To address these challenges, Amtrak, the State of New York, the State of New Jersey and the US Department of Transportation have recognized the need for the Gateway Program, a multi-phased series of investments to double capacity between Newark Penn Station and New York Penn Station from two tracks to four. Gateway would replace aging bridges, signals, catenary lines and tracks, build two new tracks under the Hudson River, and repair the existing Hudson River tunnels. The Hudson Tunnel Project and the Portal North Bridge are fully funded while other projects are in various stages of project development. Once the key elements of Gateway are completed, capacity between Newark and New York City will double from 24 to 48 trains per hour in the peak direction– if Penn Station can handle the additional service.

Proposals for meeting the existing and forecast demand include operational improvements, wider platforms, additional vertical circulation elements, different track alignments, and additional tracks and platforms. While there is consensus on the need for additional capacity, with all these variables, it is no surprise that different constituencies have proposed different strategies about how best to increase capacity at Penn Station, and even about what level and type of capacity is required.

These capacity challenges are not isolated to Penn Station. Strategies to make Penn Station operate more effectively could have negative impacts on other parts of the system. Any analysis of Penn Station operations must address the entire regional system and balance the needs of all transit riders across the different railroads.

To that end, this report examines the current operations at Penn Station and lays out the considerations that transit agencies must balance while expanding transit service throughout the tri-state metropolitan region.

While almost any operating strategy can be engineered and built, the goal is to find a solution which delivers the level of rail service that best meets regional transportation needs and supports the regional economy, maintains a high level of service reliability, preserves future flexibility, and causes the least disruption to existing operations and surrounding communities, at a reasonable cost within an acceptable time frame. In short: what is the best and most cost-effective strategy to maximize efficiency at Penn Station today, optimize regional connectivity, and provide for growth in the future?

More specifically, different groups have advocated for different approaches to resolving the capacity constraints at Penn Station. Most broadly, these debates have focused on a strategy of reconfiguring the tracks, platforms and operating systems of the railroads to provide revenue-to-revenue through-running service versus a physical expansion of Penn Station to add more tracks and platforms. And of course, there are many variations that would encompass some combination of both strategies.

RPA’s research concludes that converting Penn Station entirely to revenue-to-revenue through-running service within the existing footprint would likely reduce regional rail capacity and one-seat service to Penn Station. While there is potential for some revenue-to-revenue through-running within the existing Penn Station footprint, increasing the station’s capacity and regional connectivity requires more opportunities for passengers to board and alight in the Central Business District. That means a station expansion or some other form of system expansion will be necessary to address our region’s future transit needs.

Penn Station is owned by Amtrak and currently serves trains that are operated by Amtrak, NJ TRANSIT and the Long Island Rail Road (LIRR), which is operated by the Metropolitan Transportation Authority (MTA). Penn Station will be adding Metro-North Railroad service, also operated by the MTA, through the Penn Station Access project set to begin in 2030. The station currently operates at about three times its design capacity – with more than 600,000 daily passengers – by serving up to 24 trains per hour from the New Jersey side and up to 42 trains per hour from the Long Island/Queens side for a total of 66 peak hour trains into Penn Station at the height of the AM peak. The station also has one train per hour entering via the Empire tunnel which serves upstate New York locations including Albany and points beyond.

Train Movements and Peak Periods

The first step to understanding Penn Station operations is to look at the ways that trains can move in and out of the station.

There are five primary train movements at Penn Station:

Peak Periods

Peak periods are from 6 to 10 AM and from 4 to 8 PM on weekdays. Within each peak period there is a peak hour in the morning from 8 to 9 AM and the evening from 5 to 6 PM. They represent the busiest times of day when most trains are shuttling workers to and from their jobs. Peak period hours outside of the peak hour are often referred to as ​“peak shoulder” or just ​“shoulder” hours.

Tunnels, Interlockings, and Platforms

Station capacity is limited by the physical dimensions of the station, including tracks, platforms, stairs, elevators, and escalators. The following key factors determine train operations into, within, and out of Penn Station:

Tunnels

Penn Station is served to the west by the North River Tunnels that house a single eastbound and single westbound track across the Hudson River. Penn Station is served to the east by two sets of East River Tunnels that accommodate four tracks - two running eastward and two running westward - allowing trains to run between New York Penn Station and destinations further east. The East River Tunnels have a capacity of 21 trains each, totaling 42 trains per hour in the peak direction. The North River Tunnels have a capacity of 24 trains per hour in the peak direction.

