Pennsylvania Station, Manhattan question - Model Railroader Magazine

Amusingly, the original run through a bore of the completed North River Tunnel was with steam.

The operating concern involves the considerable grade exiting the tunnel 'in either direction'.  Either steam or diesel stalling at the bottom would result in terrific asphyxiation in a comparatively short time -- there was an accident in Italy during WWII that demonstrated the potential very effectively.  Even with the relatively short distance involved from Penn Station, by the time a recovery locomotive could be readied, brought into the tunnel, and tied on and the brakes on the train released, many passengers could be (expensively, if you remember Malbone St.) impaired or dead.

Unlike the situation at GCT, where "diesel" operation was often conducted at idle and sometimes at more than that with the FL9s that had inoperative third-rail pickup, PRR/PC/Conrail/Amtrak was always careful not to allow internal-combustion locomotive power through the tunnel.  During the time the Aerotrain ran, it was towed by GG1 going through the tunnel...

...what I don't know is whether the 567 in the Aerotrain had to be idling to supply the train with light and HVAC.  Likewise, I don't remember if all the small engines in some types of car, or the power unit of the 1950s Budd Tubular Train, were running in the tunnel or had to be shut down.

A consideration with the tunnel design is that they're not separately ventilated, as the Holland and later vehicle tunnels were.  If you are at the west portal, you can easily tell soon after a train enters from the east because displaced air comes out, proportional to speed.  Now, you can presume that there is also some of this air that presses backward into the partial vacuum created by the train's close passage.  Therefore any IC engine running to make power at the front of the train would preferentially pass heat and exhaust back along the sides of the train, which would require very good carbody sealing and recirculating HVAC circulation through the bore.  To some extent this would be minimized if the engine were pushing.  But even then, residual exhaust in the tunnel would affect a following train (likely as not to be on a close headway), and a stall or even a short traffic stoppage resulting in a stop in the tunnel would still have likely serious consequences.