Tag Archives: MARC

Capacity on the Northeast Corridor

Tucked into the testimony of Amtrak President Joesph Boardman at last week’s Senate hearing on the future of the Northeast Corridor is this graphic demonstrating the number of daily train movements by operator at different locations along the spine of the Northeast Corridor:

One interesting thing to note is the difference in the volume of Amtrak trains (light blue) north of New York, compared to south of New York. This also makes it easy to see the relative volume of Amtrak intercity trains and commuter trains, as well as a few freight movements per day north of Washington Union Station. Capacity improvements are needed to allow for a combination of increased intercity and commuter services (or even better), and other bottlenecks are likely in need of greater capacity for freight expansion on adjacent corridors.

In other Amtrak news, Systemic Failure takes note of the US rail regulatory apparatus continuing to shoot itself in the foot on even allowing efficient high speed rail and learning from everyone else around the world that has already done this hard work. The FRA rejected Amtrak’s reasoning below, with emphasis added by Drunk Engineer:

The assumption that the standards simplify the design process of the equipment and would save $2,000,000 per train set is false. The Acela example indicates the exact opposite to be true. The FRA rules, as existing and proposed, eliminate the possibility of purchasing off-the-shelf equipment. The engineering work required to design new compliant equipment alone would far outstrip any possible savings from the rules if there were any to be had.

For background on the previous history of the Acela’s regulatory weight problems, see posts here, herehere, as well as a GAO report here.

Towards a DC S-Bahn, part 2

VRE train at Franconia-Springfield. CC image from nevermindtheend

DC’s existing (yet fragmented) commuter rail network is a huge low-hanging fruit for expanded and improved transit service (see this previous post). Writing at Pedestrian Observations, Alon Levy makes the statement that nobody likes riding North American commuter rail.  Alon compares two locations in New York that have both subway and commuter rail service – and in each case, the subway ride wins a much larger share of riders despite often faster rides on commuter rail.

Though the data isn’t easily available for the commuter rail operators, the differences in ridership are substantial.  The Rockville Metro station alone has more boardings than the entire MARC Brunswick line.

Alon identifies four reasons – a poorly structured network that does not serve non-downtown destinations; poorly designed transfers, often with financial penalty; cost differential and a lack of an integrated transit fare structure across all modes; and low frequency service.

Addressing the DC region specifically, some of these are undoubtedly true.  A lack of through-routing prevents serving non-downtown destinations on the other side of Union Station MARC could easily serve dense employment clusters in Crystal City and Alexandria, VRE could offer service through to Rockville, Silver Spring, Fort Meade and others. Likewise, train frequency isn’t good – structuring it more like urban rapid transit could be a huge improvement.

Alon’s four points open the door for a comparison between Metro and the area’s commuter rail services. The commuter rail network shares several stops with Metro. Shared stops are as follows (stations in bold are those along the shared track segment of a conceptual through-routed network):

Maryland:

  • Rockville
  • Silver Spring
  • College Park
  • Greenbelt
  • New Carrollton

DC:

  • Union Station
  • L’Enfant Plaza

Virginia:

  • Crystal City
  • King Street
  • Franconia-Springfield

While transfers aren’t particularly easy, some of the physical connections aren’t terrible.  Some stations share the same basic platform access (New Carrollton), while others easily could do so (King Street) with a little construction.  The Crystal City connection is a bit of a stretch – it involves several blocks of walking, either along Crystal City’s streets or through the warren of tunnels and underground retail space.  L’Enfant Plaza does not have an actual connection to the Metro platforms, just adjacency.

Financial transfer penalties are another story.  MARC offers an add-on TLC pass (at the cost of $102/month on top of the cost of a monthly MARC pass) that allows for unlimited use of local transit (rail and bus) in both DC and Baltimore; VRE offers a similar product with a similarly-large surcharge per month.

Using MARC’s $102 per month figure, and assuming 40 last-mile Metro trips per month, that would require a minimum of a $2.55 peak fare to make the pass break even on commute trips alone (roughly the equivalent of a ~15 minute Red Line ride from Union Station to Van Ness).

The fare structures aren’t entirely integrated either, though the disparities aren’t as large as in Alon’s example from New York.  Thanks to Metro’s time-and-distance based fare structure, you don’t find the same disparity of a flat-fare subway system up against a graduated fare commuter rail system.  The example of Far Rockaway shows the disparity – a subway ride to Midtown is a flat $2.25, while the LIRR to Penn Station is $10.00 at the peak, $7.25 off peak.

Compare that to the single ride fares for MARC/VRE and Metro – all fares to Union Station as a point of comparison.