The Hudson Tunnel Project, two tubes currently under construction as part of the Gateway Program, will also cross the Hudson River and connect Penn Station to points west. When the new tunnel and related projects are completed, they will double the westside tunnel capacity to 48 trains per hour in the peak direction into the station. This capacity would be slightly more than the current East River Tunnels that connect the station to Sunnyside Yard in Queens. Penn Station is also served by the Empire Tunnel which connects to Albany and points north.

Interlockings

Interlockings, by switching track connections, allow trains from one track to move to other tracks, distributing trains from the tunnel tracks to the station platforms and funneling trains back into the tunnels when they leave the station. The ​“A” interlocking accommodates trains exiting and entering the North River Tunnels (Hudson River crossing) as well as the Empire Tunnel. The ​“C” interlocking and the ​“JO” interlocking accommodate trains exiting and entering the East River tunnels. Other interlockings such as the Harold, the ​“F”, and the Plaza interlocking are used to move train sets into, out of, and around Sunnyside Yards.

Additionally, the more tracks a train crosses to get to its platform, the longer it takes to make the transfer and the more coordination required. A train entering Penn Station from Interlocking A to track 1 must cross 7 other tracks to reach its destination, while a train traveling from Interlocking A to track 11 does not cross any other tracks. Every track crossing requires coordination and potentially slows trains trying to access other tracks.

Combined, these interlockings create a web of track connections, increasing directional options, to allow for schedule corrections, and platform optimization, but limiting how quickly trains can access or clear a particular platform.

Platforms

Penn Station has 11 platforms that range in width from 18’to 20’, with two notable exceptions. Platform 6 is only 16’ wide while platform 10 is more than double that at 33’ wide. Wider platforms allow for faster, safer and easier circulation of passengers boarding and alighting, they also provide space for better vertical circulation. Revenue-to-revenue through-running service functions most efficiently at platforms that are at least 30’ wide. Wide platforms also allow boarding passengers to ​“pre-board” – waiting on the platform and immediately boarding once the alighting passengers have disembarked, similar to the way the NYC subway functions. With the exception of platform 10, Penn Station’s platforms are too narrow to operate this way. There isn’t enough space to safely disembark passengers from a full train while other people stand on the platform waiting to board. To prevent unsafe platform crowding, alighting passengers must completely clear the platform and exit to the station concourse before boarding passengers can begin to descend to the platform level and board their departing train.

Vertical Circulation

The stairs, elevators, and escalators that transport passengers and crew between the station concourse and the platforms comprise Penn Station’s vertical circulation. Each platform has multiple egress points, typically at least two elevators, although platform 5 has just one and platform 6 has three. Depending on the platform, stairwells are narrower or wider. Vertical circulation may have the most variability of Penn Station’s elements, but on average it takes longer than the National Fire Protection Association 130 (NFPA 130) mandated four minutes for all passengers to exit the train and ascend to the station concourse.

Limited passenger vertical access capacity and limited space for passenger queuing and circulation on the existing platforms requires trains to spend more dwell time on the platforms and slows down operations at Penn Station.

Tracks, Yards, and Trains

In addition to the elements described above, the tracks, yards, and trains themselves all have capacity limits.

Tracks

Penn Station’s 11 platforms are served by 21 tracks. Tracks 1-19 can access the North River tunnels crossing the Hudson River. The four southernmost tracks, Tracks 1-4, are known as ​“stub-end” tracks because they end at 7th Avenue and do not connect to the East River tunnels. Without these connections, trains on these tracks have to be turned back at the platform; they are unavailable for through-running. Tracks 5-17 can access the southern-most East River tunnels, and tracks 14-21 can access the northern set of East River tunnels. Tracks 14-17 can access both sets of East River tunnels. Tracks 20 and 21 connect to Yard C, a small train storage yard to the west, and can connect to the West Side Yard, but trains entering on these tracks are not able to connect to tunnels running to New Jersey and thus are not able to through-run to New Jersey. As a practical matter, only the center 15 tracks (tracks 5-19, which access platforms 3-10) can be programmed for through-running in the existing configuration.

