How to Build New York's Interborough Express

New York City is building a new transit line. No, not the second phase of the Second Avenue Subway, or the Queenslink proposal on the old Rockaway Beach Branch, or the subway extension to LaGuardia Airport that is now back on the table, or the recently announced desire to put a subway station in Red Hook, or the now over 100-year-old Utica Avenue subway proposal. Though all of those projects are in some stage of evaluation and may or may not be built, we are talking about the Interborough Express, which was announced by Governor Hochul at her 2022 State of the State address. The Interborough Express is a proposed rapid transit route that will link Brooklyn and Queens along the Bay Ridge Branch and Fremont Secondary, two under-utilized freight rail lines that travel in a continuous, grade-separated path from Sunset Park in Brooklyn to Woodside in Queens. If completed, it would be only the third rapid transit route in New York City to not touch Manhattan alongside the G subway line and Staten Island Railway.

Interborough Express, shown in light blue, alongside existing rail lines in New York City, courtesy of the MTA

The Proposal

The MTA began a feasibility study for the line in 2020 and released its report in January 2022. The study narrows the project down to three potential alternatives, each with projected ridership and overall ride time: conventional heavy rail, light rail, or bus rapid transit. Conventional heavy rail can be either subway-style rail with its own dedicated tracks, as used by the rest of the New York City Subway, or regional-style rail that can share tracks with freight trains, as used by the Long Island Rail Road, or LIRR, and Metro-North systems. The study's conventional heavy rail alternative proposes a hybrid of these two options, with the same dedicated tracks needed for a subway-style service but with compatibility instead with regional-style trains like those used by the LIRR. The study also does not specify any particular number of stations or station locations at all, with the public encouraged to suggest locations on an interactive map published by the MTA. The only station locations suggested are the proposed line's intersections with existing transit lines, as the study mentions the line will connect to 17 different subway stations and the Long Island Rail Road.

Comparison of the final Interborough Express alternatives, courtesy of the MTA

This is where issues with the study start to appear. The study projects daily ridership and trip time numbers, but it is unclear what stations are used to measure these numbers. Given the only semi-confirmed locations are at intersections with existing subway or LIRR service, it is unknown if the study assumes massive unserved gaps will be left along the route, such as the more than 3.25 miles between Flatbush Ave/Brooklyn College station on the 2 and 5 and New Lots Ave station on the L. Even assuming each alternative has the same station locations, it is unclear how certain conclusions are reached. The heavy rail alternative has the highest overall trip time, but it is the only option that will be in an exclusive right-of-way for its entirety, meaning it will never have to deal with vehicle traffic that can drastically slow down transit. Additionally, the heavy rail alternative projects needing fewer train consists than the light rail alternative despite having the same headways and a higher projected trip time. Given the study implies both fleets will operate exclusively on Interborough Express, this is the inverse of what should occur, as the number of required consists should decrease as trip time decreases and headways remain the same.

Beyond these issues, the selected alternatives are fairly questionable. The conventional heavy rail option uses regional-style trains but proposes fitting them with additional doors and subway-like interiors to have them closer resemble subway cars. So why not just make them subway cars to match the rest of the intracity network? Two reasons are given for this. First, this will allow for LIRR trains to use the line, which could be useful if not for the fact that the study admits this will be extremely complex to operate and cost a ton of additional money to build. The purpose of this line is to move passengers between boroughs and not to and from the rest of Long Island like the LIRR does. Second, the study says that a shared freight corridor like this would require trains to be FRA-compliant without proper separation, and regional-style trains are FRA-compliant while subway-style trains are not. This is a true statement, but it is misleading to present it as an insurmountable problem, as it is relatively easy and straightforward to properly separate the rapid transit line from the freight line to be compliant as demonstrated by the L, N, and M trains that already do this on the corridor. Even their own proposed light-rail alternative in the study is properly separated from freight trains, showing they understand it is possible.