Maryland (Penn Line fares, Camden and Brunswick line fares; Metro fares from Union Station):

  • Station – MARC fare to Union – Metro fare to Union
  • Rockville – 5.00 – 5.75 (time: 35-42 mins via MARC; 34 mins via Metro)
  • Silver Spring – 4.00 – 3.35
  • College Park – 4.00 – 3.65
  • Greenbelt – 4.00 – 4.30
  • New Carrollton – 4.00 – 4.20

Virginia (VRE fares – single ride price used; Metro fares from Union Station)

  • Station – VRE fare to Union – Metro fare to Union
  • Crystal City – 6.20 – 2.60
  • King Street – 6.20 – 3.70
  • Franconia-Springfield – 6.80 – 5.60 (time: 36-41 mins via VRE; 45 mins via Metro)

MARC fares are all rather close to Metro; VRE fares have a different problem of the LIRR-Subway comparison at Far Rockaway; the longest  possible competing trip (from Franconia-Springfield) has the smallest fare differential, it’s the shorter trips that are out of whack (a one-station VRE ride from L’Enfant to Union Station costs $5.55 on VRE, compared to the minimum Metro rail fare of $2.10).  This structure obviously reflect’s VRE’s role as an AM-peak-inbound, PM-peak-outbound operation, but certainly discourages usage within the core of the region for rapid transit.

Commuter rail isn’t always more expensive, either.  Looking at Rockville, (which, again, draws more boardings than the entire Brunswick line) a monthly pass to Union Station costs you $125 with the various discounts, while 40x trips per month via Metro at $5.75 a pop totals $230 ( !!! ); a weekly MARC pass totals $37.50 compared to $57.50 for ten peak-hour rides on Metro.  A person who was dropping $230 a month on Metro fares could easily purchase a $125/month MARC pass, add on the $102/month TLC pass and still get unlimited Metro usage off-peak for about the same cost.

A unified fare structure would likely involve lowering fares within the inner VRE territory (further integration would assume a single fare table for a through-running merged S-Bahn-like operation between MARC and VRE) to better mirror the various Metro fares for similar distances.  I would imagine this to be an easier organizational lift than, say, trying to bridge the gap in peak fares at Far Rockaway between $2.25 and $10.00.

The biggest difference in usage (thereby indicating usefulness) would appear to be frequency.  MARC’s Penn Line features the most frequent service, and even that is 20-40 minutes between trains at best during the peak, hourly trains at mid-day, and longer headways in the evening – plus, no weekend service.  Frequency is freedom, after all.  Thus, the purpose of any effort for through-running commuter rail services should be to help the transition of DC’s commuter rail network into a frequent S-Bahn-like network of interlined rapid transit services.

Towards a DC S-Bahn

S-Bahn logo. From wiki.

This week, Greater Greater Washington has run a series of posts on the hurdles to implementing through-routed commuter rail services in DC. The technical reasons include many basic incompatibilities between the region’s two commuter railroads (MARC and VRE), ranging from type of locomotion to platform height, as well as the infrastructural shortcomings of DC’s rail infrastructure to handle high frequency transit-like operations.

Lost in the wash, however, is the reason to do this.  The reasons are two-fold: First, through-routed service expands the transit network relatively inexpensively, offering mobility benefits to current and future riders.  Second, such a service (and the technical changes required to implement it) help solve some of the other challenges of DC’s commuter rail network (such as insufficient storage capacity at Union Station for trains to layover mid-day). Through-running is both a means to an end as well as an end itself.

This isn’t exactly a new concept, as it has been raised in numerous places over the years:

Plus the southern intentions of intercity rail:

CSX’s plans for increasing freight traffic will also mean added capacity through DC, with implications for passenger rail.

None of these plans hints at even the possibility of the kind of S-Bahn integration potential that through-routing unlocks.

The goal should be to turn the core of DC’s commuter rail network into a system like Germany’s S-Bahns (touched on previously here and here).  David Alpert essentially suggested the same thing with his conception of the ‘Metro Express.’  The level of service would be more like rapid transit than commuter rail.  The geographic extent might not be as expansive as the current commuter rail network (there’s no sense in running rapid transit to Martinsburg, WV) but those outer extensions could easily be serviced by a service that mirrors today’s commuter rail.  The core of the network (say, to Woodbridge in the south, Manassas in the west, Germantown or even Frederick to the north, and Baltimore to the east) would see higher frequencies, through-routed service, and all-day, full week service.

One common characteristic of S-Bahns is the use of interlining and shared tracks in the core of the system (this Wiki diagram illustrates), where interlining produces short headways on the shared portion while the outer parts receive less frequent service due to the branching of the network. The three MARC lines feeding DC, each running on 30 minute headways would combine for 6 tph in the shared segment (Union Station to Alexandria).

VRE’s timetable shows Alexandria-L’Enfant at about 17 minuntes; Alexandria-Union Station at about 25 minutes.   There’s room for massive improvement here – Metro’s trip planner shows King Street to Union Station (Yellow to Red) as 20 minutes on Yellow, 4 minutes on Red; King Street to L’Enfant Plaza is the same 17 minutes on the Yellow line.

Give a DC S-Bahn transit-like service and that can be reduced.  Electric multiple unit trains would accelerate faster; level boarding and more frequent service would shorten station dwells; etc. Then, the commuter rail infrastructure would function as a key part of the region’s rapid transit network.