Track configurations vary across the system. There are at least three critical segments for trains entering and leaving Penn Station: between the Harold Interlocking and Jamaica station in Queens, between Millburn Station and Kearny Junction in New Jersey and between Floral Park and Hicksville. In each of these cases, more tracks serve peak direction than reverse-peak direction trains.

There are four tracks that serve trains on the seven-mile stretch between the Harold Interlocking in Sunnyside and Jamaica, Queens. Three of these tracks are used by trains running in the peak direction, with one track set used for trains running in the reverse-peak direction.

Similarly, from Millburn on the Morris and Essex Line in New Jersey, to the Kearny Junction where the Morris and Essex joins the Northeast Corridor, and in the stretch between Floral Park and Hicksville on Long Island, there are three tracks, two of which are used in the peak direction and one reserved for trains running in the reverse-peak direction.

In these cases, more tracks are devoted to trains heading to Penn Station in the AM and away from Penn Station in the PM; this accommodates the vast majority of trips.

Yards

Train yards are where trains are stored and maintained between passenger service. For drop-and-go service at Penn Station, LIRR trains entering from the east drop passengers at Penn Station platforms and then continue west, where they are stored and turned at the West Side Yard (WSY) in Manhattan. WSY is owned by the MTA. Many NJ TRANSIT trains entering from the west drop passengers at Penn Station platforms and then continue east under the East River to be stored at Sunnyside Yards in Queens (SSY) which is owned by Amtrak.

WSY, which is primarily a storage facility, has capacity to store 30 trainsets. SSY can hold a similar number, but being a storage and maintenance facility, the operation is more dynamic. NJ TRANSIT has an agreement with Amtrak that allows them to store 19 trains at SSY during the day and 6 overnight. Amtrak cleans and services their trains at SSY during the day and uses SSY to store trains overnight.

Trains

While some trains are single level, NJ TRANSIT’s fleet includes many double-decker passenger cars, including all the new cars purchased since 2006. NJ TRANSIT locomotives entering Penn Station are all electric or dual fuel (electric and diesel) units with overhead catenary power connections. Long Island Rail Road (LIRR) is also committed to purchasing double-decker passenger cars where they make sense, but tunnel height restrictions will require single level vehicles well into the future. LIRR locomotives are also electric or diesel. The electric locomotives are powered by third rail electric delivery. As a result, all trains serving Penn Station enter under electrical power, but NJ TRANSIT and LIRR trains use different electrical systems and the trainsets are built according to their respective power requirements.

All commuter trains are designed to balance the trade-offs between capacity, passenger comfort, and speed of boarding and alighting. NJ TRANSIT and LIRR average commute lengths are 26 and 21 miles respectively and many trips are considerably longer, making amenities such as comfortable seating and restrooms important priorities. While subway and bus riders on shorter trips may not mind standing in a vestibule, this would be a problem for commuters on longer trips from further stations.

The physical constraints of the tracks, platforms, vertical circulation, train design, interlockings, tunnels, and yards, combined with current day-to-day operational decisions create Penn Station’s current capacity limitations.

Capacity between Penn Station and New Jersey is limited by the trans-Hudson tunnels. The existing tunnel capacity is 24 trains per hour in each direction. When new Hudson River tunnels are built as part of the Gateway Program, the tunnel capacity will double to 48 trains per hour in the peak direction and the capacity bottleneck will shift to the east and west of the Hudson River, including New York Penn Station’s platforms and vertical circulation to the east, and the two-track alignment to the west. Other elements of the Gateway Program, including Secaucus Junction Capacity expansion, the Portal South Bridge, replacement of the Sawtooth Bridge, and the Newark-Harrison Systems Modernization, will double capacity between the Hudson River and Newark Penn Station. This section explores the limits to New York Penn Station capacity once the full Gateway program is completed.

Averaged across the day, one train enters and another exits the station every 2.25 minutes. In the AM peak period (6 am - 10 am), that frequency is closer to one train every 1.5 minutes, and in the AM peak hour (8 am - 9 am), it is one train every 1.25 minutes. Throughout the day, about 38% of the trains drop passengers at Penn Station and then run through to a yard (either WSY or SSY), while another 18% through-run in passenger service on Amtrak trains. While 44% of trains are turned back daily, since these operations are less efficient only 11% of the trains are turned during the peak period and just 3% during the peak hour. Turn-backs are very time intensive and should be avoided, except for the stub end tracks where they are the only option.