The study compares the Interborough Express to the Paris RER with its FRA-compliant trains, but the Paris RER operates more akin to a commuter rail like LIRR or Metro-North than an intracity train line, courtesy of Ile-de-France

Lastly, there is the issue of looking at this line in isolation and not factoring in the surrounding environment. This can be seen by simply looking at the alternatives presented. The MTA started this study from scratch with six alternatives, but they were all for the same exact route with the same theoretical, if perhaps not complete, list of stations. The only difference between alternatives was the mode of transit. Mode of transit is not the most important question here; That should be asking what the goals and objectives of the overall project are. Is the MTA attempting to improve travel times between all outer boroughs? In that case, why is there no service to Staten Island or the Bronx? Or, is the MTA only attempting to improve transit between Brooklyn and Queens? After these questions are asked and answered, looking at existing conditions is a good next step. For example, Metro-North has plans to utilize the corridor north of the proposed northern terminus for service to the Bronx, Westchester, and Connecticut. Would it make more sense to integrate the system with Metro North instead of LIRR? Alternatively, would it instead make sense to interline with the existing subway service on the corridor? Questions like this can then directly lead to solutions that then include selecting the transit mode. Starting at this endpoint and examining which of six transit alternatives is best for a project is completely backward, and ending with no subway alternatives in a city with more than 850 miles of subway track and a bus rapid transit option when the governor has clearly advocated for a rail line makes the entire process a joke that cannot be taken seriously.

A Better Way Forward

The concept of the Interborough Express is one worth exploring and likely building. To start, we can establish an overall goal and objective of the project: to improve connectivity in the outer boroughs of New York City along the under-utilized Bay Ridge Branch and Fremont Secondary corridor in a cost-effective manner while preserving the corridor's freight capacity with room for two dedicated tracks. Next, we can examine existing conditions. Although this corridor has no rail service for its entirety, it actually does share a right-of-way with subway services for three separate segments: the BMT Myrtle Ave Line (operated by the M) for less than a quarter of a mile, the BMT Canarsie Line (operated by the L) for a little less than 2.5 miles, and the BMT Sea Beach Line (operated by the N) for about 1.5 miles. Additionally, the corridor has room for two Interborough Express tracks alongside 2 freight tracks for most of its entirety south of Bowery Bay Junction, where the Fremont Secondary Branch intersects with the Northeast Corridor, with two notable exceptions standing out among a couple of other minor locations. In between Bowery Bay Junction and the IRT Flushing Line, there is physical room for the four total tracks sought after, but constructing them would require building a couple large viaducts parallel to existing structures, widening a tunnel under the intersection of Broadway and 37th Ave, and shifting the existing freight tracks. Separately, the roughly 2/3-mile East New York Tunnel presents a challenge because even though it was built for four total tracks, one of those tubes is currently used by the Buckeye Pipeline to serve aviation fuel to both LaGuardia Airport and JFK International Airport. Even more complicated, the East New York Tunnel bypasses two major potential rail connections at East New York with the LIRR and Broadway Junction with the A, C, J, Z, and L subway lines.

East New York Tunnel, with the closed leftmost portal containing the Buckeye Pipeline, courtesy of LTV Squad

The former obstacle mentioned can be bypassed by simply cutting the Interborough Express short at the IRT Flushing Line and leaving a provision for northern expansion to Astoria and the Bronx in the future, with a future study determining whether the potential costs outweigh the benefits. However, the latter obstacle cannot be avoided given its location in the middle of the Bay Ridge Secondary. As mentioned earlier, the BMT Canarsie Line travels alongside the Bay Ride Branch for a little less than 2.5 miles, and part of that segment happens to cross through this exact obstacle. So, could the Interborough Express share tracks with the L and use the BMT Canarsie Line for this segment and potentially beyond? This introduces the concept of interlining, where different rail services use the same tracks. This is a common occurrence in rapid transit to create a more frequent service through one area, called the line's trunk, with different services then branching off to serve other areas with less demand than the trunk. For example, the 2 and 3 trains both use the same tracks on the IRT Broadway-Seventh Avenue and IRT Eastern Parkway Lines through Manhattan and Brooklyn, but the 2 branches off to the south to serve the IRT Nostrand Avenue Line while the 3 branches off to the east to serve the IRT New Lots Line. The New York City subway actually interlines extensively across its network, with the L being one of only four services that have no interlining.

