Penn Station's Missing Model Is Now Public

FRA has now tested the hybrid RPA omitted

The RPA’s analysis screened only two extreme scenarios: through-running on today’s platform layout, and a wider-platform concept requiring the removal of nine tracks. By comparing these options against its own 90-train benchmark, the RPA treated the resulting capacity shortfall as automatic evidence for physical expansion.1

The FRA’s Phase I analysis took a different route, retaining terminal operations and introducing only the specific through-running needed to deconflict the throat and shorten platform dwells. Concept 1 preserves existing Long Island Rail Road (LIRR) volumes, expands NJ Transit and Amtrak service, and routes up to four commuter trains per hour through the station in each direction. Once the proposed platform and circulation upgrades are in place, the RTC simulation shows that a 32-train morning peak is dynamically feasible under current delay conditions, though a 30-train plan maintains better reliability.23

While this capacity increase does not prove the existing station can absorb the full 48-train Gateway target, nor eliminate the eventual need for expansion, it completely invalidates the RPA’s core inference: that the failure of total through-running proves the immediate necessity of an annex. The controlling design baseline is now a staged hybrid, not the binary choice the RPA evaluated.

The model exists; the reproducible record does not

The Phase I report significantly narrows the earlier evidence gap. It names the specific RTC and pedestrian simulation software, publishes three distinct structural dwell profiles, provides platform-access diagrams, outlines directional train counts, describes how delays are seeded, and clearly distinguishes a paper timetable from a successful dynamic simulation.23

However, the FRA has withheld the raw RTC and PTV (pedestrian simulation) files, complete timetables, run logs, and sensitivity outputs, preventing independent verification or reproduction of the results. To simplify the model, the study standardizes train lengths, treats equipment as indifferent to specific origins and destinations, and assumes every modeled LIRR train runs directly to or from West Side Yard—a simplification the report itself notes is an unlikely tactical operating plan.23

These simplifications directly affect both concepts. While Concept 1 passed dynamic testing at 32 peak trains per hour, Concept 2 failed to survive simulation once real-world arrival variability and West Side Yard congestion were introduced, despite producing a viable paper timetable. The FRA has deferred the operational corrections and necessary regional investments required to make Concept 2 work to Phase II.23

Furthermore, this new model is not the missing simulation behind the controversial October 2024 engineering feasibility study. Amtrak previously confirmed under FOIA that no simulations were performed for that earlier review. The Phase I study is an entirely separate 2025–26 effort with different objectives, inputs, and hybrid operating concepts.24

The disclosure debate has thus narrowed: the public no longer lacks an operating model; it lacks the raw simulation binaries and the Phase II regional service plan needed to verify these results before locking in irreversible track geometry.

The arithmetic behind the conclusion

The RPA organizes Penn Station’s capacity limits around five specific movement types. This taxonomy is critical because the RPA’s resulting capacity hierarchy is entirely driven by the specific dwell times, clearance periods, and platform-use restrictions it assigns to each movement.

The FRA’s modeling recalibrates this arithmetic. Its pedestrian simulations analyze platform-level flows and conclude that adding vertical circulation, decluttering platform space, distributing Amtrak passengers, and providing earlier NJ Transit track announcements can reduce average platform occupancy times by 2.4 minutes. Crucially, this benefit is highly conditional on those specific operating practices; building the physical stairs alone will not generate the modeled dwell reductions.23

The benchmark must match the operating question

RPA's 90-train benchmark and FRA's 32-train finding measure different dimensions of capacity. RPA's benchmark represents a combined east-and-west terminal throughput requirement (48 peak-direction approach trains from the west added to 42 from the east). In contrast, FRA reports practical, reliability-constrained trains per hour per direction solely under the Hudson River (the western approach): 24 today, up to 32 in the morning peak, and up to 30 in the evening peak under Concept 1.123

The FRA’s practical capacity results are substantial but incomplete. They unlock 25 to 33 percent more cross-Hudson throughput within the existing footprint, falling short of the full 100 percent increase associated with the Gateway project. Phase II must explain how future demand, regional power and signaling investments, and any remaining expansion plans connect this near-term 32-train concept to the long-term 48-train objective.23

The 2024 feasibility study’s most compelling objection to through-running was the complex interaction among reverse-peak schedules, mixed intercity and commuter trains, platform allocation, and tunnel headways.2 The Phase I study does not erase these physical limits. Instead, it sidesteps the earlier study's 'fatal flaw' conclusions by retaining terminal operations and restricting commuter through-running to a modest four trains per hour under Concept 1.

The federal baseline is now a staged hybrid

RPA's principal screens do not match the Phase I concept selected by FRA. The federal baseline keeps the existing station footprint and most of its track configuration, separates the station into three operating zones, retains terminal service, and adds limited commuter through-running.