Dwell Time

The amount of time a train spends at the platform is called ​“dwell time.” Operators want to minimize dwell time to free up platforms for the next incoming train.

Still, it takes some time to stop at the platform, unload passengers, sweep the train for people who may have fallen asleep or not realized it was the last stop and exit the platform. The time that the train dwells at the platform is 7 minutes in reasonable operating conditions: 5 minutes to unload passengers, 1 minute to check for stray passengers and then 1 minute for schedule recovery. There’s an additional 2.5 minutes to clear the interlockings and for train safety separations.

Accounting for all these aspects, under reasonable operating conditions,each track can accommodate 6 trains inbound per hour using drop-and-go service. Inbound trains have different dwell times from outbound trains, and long-distance Amtrak trains have longer dwell times than the commuter railroads.

To execute the non-revenue-to-revenue sequence, a train is brought into the station and only then is the track announced so passengers can move from the station hall down to the platform level to board. Just like boarding an airplane or bus, this process is inherently slower than alighting the train as passengers have to find the track, reach the track, and board the train before it can depart. The process of bringing an empty train into Penn Station, loading passengers and then departing takes 9 minutes in reasonable operating conditions during the PM peak period. This includes 7 minutes for passengers to hear the track announcement, descend to the platform and load into the train and 2 minutes for schedule recovery. Once loaded an additional 2.5 minutes must be considered for the train to clear the interlockings and for train safety separations.

Accounting for all these aspects, under reasonable operating conditions, each track can accommodate 5 trains per hour using load-and-go service.

Track and Platform Capacity

An inbound track capacity of 6 trains per hour might seem to indicate that two tracks should be able to serve 12 trains per hour. However, most tracks at Penn Station share a platform, and it is the platform capacity, rather than the track capacity, that determines how many passenger loads can be served. Due to Penn Station’s narrow track platforms, two trains should not arrive on either side of a platform and have their passengers disembark simultaneously. The crowding from people standing on the platform, waiting to access the stairs and elevators prevents the next train from using the platform. However, by staggering when trains arrive on a common platform, the platforms can serve 9 trains per hour in drop-and-go operations.

With each drop-and-go platform able to serve 9 trains per hour, the effective station capacity for an all drop-and-go operation in the AM peak is surprisingly large.

The platform can serve slightly more trains in the PM peak using load-and-go, up to 10 trains per hour. This increase is because with load-and-go passengers board the train directly keeping the platform clear, making it safe to pull an empty train on the other side of the platform. This is in contrast to drop-and-go where there is crowding at the platform due to the bottleneck of the vertical circulation.

Estimated Capacity at Penn Station

If no train is delayed and all trains travel in the peak direction for the AM peak hour, under all possible revenue to non-revenue (drop-and-go) or non-revenue-to-revenue service, the current station could hypothetically achieve 73 inbound trains per hour or 80 outbound trains. To match the future tunnel capacity (42 trains to and from east, 48 trains to and from west) our station capacity must be able to reach 90 trains per hour in the peak direction. Even if we devote all of our resources to drop-and-go or load-and-go, Penn Station with its current orientation still cannot meet the peak tunnel capacity. This does not even account for reverse peak movements which will still be necessary under any operating scheme.

The estimate also assumes that trains leaving Penn Station have somewhere to go – ideally a yard. At the moment there is insufficient yard capacity. Both SSY and WSY are already operating at their maximum capacity. SSY has some potential to increase storage, but at the expense of maintenance and other operations. Additional land would have to be acquired to increase storage. The land acquisition question is discussed in the section ​“other considerations.”

These constraints raise the question of whether there are any changes to railroad operations that could increase capacity at Penn Station. Revenue-to-revenue through-running has been proposed as a potential operational improvement. In the next section we will examine how revenue-to-revenue through-running at Penn Station would affect station and system operations.

