# Trains, tunnels and air

I was recently thinking about trains in long tunnels. They have to displace air in the tunnel when travelling fast which slows them down and although running in a vacuum would be ideal, it would be difficult (and dangerous) to maintain.

How feasible would it be to have a train with a large impeller at the front which would draw in and compress the air and vent it (through a tube) at the rear or the underside of the train? Could an impeller, powerful enough to suck in the air in front of a fast moving train, be driven by electricity? Would an increase in efficiency balance out the power needed to drive the impeller? Could speed be greatly affected (I’m thinking of the reduction of friction rather than any development of thrust which would require a fuel - hence impeller, not jet)?
I remember seeing a train once with a big propeller at the front, but that was designed as a snow-plough, but that’s OT…

I think it would be hard to fit a passageway through the train that had anywhere near the surface area that already exists outboard of the train and inside the tunnel. The pressure it takes to accelerate air is proportional to the square of the velocity to which you accelerate it, so the incentive to have a bigger passageway is tremendous. Plus, the compression and so forth would be imperfectly efficient.

It might be that trains also slide the column of air along the length of the tunnel. Air weights a kilogram per cubic meter, and a train that fills more than a percent of the tunnel’s length will weigh more than the air filling the tunnel unless the tunnel is much wider/taller than the train (which would fix the problem anyway). So before it does much work to fight against the air, the train might also just shove it up the tunnel.

Most trains run through tunnels for only a small proportion of the time, making any saving in energy fairly insignificant. And in rail networks where trains run primarily underground, such as the London Underground, pushing air through the tunnels is often the only way to ventilate the system.

In other words, an impeller on the front of each train would be an expensive solution to a problem that is either minor, or not even a problem if you look at it the right way.

The thing is, the trains in the United States that use long tunnels really aren’t going all that fast anyway - 79MPH being the upper speed limit in 90+ percent of the territory. Add the fact that tunnels usually occur in mountainous territories, which limit speed by grade and curve gradient, and it’s just not much of a problem.

On the other hand, trains in most of the rest of the world that travel through tunnels are already electric - I’m thinking specifically about the Eurostar that runs through the Chunnel. In fact, I’m not aware of any high speed train in the world that doesn’t run on straight electricity (jets and various turbines have been used, but that’s never really been more than experimental.) I don’t know enough about the TGV or Shinkansen to know whether they travel through many tunnels, but both are electric.

Some (most?) of the larger tunnels in the US have exhaust systems to clear the diesel fumes - both for the benefit of the next crew as well as the diesel engines, which obviously need to breathe. So it’s not as if the two portals are the only intake/exhaust ports. Quoting from the current issue of Trains magazine (October 2007): “…Stampede Pass [Washington] had one key thing going for it: The 7.79-mile Cascade Tunnel at Steven’s Pass must be ventilated after each train passes, a process that takes at least 30 minutes.”

In short, I’m guessing that no, the relatively minor increase in efficiency wouldn’t justify the cost of powering it - either we’d electrify the railroad (which would then put us on par with Europe, but since we have hundreds of thousands of miles of track, it’s not feasible) or we’d modify the diesel-electric locomotives to supply the power (or, I suppose, have each tunnel supply its own power by some means.) Then, GE and EMD would have to offer that option on locomotives - an expense - that only a few power divisions would need (the mountain divisions with tunnels) which would be subsidized by all the other railroads/divisions that don’t need the modification.

Also keep in mind that many tunnels are at grade - so (for example) the Westbound train would be using the diesel engine to load amps to the traction motors, effectively ‘working’ to get the train through the tunnel, the Eastbound train would be going downhill, probably in dynamic braking, and would have no need to displace air in the tunnel - that train is trying to decelerate.

Sometime back in the maybe '60s, Scientific American had an article about future railways. The author(s) imagined, among other routes, a tunnel bored straight (as in shortest distance) through the earth. If we suppose that the angular distance between NY and LA is 45 deg (which is the width of three time zones), then at the midpoint of its trajectory, the train would be (1 - cos(22.5))*4000, a bit over 300 miles deep. The basic idea is that the tunnel would be evacuated fore and pressurized aft and the train would accelerate under the influence of gravity as it descended and decelerate as it ascended and the pressure differential would make up for frictional losses. What would happen if there were an accident 300 miles underground ws not addressed. Frankly, I think it belonged more in Astounding than in Sci Am, but I assure you it was there.