Concept 1 provides the actionable design basis. The master developer should integrate its platform extensions, track deconfliction, and vertical circulation into the Penn Transformation project immediately. In contrast, Concept 2 functions as a spatial preservation requirement rather than an active design plan. The design should reserve space for future stairs and avoid placing new structural columns where Tracks 10 and 11 might eventually be removed, but the agencies must defer permanent track removal until a Phase II service plan passes dynamic simulation.23

The distinction controls the next design decision. Penn Transformation can lock in circulation capacity and preserve future platform widening without claiming that the super-wide-platform operating plan has already passed reliability testing.

The connectivity substitution

The RPA identifies ten distinct rail branches on each side of Penn Station, creating a matrix of 100 possible branch-to-branch routes. Based on this matrix, the RPA argues that any single through-running train is highly unlikely to match a rider's desired origin and destination, illustrating the point with an obscure Hempstead-to-Gladstone pairing.1

This framing substitutes raw combinatorics for actual origin-destination market analysis. Modern transit agencies build high-capacity trunk lines around established passenger demand and track geometry, not randomized branch-to-branch pairings. A credible service design would select a small number of high-volume cross-regional corridors and use timed transfers elsewhere. Evaluating every theoretical branch pairing misses this basic planning reality.

The FRA adopts a narrower, more defensible position. Its Phase I study treats through-running strictly as an operational tool to deconflict tracks and shorten platform dwells, without claiming that a cross-regional travel market already exists. Phase II will conduct actual demand forecasting and model structured service networks—an essential planning sequence that the RPA's abstract branch matrix cannot replace.23

West Side Yard is the controlling unresolved operating issue

Just as theoretical branch pairings can oversimplify regional travel demand, idealized yard routing can mask severe station constraints. Phase I makes this yard issue concrete. To simplify the simulation, the model assumes every LIRR train begins or ends its run at West Side Yard, meaning no LIRR trains turn back at Penn Station's platforms. The FRA explicitly calls this an 'unlikely tactical operating plan' because reversing a train inside a storage yard is slower and introduces more operational variability than turning it at a station platform.23

To resolve the Concept 2 simulation failure, the FRA identifies two possible corrections: turning some LIRR trains directly at the Penn Station platforms, or routing a larger share of commuter trains through the station in revenue service. Phase II must test these operational alternatives, define a regional fleet and routing plan, and identify the supporting line investments required. Until that dynamic modeling succeeds, the proposed super-wide platform should remain a protected future option rather than a committed physical footprint.23

The technical record has changed

The RPA’s historical support for through-running relied on significant new infrastructure, specialized fleet equipment, and unified regional operations. Its April 2026 report shifted focus, screening two extreme, within-footprint options. The FRA’s July 2026 study supplies the critical, missing intermediate case: selective through-running integrated with a retained terminal station.

The new study does not settle Penn Station's ultimate capacity requirement. It does establish that RPA's April screening result cannot serve as the controlling design basis without confronting the federal hybrid model.

Advocacy context raises the disclosure standard

The RPA leads the *Build Gateway Now* advocacy coalition. Freedom of Information Act (FOIA) disclosures show that the RPA administered Amtrak-funded coalition management and research.20 Specifically, funding records show that Amtrak paid the RPA $1.825 million from FY2015 through FY2024 for restricted Gateway advocacy and Build Gateway Now campaigns.22 Furthermore, Amtrak's 2024–25 campaign proposal explicitly added through-running research to the RPA’s funded work plan, noting that the resulting report would present full through-running as disruptive and frame the Penn South expansion as the only practical path forward.20

While this financial history does not prove that Amtrak dictated the RPA’s conclusions, nor that the resulting report is technically incorrect, it demonstrates that the analysis emerged from an advocacy partnership. This relationship, wherein Amtrak funded the RPA's coalition work and designated through-running research in the active work program, significantly raises the disclosure standard for the report's decisive, underlying operating assumptions.

What remains before final geometry

The FRA’s Phase I study provides the first model-based evidence that the existing Penn Station footprint can deliver a 25 to 33 percent increase in cross-Hudson capacity through a hybrid operating plan and targeted physical upgrades. This finding materially weakens the RPA’s claim that the failure of total through-running proves the immediate necessity of terminal expansion.23

The Penn Transformation project should actively incorporate Concept 1 and preserve spatial options for Concept 2. It should not permanently remove Tracks 10 and 11, treat 32 trains per hour as the ultimate Gateway solution, or use Phase I to foreclose future expansion until Phase II fully validates a regional service plan and evaluates all remaining alternatives on a common demand and reliability basis.