The operating question has narrowed.
FRA’s Phase I study finds that Penn Station can raise peak-direction trans-Hudson capacity from 24 trains per hour to as many as 32 in the morning and 30 in the evening without expanding the station footprint. The tested plan keeps terminal service, introduces four commuter through-running trains per hour in each direction, and depends on targeted track, platform, and vertical-circulation work.
This does not settle the Gateway end state. It does settle the immediate design obligation: Penn Transformation should incorporate the tested Concept 1 package, preserve the wider-platform option for Concept 2, and avoid treating 32 trains per hour as the final regional capacity target.
Updated July 17, 2026 · Phase I published July 15 · Phase II due in 2028
Phase I replaces the old binary with a tested hybrid.
The earlier debate treated Penn Station as a choice between terminal operations and full through-running. FRA tested a third option: separate operating zones, retain useful terminal functions, reduce conflicting train movements, and add selective through-running where it improves capacity.
The 32-train morning case produced slightly more delay than the existing baseline; the 30-train case produced slightly less. Phase I therefore establishes a capacity range, not a guaranteed operating level.
Concept 1 is the actionable baseline. It retains the present number of station tracks, extends Platforms 1–3, adds up to 23 stairs and escalators, clears obstructions from platforms, and reorganizes train flows into three operating zones.
The plan supports four commuter through-running trains and four Amtrak through-running trains per hour in each direction. FRA estimates that the circulation package can shorten average platform clearance and boarding by 2.4 minutes, or 28 percent.
Neither Concept 1 nor Concept 2 reaches the 48-train Gateway objective under Phase I assumptions. The remaining capacity question moves to Phase II, where regional infrastructure, fleets, yards, service patterns, governance, and labor become part of the analysis.
The study now controls several track-level design decisions.
Penn Transformation is no longer free to treat service optimization as a later operating exercise. The pre-development agreement and station design must convert Phase I into specific geometry, access, and preservation requirements.
Dynamic simulation supports up to 32 morning and 30 evening peak-direction trains under the stated assumptions.
Build the Platform 1–3 extensions, new vertical circulation, platform decluttering, and track deconfliction into the base design.
A wider combined Platform 5–6 can support eight commuter through-runners per hour in a timetable, but the initial dynamic simulation failed under current variability and West Side Yard congestion.
Reserve structural and circulation space for the future platform. Do not remove Tracks 10 and 11 until a revised operating plan passes simulation.
Phase I does not test actual branch pairings, a fleet plan, yard and turnback strategy, governance, labor, or the full set of projects needed to approach 48 trains per hour.
Use Phase II to define the regional railway and compare any remaining expansion need against a fully specified optimized baseline.
Amtrak and PTP should show the platform extensions, vertical-circulation locations, new column positions, Hudson River Tunnel connections, preserved Platform 5–6 option, track assignments, dwell assumptions, and construction phasing before the design advances.
The next risk is premature design closure.
The master developer is under contract, the public architectural concept is established, and groundbreaking remains scheduled before Phase II is complete. Track-level choices made during the 2026–27 PDA phase could therefore limit the service plan that FRA is still developing.
The near-term test is concrete: the project must carry Concept 1 into design, keep Concept 2 constructible, and publish enough operating and modeling detail for independent review. Phase II can then determine whether regional investments raise practical throughput beyond 32 trains per hour and whether any external station expansion remains necessary.
The public record still lacks the native RTC and pedestrian-model files, the full 24-hour timetable, detailed platform assignments, and dispatching rules. Those materials would materially improve confidence in the capacity and reliability findings.
-
1
Integrate the tested package
Put the Concept 1 platform, circulation, and track requirements into the contract and design basis.
-
2
Preserve the future option
Protect the width, structure, and vertical circulation needed for a combined Platform 5–6.
-
3
Validate regional operations
Test actual routes, yards, fleets, reverse-peak service, reliability investments, and governance in Phase II.
-
4
Resolve the 32-to-48 gap
Identify which external projects can raise throughput and isolate any remaining case for expansion.
Core record and analysis
The links below separate the official study record from independent analysis, earlier feasibility work, and the underlying procurement and public-record files.
The official June 2026 report released July 15, including the operating concepts, simulations, infrastructure packages, and next steps.
→ Service Optimization Study WatchA focused implementation tracker for the Concept 1 package, Concept 2 preservation, model record, and Phase II.
→ Penn Station’s Missing ModelAn independent assessment of what the FRA model establishes, where the reliability limits sit, and what remains unresolved.
→ Through-running operating caseBackground on selective through-running, platform performance, yards, and regional rail integration.
→ Public-record archiveProcurement documents, agency correspondence, modeling records, and oversight materials related to Penn Transformation.
→Source note. Project-status statements derive from USDOT and Amtrak announcements dated May 20, June 8, and June 22, 2026. Capacity, service, infrastructure, and modeling statements derive from the Federal Railroad Administration’s New York Penn Station Service Optimization Study, Phase I Report, released July 15, 2026. The design and disclosure recommendations are independent interpretations of that record.