So, what would an interlined Interborough Express/L service look like? The combined line's trunk would be between New Lots Avenue station and the existing tunnel portal north of Wilson Avenue station, covering the entirety of the BMT Canarsie Line's service that utilizes the Bay Ridge Branch corridor. Before and after the trunk, the L would retain its existing service pattern along the BMT Canarsie Line north to Manhattan and south to Canarsie, while the Interborough Express would continue on new tracks north to the Fremont Secondary's intersection with the IRT Flushing Line and south to the Bay Ridge Branch's terminus at 65th Street Yard. Beyond the benefits of avoiding having to touch the East New York tunnels, approximately 2.5 of the 14 total track miles, or 18%, for the proposed Interborough Express route would already be complete in addition to 7 stations. Additionally, by interlining the service, the Interborough Express would use the same rolling stock and maintenance facilities as existing the existing subway network, allowing for even less new infrastructure. In total, the approximate savings from doing interlining with the L are likely almost $2 billion given recent cost estimates.

Potential interlined service pattern, with the L, shown in grey, the Interborough Express, shown in blue, and their shared section, shown in black, courtesy of Google Maps

Preserve the L

This proposal sounds great, but what about L line capacity? The BMT Canarsie Line currently has a maximum capacity of 19 trains per hour (tph), all of which are used by the already crowded L. Luckily, the tracks are already able to handle 26 tph, and the signals can handle more than 30 tph, with only the electrical systems needing to be upgraded to match this capacity. The necessary electrical upgrade can be done with a fraction of the money saved from interlining the Interborough Express with the L, as the project is estimated to cost well less than $1 billion. This alone can allow increased L service at 20 tph and leave 6 tph for the Interborough Express.

While 6 tph is a good start for the Interborough Express, the MTA has stated a desire to achieve 12 tph for any rail alternative, requiring a combined tph of 32 accounting for the 20 L trains. As mentioned prior, the signals can handle more than 30 tph, with other similar similar systems around the world handling up to 36 tph. So, what is currently limiting the tracks from being upgraded to a similar capacity? Three factors: a lack of available trains, a lack of tail tracks at both termini of the BMT Canarsie Line, and the topography of the line in Williamsburg and East Williamsburg. Additional trains can be purchased alongside the Interborough Express construction, as more trains will be needed for a new line anyway, and the latter two issues should not be problematic because the trunk section of the line handling both the Interborough Express and L train is not impacted by those issues. So, with those issues mitigated, the system should be capable of operating 32 tph through the shared segment, with the Interborough Express having 12 tph and the L train having 20 tph.

A total of 32 tph would make the shared section of the BMT Canarsie Line the most heavily used portion of the NYC Subway System, and that could draw pushback from the MTA even though it is feasible. To be fair, there is some validity in the argument that achieving this level of frequency is more practical on paper than in reality, and that is due to dispatching. Dispatching refers to the ability to send trains from each end of the line on a schedule, and more frequent service requires more precise dispatching, as any small delay can easily cascade into a long string of delays without slack in the schedule to reset operations on time. The MTA has sometimes struggled with proper dispatching, and while it could learn to improve based on the operations of other transit agencies around the globe, an alternative approach is to remove the human element from the equation. Much of the issues surrounding irregular dispatching are due to human error, as humans cannot achieve the same levels of consistency or precision as computers. To overcome this, other transit lines around the world that want to achieve frequencies as high as 48 tph have resorted to GoA Level 4. GoA refers to how automated a transit system is, with Level 4 permitting trains to drive, open and close doors, and operate in irregular occurrences without the need for a driver or attendant. Converting the BMT Canarsie Line to GoA Level 4 and building the Interborough Express to this standard from the start will eliminate any concern about frequencies being too high for interlined operations. Line 4 of the Paris Metro was recently converted to GoA Level 4 for a cost of ~$500 million. Factoring in the fact that the BMT Canarsie Line is about 18% longer than Line 4 and that the MTA has never done a GoA Level 4 conversion before, we can generously estimate such a project will cost ~$1 billion, which should still be within the budget given the savings from interlining. However, it will also bring the additional benefits of lower operating expenditures without the need for labor, and each existing train used on the L can be redeployed to other lines across the B Division of the NYC Subway to increase capacity elsewhere.