Through-running can deliver regional connectivity by providing riders with one-seat rides across a metropolitan area. There are many examples of through-running service providing excellent regional connectivity and service frequency, including the RER in Paris and the Elizabeth Line in London. However, revenue-to-revenue through-running is not guaranteed to increase capacity or connectivity. The success of these operations depends on station dimensions, track alignments, and where passengers are coming from and where they are traveling to.

A critical question for Penn Station operations is whether revenue-to-revenue through-running in the New York metropolitan region could provide a similar benefit of connectivity between areas east and west of Manhattan. Although technically possible, an analysis of Penn Station and the larger commuter rail system shows that an all revenue-to-revenue through-running system at Penn Station would actually decrease the station’s peak period passenger through-put while also disrupting wider system capacity and regional connectivity.

To better understand the potential impacts, RPA analyzed several through-running scenarios. The first scenario assumes that all Penn station tracks that connect to tunnels on both sides of the station are used for revenue service through-running. In this scenario we maintain the same number of tracks (21 tracks, of which 17 are configured for through-running) and the same number of platforms with the existing widths (11 platforms, of which 8 serve through-running tracks). All trains run through except those on stub-end tracks which cannot through-run.

As with drop-and-go, in revenue-to-revenue service the train enters the station and discharges passengers. Unlike drop-and-go, additional time must be budgeted for the passengers to clear the platform and reach the concourse before reboarding can commence. The platform must be completely cleared before a track announcement can be made to avoid the conflict of passengers trying to come down the stairs and escalators to board while the departing passengers are heading up. Once the announcement is made, passengers must find their track, come down to the platform level, and board the train.

It would take an estimated 6 minutes for passengers to deboard, clear the platform and reach the concourse and another 4 minutes for passengers to descend to the platform and board the train during the AM peak. 2 additional minutes must be added to provide a buffer that ensures schedule adherence. The trains also require 2.5 minutes for clearing interlockings and safety separation, resulting in an hourly capacity of 4 trains per track and potentially 8 per platform.

In the PM peak it would be reversed where it would take 4 minutes to deboard and clear the platform and 6 minutes for passengers to board. These differences are due to the asymmetrical demand; more passengers arriving in the AM peak and more departing in the PM peak. With 2 additional minutes to provide a buffer for schedule and 2.5 minutes for clearing interlockings and safety separation. This results in a similar track capacity of 4 trains per track and up to 8 trains per platform.

With wider platforms, passengers can be waiting on the platforms when the trains pull in, reducing dwell time by the time reserved for track announcement and allowing passengers to descend to the track level before the train arrives. We estimate that the dwell time for a revenue-to-revenue through-running train on wide platforms would be 7 minutes: 5 minutes to load and unload and 2 minutes for schedule correction. In addition, 2.5 minutes would also be needed to clear the interlockings and allow for safety separation. This would decrease the dwell time, but widening platforms would require a reduction in the number of tracks and platforms. Under one proposed scenario, platforms could be widened and 9 tracks removed (leaving a total of 10 tracks), which would generate a capacity of only 63 trains per hour, which is only 70% of the future tunnel capacity.

This analysis shows that moving to an all through-running station would reduce overall capacity at Penn Station.

Other Constraints and Considerations

Along with capacity constraints at the station, a full revenue-to-revenue through-running scheme within the footprint of the existing Penn Station presents additional challenges:

Turning and tunnel constraints

While some of today’s trains do turn at Penn Station, all of the trains must turn somewhere on their branch lines. Given the highly asymmetrical nature of the demand, with almost all passengers looking to travel to the Manhattan central business district (CBD) during the AM peak and departing from the CBD during the PM peak, station pairing for revenue-to-revenue through-running would generally be relatively close to Manhattan. It is not practical to schedule trains from the end of one LIRR branch to the end of a NJ TRANSIT branch (such as Babylon, NY to Trenton, NJ) because the passenger demand doesn’t exist to justify the crew expense and the costs of purchasing and operating the equipment. The most effective utilization of through-running would be for relatively short runs making stops within the metropolitan core and then turning back. For example, a service running from Hempstead, NY to Secaucus, NJ would have higher demand at the terminals than a service from Trenton, NJ to Babylon, NY.