Metro Boy arrives!

I can’t speak for bigger systems, but ventilation is a significant issue in the tunnels in the Montreal metro, for the following reasons: 1) the piston effect (what you describe), 2) fire safety, 3) passenger comfort (too hot/too cold). These problems are made more severe by the fact that the entire system is underground, with no open-air sections providing constant large-scale air circulation via their portals.

In the case of the piston effect, the problem is not only a loss of operating efficiency, but also air pressure changes that can make it difficult or even dangerous to enter or exit certain stations. (In the case of one especially vexed station, a correspondent reports having seen a little old lady blown out of a station door into the air like a champagne cork, landing a few metres away. She was undamaged, aside from her dignity.)

There are two types of vents used: forced and natural. Forced air vents involve the use of large blowers. (These are automatically computer-controlled to also provide protection in case of fire or similar incident, by drawing smoke away from evacuating passengers.) In the case of the very longest stretches, such as the more than two kilometres underneath the Saint Lawrence River between Berri-UQAM and Jean-Drapeau station, such powerful blowers are used that they could evacuate all the air from a bungalow in ninety seconds.

Natural vents are simply shafts linking the tunnel to the open air. These are of course much simpler to operate than blowers, make less noise for people living above, and can be built into the architectural design of stations; however, they must be fully or partially closed in the wintertime (although frankly it could stand to be a little cooler in the metro in the winter when everyone is bundled up anyway).

In any case, at stations where there is insufficient ventilation in the tunnels nearby, the piston effect is especially noticeable, much to the irritation of all who use them; especially if the station is built in tunnel with no large volume to mitigate the effects.

There’s also the German ICE 3 (Inter-City Express, phase 3). I took the Cologne-to-Frankfurt line a few years ago, and there were many tunnels, but all very short. The terrain was rolling hills; I think the point of the tunnels was to just go through the very tops of the hills and avoid the lightweight, roller-coaster feeling you’d have by going over the crests at high speed. We were running late, so the driver kept the speed pegged at 296 km/h, and I was in the front car so I could look through the partition and out the windshield. That was one hell of a ride.

The OP is describing a system of pressurizing air and channeling it through a duct from one end of the train to the other. But that’s what’s happening already. A train going through a tunnel creates high-pressure in front of it, and the space between the train and the walls of the tunnel carries air from the front to the back (relatively speaking). And I would think you’d get better efficiency by putting the power to use in the traction motors than in spinning a fan to blow air.

Look at it this way. Instead of a duct through the train, make your train with a smaller cross-section. Instead of a fan, use that power to go faster which would force more air into the space around the train.

I could see this turning into a plane-on-a-treadmill thing pretty easily.

The new Minneapolis light-rail system has a tunnel going under the airport.

They had a problem with this, in that during winter the trains pulled cold air with them into the tunnel, which then caused ice to form on the tracks. The designers had not planned for this, because the tunnels were far enough below the frost line that they thought ice would not be a problem. They failed to consider the effect of pulling Minnesota winter air into the tunnel.

Local rail experience is that most tunnels are dual track anyway, so the problem is reduced, and that the often single-track Sydney underground was scoped out in the 1910s as being either a high level (cut and cover method) system, or a low level (tube) one, and these being the pre-aircon days, the designers decided on a high level system, partly because Sydney’s summer climate made the piston effect necessary to ventilate the stations. And it works reasonably well to this day.
Of course, there are some old single track tunnels in the rural areas, but as they were designed in the steam days, energy efficiency wasn’t nearly as important as simply staying alive. There have been deaths of steam locomotive crew by suffocation when a train has stalled in the tunnel (particularly on the second or third loco in multiple unit trains). These “rathole” tunnels often had the crew getting the train through by jamming the controls open, and lying on the footplate with rags covering their mouths and noses.

In most modern underground rail situations, it’s either more efficient to have the train itself do the job of moving the air, or it’s just not a consideration.

Of course, for the opposite of all this, where the air drives the train, have a look at the Beach Pneumatic Tube in NYC.

I took the Shinkansen from Tokyo to Kyoto last year. There were a handful of short tunnels that we blasted through, and you could definitely feel the piston effect (there were usually two single-track tubes side by side in a given mountain). I wondered if it would be useful to have ducting and fans that would actually get the air moving in the tunnel before the train actually got there, reducing or eliminating the piston effect. Probably a lot of effort for little (if any) gain, however.