Paris Metro Line 4 GoA Level 4 Conversion overview, advertising frequencies as high as 42 tph, courtesy of Siemens

The Path from Idea to Reality

Although no timeline was announced for the Interborough Express alongside the report, it can be assumed that it will be one of the MTA's top projects for at least the next four years given Governor Hochul's prioritization of the project and her recent re-election. With this in mind, it is important to move quickly in advocating for a better Interborough Express. Allowing the MTA to take a good idea and plague it with operational inefficiencies will forever doom the Interborough Express as a problem child in need of money and attention by an agency already struggling to maintain century-old infrastructure. The time is now to make sure the MTA builds the Interborough Express properly to ensure its success in the future.

Boston's REM

Boston is a city of transit. The first subway system in the United States opened in Boston under Tremont Street in 1897, and the system has since expanded throughout the city and into a few neighboring municipalities. In 2019, the MBTA, the organization that operates Boston's transit, carried over 152 million riders on its three heavy rail lines, the Orange Line, Red Line, and Blue Line, and over 47 million riders on its two light rail lines, the Green Line and Mattapan Trolley. Collectively, this gives Boston the honor of having the fourth-largest rapid transit system in the country by ridership, yet its network is actually relatively small compared to its peers. The Chicago L, the third-largest rapid transit system, only carried about 10.5% more passengers in 2019 than Boston's combined heavy and light rail lines. Yet, the L has more than 60% more track miles than Boston's rapid transit network. In fact, Boston's network is so dense that its light rail is the most used per mile of any system in the country, and its heavy rail is the most used per mile of any system in the United States outside of New York City.

Boston's Existing Rapid Transit Network

A rapid transit system's size can be small and still be impactful, but Boston's small size results in a major flaw: the system does not cover the entire city, including many dense, heavily populated regions. As shown on the map below, sizeable parts of Charlestown, Allston, Brighton, South Boston, Jamaica Plain, Roxbury, Dorchester, Mattapan, and the Seaport have no rapid transit within walking distance, with West Roxbury, Roslindale, and Hyde Park having no rapid transit access at all.

Map of Boston and its transit lines courtesy of Boston Planning & Development Agency, with yellow circles added to show a half-mile radius around all rapid transit stations, or around a ten-minute walk

An eagle-eyed reader may note that there is a rail line going between Downtown Boston and Hyde Park through Dorchester and Mattapan not marked on the map with yellow circles. This is the Fairmount Line, and it is one of the MBTA's commuter rail lines that is fully grade-separated. Unlike every other MBTA commuter rail line, the Fairmount Line does not have any stations beyond Boston's boundaries, making it the shortest commuter rail line in the system at just 9.2 miles. Realizing the area's lack of rapid transit, the MBTA started offering some of its most frequent commuter rail services along the corridor in recent years, with weekday schedules having a train depart every 45 minutes in each direction. While this does provide additional transit for neighborhoods along the corridor, 45-minute headways pale in comparison to true rapid transit, where schedules should be frequent enough that a rider does not need to worry about when the next train is coming when needing to use the train. While there is no exact headway that trains must achieve to qualify as rapid transit, the MBTA's Orange Line, which has the medium ridership of the MBTA's heavy rail rapid transit lines, currently has weekday headways of 6-11 minutes.