Once these short runs beyond Penn Station are complete, the trains will either need layover space at the destination, or to turn back from the destination preempting trains from further away despite their likelihood of having fewer passengers. For example a train traveling from Babylon, NY through Penn Station and on to Secaucus, NJ in the AM peak would have to layover in Secaucus or turn around at Secaucus and then continue back through Penn Station and on to travel in the reverse peak direction east, back to Babylon. To layover, that train would need to have a railyard in Secaucus where it could be stored until making its return trip to New York and Babylon. Because regional demand is not balanced – many more travelers want to access Penn Station in the morning, and many more want to leave in the evening – it would not be beneficial for the railroad to turn that train at its terminus and immediately run it back through the core of the region. Doing so would compromise peak direction capacity.

During the morning peak, when demand to enter the core is highest, this train turning at Secaucus and traveling East would add to the queue of trains traveling into Penn Station during the AM peak, competing for limited tunnel space with trains traveling from further out. In this scenario, we would expect that the train turning at Secaucus will have relatively few riders compared to trains traveling from further away, because it will have served far fewer stations. In the PM peak, the train turning at Secaucus will be traveling against the regional demand, once again reducing overall capacity.

Another constraint is that many parts of the system rely on sections of single tracks while others rely on multiple tracks with unequal travel.

For example, in the Kearney-Milburn section: there are three tracks, two tracks devoted to the peak direction, and one track devoted to the reverse-peak trains. The single reverse-peak track limits the ability of this line to receive LIRR trains and provide both local and express service to the Morris & Essex Line.

Many branch lines throughout the region have portions of single track. New tracks would have to be laid throughout for a consistent through-running operation.

Branch combinations and connectivity

While a train enters and exits Penn Station on average every two and a quarter minutes, these trains are convergent and not substitutable. Unlike the NYC subway, the Paris regional rail (RER), or the Elizabeth line in London, where trains follow each other along the same paths, the intensity of service at Penn Station is the sum of one to four trains each hour on branch lines that all converge at the tunnels leading to Penn Station. With 10 branch lines on either side of Penn Station there are 100 branch-to-branch combinations. With through-running, the odds that the branch you started on is scheduled to continue on your destination branch are very low. Hence the connectivity upgrade would also be very low. In other words, even in a revenue-to-revenue through-running service, if you are a passenger traveling from Long Island on the Hempstead branch and going to a New Jersey destination on the Gladstone branch, you are likely going to need to transfer at some point during your trip, because the train you boarded on the Hempstead branch could be connecting to the NEC, or Morris & Essex or some other branch at the time you wish to travel.

Another connectivity question is whether greater connectivity is created by through-running or by creating more direct service between Penn Station and the NJ branch lines that do not currently serve Penn directly. Currently, there are three NJ TRANSIT lines (Raritan Valley, Main and Bergen Lines) and two MTA supported lines (Port Jervis and Pascack Valley lines) that do not have a one-seat ride to Penn Station during peak hours.

The service plan that undergirds the need for 48 peak hour trains across the Hudson River provides greater connectivity between stations on the Raritan Valley Line, the Main Line, and the Bergen County Line as well as the MTA-supported Port Jervis and Pascack Valley lines. A revenue-to-revenue through-running scheme within the current Penn Station footprint would not be able to accommodate the full 48 trains per hour in the peak direction, meaning all or some of these lines would not get the connectivity benefits.

Land acquisition

To accommodate the additional remote layover and/or turning operations, thus integrating the three commuter rail operations to a full through-running operation would require additional railyards to allow trains to turn around without running to the ends of their lines. Some options that have been proposed by outside advocacy groups include:

• Land acquisition in the Bronx (approximately 40 acres) to build a new railyard
• Land acquisition in Secaucus (potentially 46 acres) to build a new railyard

Building these yards would require significantly more land than expanding Penn Station to create additional track capacity.

Power systems

The train sets that use Penn Station are currently powered with different technologies. NJ TRANSIT uses an overhead catenary for its power supply. LIRR relies on a third rail power supply. The fleets will have to be brought into compatibility using either hybrid designs, relying on emerging battery technology or by switching over the power delivery mechanism of one of the systems. It will take time to effect a transition and it must be addressed along with the other challenges.

