City-Niagara Falls corridor. In theory, this could attract more riders than would be lost from overnight services.

Old schedule

	New York City	Albany	Depew	Cleveland	Chicago
Departure/Arrival	3:40 PM	6:10 PM	12:12 AM	3:33 AM	9:50 AM
	Chicago	Cleveland	Depew	Albany	New York City
Departure/Arrival	9:30 PM	5:38 AM	8:46 AM	2:31 PM	6:23 PM

New schedule

	New York City	Albany	Niagara Falls	Niagara Falls	Cleveland	Chicago
Departure/Arrival	9 AM	11:30 AM	6:50PM	9 AM	10:50 AM	6:20 PM
	Chicago	Cleveland	Niagara Falls	Niagara Falls	Albany	New York City
Departure/Arrival	9 AM	5:30 PM	8:20PM	9 AM	4:20 PM	5:50 PM

Amtrak’s proposal will have to address fundamental challenges of US passenger service 

First, all long-distance trains run on tracks controlled by freight railroads. Reconfiguring service would require their cooperation as schedules would move from overnight to the daytime.

The good news is that freight railroads are moving from more frequent unscheduled service to less frequent scheduled service. In theory, this creates more track capacity and would allow Amtrak to better schedule its trains to avoid delays due to freight trains. In addition, the number of coal trains has declined, typically the slowest and most capacity intense ones, and fewer oil trains are running due to the decline in crude oil prices.

Even if a freight railroad agrees to allow more passenger trains, running more than one train per day would likely require some investment, particularly if Amtrak wants its new service travel times to be competitive with driving. Presumably, its reauthorization to Congress would include specific projects and cost estimates to be funded by Federal grants, loans, or state grants.

Second, while operating day trains cost less, ticket revenue is also substantially lower. While the Empire Service to Niagara Falls averages $61.76 a ticket, the Lakeshore Limited averages $77.63. Amtrak’s Silver Palm used to run between New York City and Miami but was truncated to Savannah in 2004. While the Silver Meteor averages $109.38 a ticket on the Miami run, the daytime Palmetto receives just $75 a ticket. However, the Palmetto costs $32.7 million and loses just $3.3 million but the Silver Meteor costs $72 million, more than double the Palmetto, and loses $34.7 million, ten times as much.

Third, labor costs should be better addressed on existing routes prior to any major change. Railroad unions remain strong but there needs to be a frank discussion of fewer staff members on long-distance trains in exchange for more trains overall for staff to operate. While sleeping car attendants in every car made sense at some point in the past, some functions should be automated such as dinner reservations and setting up beds. Amtrak has tried replacing dining car service with cheaper options, most recently last year, and ended up returning some hot food items. Whether this option balances between cost and customer satisfaction remains to be seen.

Fourth, Amtrak’s financial accounting has long been described in many circles problematic or even fatally flawed. For example, the Rail Passenger Association claims that the Amtrak Performance Tracking system allows the company to understate its Northeast Corridor costs and overstate its long-distance train costs. Congress has required Amtrak to substantially improve its accounting system but a 2013 Amtrak OIG report and 2016 GAO report both found the company was overly reliant on cost allocation and still lacked accurate cost information. Amtrak has made smaller changes without accurate cost information before. While the company made substantial changes to its private railcar program in 2018, it did not know the true cost of providing that service before making those changes.

Fifth, PRIIA required Amtrak to write a series of performance improvement plans for its long-distance trains. While staff came up with a lot of great ideas, relatively few have been implemented. Prior to a major change, implementing more of those improvements may attract more passengers and more revenue. 

While I understand Amtrak’s rationale, it is hard to say whether two corridor trains will attract more passengers and cost less overall than a single overnight train. I think it makes more sense to provide additional service along current long-distance routes first. Amtrak has the key ingredients for developing realistic service expansion plans. In May 2007, Amtrak made a successful presentation for additional service to Lynchburg along an existing long-distance route. This service was a runaway success and paved the way for substantial state investment in passenger trains. There are other good candidates such as Chicago-St. Paul and Chicago-Cincinnati which would generate strong ridership with some travel time reductions. While tempting to gamble a lot for a reauthorization opportunity, a cautious approach may pay off instead.

Nine long-distance routes possibly targeted for replacement corridor service
Capital Limited: Chicago-Cleveland-Pittsburgh-Washington>
Chicago-Cleveland-Pittsburgh
Washington-Pittsburgh

Cardinal: Chicago-Indianapolis-Cincinnati-Charleston (WV)-Washington-New York City>
Chicago-Indianapolis-Cincinnati
Washington-Charleston (WV)-Cincinnati

City of New Orleans: Chicago-Carbondale-Memphis-Jackson-New Orleans>
Chicago-Carbondale-Memphis
Memphis-Jackson-New Orleans

Coast Starlight: Los Angeles-San Jose-Sacramento-Portland-Seattle>
Los Angeles-San Jose-San Francisco
San Jose-Sacramento-Portland-Seattle

Crescent: New York City-Washington-Charlotte-Atlanta-New Orleans>

New York City-Lynchburg-Charlotte
Charlotte-Atlanta
Atlanta-New Orleans

Lakeshore Limited: New York City-Albany-Buffalo-Cleveland-Chicago>
New York City-Albany-Buffalo-Niagara Falls
Chicago-Cleveland-Buffalo-Niagara Falls
Boston-Albany

Silver Meteor: New York City-Washington-Charleston-Savannah-Orlando-Miami>
New York City-Washington-Charleston-Savannah
Miami-Orlando-Savannah

Silver Star: New York City-Washington-Columbia-Savannah-Orlando-Miami>
New York City-Washington-Columbia-Savannah
Tampa-Orlando-Savannah

Texas Eagle: Chicago-St. Louis-Little Rock-Dallas-Fort Worth-San Antonio>
Chicago-St. Louis-Little Rock-Dallas-Fort Worth
Fort Worth-Dallas-Austin-San Antonio