It is also worth noting that the MBTA operates a single bus rapid transit line that has yet to be mentioned. The Silver Line is a collection of five bus routes marketed as bus rapid transit, but in reality, only three of the five lines actually qualify as bus rapid transit and only for a short segment between South Station and the Silver Line Way in the Seaport via a tunnel and two underground stations. This tunnel and its two stations are the only rapid transit option for the Seaport, which has been redeveloped over the past twenty years into one of Boston's primary business hubs and fastest growing neighborhoods, housing the headquarters of General Electric, Reebok, Vertex Pharmaceuticals, and PTC among other companies. This development has been such a success that it has overwhelmed the Silver Line, with an additional service called "SLW" needing to run during peak hours on the bus rapid transit section between South Station and Silver Line Way. Additionally, since only SL2 serves the eastern portion of the Seaport, transit access is so poor that a developer proposed spending $100 million of private money to finance the construction of an aerial gondola to connect the area to South Station in 2018. No matter how many additional busses are added, the Silver Line will never be able to achieve the same capacity as Boston's rail rapid transit routes, as the Silver Line busses can only hold a maximum of about 100 passengers while trains can hold multiple times that amount.

MBTA Silver Line network, courtesy of the MBTA

A Solution North of the Border

To solve both of these problems, Boston can turn to a nearby city north of the border. Not the state's border in New Hampshire, but the nation's border in Canada. In 2015, Montreal found itself in a similar situation, with a commuter rail line traveling primarily in the city's borders through an area not served by the city's Metro and an underutilized tunnel connected to the city's primary rail station. With this, the Réseau express métropolitain, or REM for short, was born. The project, which is scheduled to open in phases between 2023 and 2024, is a fully automated metro system that combines the former Deax-Montagnes commuter rail line and Mont Royal Tunnel with additional spurs to create a 42-mile long system with headways as low as 2.5 minutes through the core. This is all wonderful, but arguably the most impressive aspect is the system's cost. The entire project is only projected to cost $6.9 billion in Canadian dollars, or a little over $5 billion in U.S. dollars, even though Canada is not immune from very expensive infrastructure costs like the United States.

Boston does not need a 42-mile long system, but a scaled-down REM could link Dorchester, Mattapan, and Hyde Park to the Seaport and beyond by converting the Silver Line tunnel and Fairmount Line to a similar rapid transit system. Infrastructure-wise, both key segments are already fully grade-separated, with the Fairmount Line only needing the addition of either catenary or third-rail electrification and the Silver Line tunnel needing both rails and the same electrification system. Connecting the two segments would be an approximately 1.25-mile-long tunnel, which would extend south beyond the current terminus of the Silver Line tunnel at South Station and go under Atlantic Avenue, South Station's approach tracks, Fort Point Channel, and the current Dorchester Branch alignment through the MBTA's Cabot Yard and Amtrak's Southampton Yard before rising to the surface prior to the line's existing rail bridge across the Southeast Expressway.

The existing Silver Line tunnel, in silver, and Fairmount Line, in purple, connected by the proposed tunnel, in brown, with all existing stations shown in blue, courtesy of Google Maps

All existing stations on both segments will need to be renovated to accommodate a new type of train. Additionally, given the wide spacing between both Blue Hills Avenue and Fairmount stations and Uphams Corner and Four Corners/Geneva stations on the Fairmount Line, new infill stations can be built at River Street and Quincy Street, respectively.

Fairmount Line segment, in purple, with renovated existing stations, in blue, and new infill stations, in green, courtesy of Google Maps

On the other end of the line, the existing Silver Line tunnel can be extended approximately 1.1 miles to the Boston Design Center, effectively replacing SL2's current route, before going under Reserved Channel and into the heart of South Boston and City Point, which is currently another rapid transit desert. With this extension, the remaining two Silver Line routes utilizing the current tunnel, SL1 and SL3, would instead terminate at World Trade Center, where existing bus bays can accommodate passengers wishing to transfer to the new REM-like service.