Overall System Tradeoffs

In summary, a full revenue-to-revenue through-running operation at Penn Station appears to require a significant system expansion– perhaps even greater than an expansion of Penn Station alone. It would require greater land acquisition than expanding Penn Station without through-running. It would require new tracks to be laid and new points where the trains would turn. It would require new tracks and platforms at Penn Station in order to accommodate the full forecast demand. It could require additional tunnel capacity as well.

Operationally, storing and turning a train in Secaucus (or some other new destination) is effectively no different than storing and turning it at the WSY. The difference would be that passengers may stay on the train to Secaucus. Demand for that service has not been analyzed, but it is unlikely to supersede demand for travel to the CBD. Elected officials and agency heads should weigh these trade-offs in making any decisions related to through-running and system expansion.

RPA’s analysis shows that Penn Station is currently constrained by a number of factors, including tunnel capacity. Once the new Hudson River Tunnels are built and other track improvements from the Gateway Program are completed, the constraining bottleneck will become a combination of conditions at the station and the railyards on either side of the station. Operational changes can unlock some of the station constraints, but wouldn’t achieve the goal of creating direct service from each branch with a minimum service of two trains per hour during peak periods. Platform and yard constraints will still be a concern.

The recommendations below can begin to be implemented today and will contribute to greater capacity when the tunnel bottleneck is eliminated. Further capacity improvements will rely on increased physical space and capacity of the station and/or rail yards.

Near-Term Through-Running Applications

Based on this analysis, converting all operations at Penn Station to revenue-to-non-revenue (drop-and-go) and non-revenue-to-revenue service would result in the highest passenger through-put at Penn Station, but implementing this operation, under existing operating parameters, would require more yard storage space and likely more rolling stock to implementing this system.

Nevertheless, there are some applications of revenue-to-revenue through-running service that could increase efficiency while acting as a pilot to examine demand for one-seat ride commuter rail travel between New York and New Jersey.

Increase existing special event through-running

The commuter rail system already has the capability to operate revenue-to-revenue through-running, and has been previously used for a limited number of events at the Meadowlands in New Jersey including the ​“Train to the Game” program that ran NJ TRANSIT trains on Metro-North tracks on the New Haven Line. This strategy of revenue-to-revenue through-running should be expanded to include additional sports games and special events.

Peak period ​“second run”

Another opportunity is to reevaluate the use of rolling stock during the shoulder period. To make efficient use of our rolling stock, commuter railroads must utilize the same train sets for multiple runs throughout the day, resulting in the train turnbacks that occur at Penn Station during the shoulders of the peak periods. Some of these trains run back in revenue service and others run back empty (deadhead) to an origin station to pick up a second run.

An operating regime where the shoulder trains run through in revenue-to-non-revenue and/or revenue-to-revenue service to the start of a branch line, replacing the trains that are turned today, would increase the efficiency of the existing infrastructure. This would replace the current deadhead trains and serve the low percentage of trains that operate in the reverse-peak direction. For example, a LIRR train set could enter Penn Station in the morning peak period, discharge its passengers and then deadhead or operate in revenue service to a branch line in New Jersey to pick up a second run. Likewise a NJ TRANSIT train might drop its passengers and continue to the LIRR branch line to pick up a second run, replacing an LIRR train that would have otherwise been turned at Penn Station. This kind of exchange would eliminate the need to turn a train and would save space in the railyard.

A prerequisite for this efficiency gain would be interoperability meaning rolling stock that can operate on different power systems, and renegotiated labor agreements.

Standardization of rolling stock and power delivery

The systems differ in how the trains are powered, creating incompatibility for through-running on the commuter railroads. Future rolling stock purchases should be hybrid, i.e. trains that can run on either overhead or third rail. Other forms of compatibility could include battery power, but that technology has limited distance applications. The standardization of rolling stock and power delivery is a minimum pre-requisite for interoperability between networks which could support running an MTA train on NJ TRANSIT tracks and vice versa.

Through-ticketing

Branch-to-branch and system-to-system travel can and should be facilitated via through-ticketing, fare integration, good wayfinding, and easier transfers. A passenger from Montclair, traveling to Kennedy Airport should be able to purchase a ticket from Montclair to Jamaica, or any other station combination. A through-ticketing system should look to encourage regional connectivity by offering a discount to those traveling from New Jersey to Long Island and vice versa such that a ​“through-ticket” would be less expensive than the two tickets purchased separately.