The existing Silver Line tunnel, in silver, with the extended tunnel to South Boston, in yellow, with existing stations, in blue, and proposed extension stations, in green, courtesy of Google Maps

In total, Boston's REM will extend over 11 miles, connecting the rapid transit deserts of Hyde Park, Dorchester, Mattapan, and South Boston to the historic economic center of the city at South Station and the booming business hub of the region in the Seaport. At South Station, riders will be able to connect to the Red Line, commuter rail services, and Amtrak via the existing station complex, while commuter rail connections will also be available at Readville. These connections to the rest of the MBTA network are critical for people along the corridor to be able to live a car-free or car-lite lifestyle. According to the Boston Planning and Development Agency, 31% of Dorchester households, 30% of South Boston households, 27% of Mattapan households, and 19% of Hyde Park households are already without a car, and the addition of Boston's REM will surely bolster these percentages to match the approximately 50% of households without a car in neighborhoods of the city that currently have rapid transit, such as South End and Mission Hill.

The same Boston Planning & Development Agency map from earlier with purple circles added to show a half-mile radius around all proposed Boston REM stations

The MBTA's Pushback

Without having ever discussed this plan with a single MBTA official, it can already be determined what one of their biggest, if not their absolute biggest, concerns will be: preserving access to the commuter rail's Readville Yard 2. Currently, the MBTA's South Station commuter rail operations utilize four layover yards to store their trains midday while not in use: Readville Yard 2, South Side Service and Inspection Facility, Amtrak's Front Yard, and Amtrak's Southampton Yard. While the latter three yards are within a 1-2 mile radius of South Station, Readville Yard 2 is about 9 miles away and requires trains to traverse the Fairmount Line. Commuter rail operations are not compatible with the metro-style service proposed under Boston's REM, so the commuter would effectively be blocked out of this space. 

This is definitely problematic at first glance, but the problem can be solved in two ways. First, the MBTA can run more commuter rail trains midday. In the current commuter rail system, layover yards are needed in Boston because schedules call for numerous city-bound trains in the morning hours and suburb-bound trains in the evening hours, meaning they need a place to sit during midday. However, the MBTA has already committed to reorienting as a regional rail rather than a commuter rail, meaning trains and schedules will be consistent throughout the day in both directions. This means layover facilities will not have as much use for midday storage and instead be for overnight storage, which can take place at the existing MBTA layover facilities at the suburban termini of commuter rail lines. However, the MBTA's Readville Yard 2 also has an equipment shop and other facilities, which leads directly to the second solution: Build a new commuter rail yard in the suburbs where land is cheap. Many European rail systems realized long ago that land is cheaper farther away from the city center, so that is where many systems chose to construct their rail maintenance facilities. The United States has been slow to learn this lesson, but the MBTA can follow this proven international model by constructing a new equipment shop and maintenance facility to replace Readville Yard 2 outside of Boston's city limits. The commuter rail's South Side Service and Inspection Facility is also not going to be touched, so midday maintenance can still be performed in Boston if needed, and the existing Readville Yard 2 is the perfect location to construct the vehicle maintenance and storage facility for Boston REM's.

The Path from Idea to Reality

The MBTA and state legislature have already shown a willingness to fund economical transit projects and avoid those whose costs are too high, which was demonstrated by the recently completed Green Line Extension that was only approved after shedding approximately $1 billion in costs. Cost-saving measures pioneered by the REM, such as reutilizing existing infrastructure, using fully automated trains, and building simpler stations, can be applied to make the project more appealing to both politicians and the public. However, what may be the most important part is that the Caisse de dépôt et placement du Québec, or CDQM, which is the institutional investor behind the REM, has expressed interest in exporting the REM business model elsewhere. 

There has not been a better time to build transit in Boston in recent memory. The mayor rides the MBTA to work every day, the governor-elect has prominently discussed her desire to invest in public transit, and the MBTA recently opened its first rapid transit expansion since the 1980s with the Green Line Extension. The time is now for the MBTA, the City of Boston, and the Commonwealth of Massachusetts to court the CDQM to build a REM-like system along the Fairmount Line/Silver Line corridor.

Welcome to Infrastructure Week!

Welcome to Infrastructure Week! On this blog, I will be writing weekly posts about infrastructure projects in the United States. 

Follow me on Twitter @infrstrctreweek to keep up with my posts!