Is Amtrak greener than flying?

[joelkfla]

NY Times writer questions whether the train is really cleaner than flying cross-country:

Trains Are Cleaner Than Planes, Right?

 

[bonzoesc]

Add to that generous seat pitches, large old-fashioned private rooms for longer-distance trains, a longer, winding route across the country and “per-passenger-mile emissions go through the roof,” said Justin Roczniak, a co-host of “Well There’s Your Problem,” a podcast about engineering.

I love this podcast, really hoping I can make it to a live show some time. Towards the end of the episode about battery-electric trains, Rocz really goes into the physics advantage that overhead electrification offers: you don't have to bring all your stored energy with you!

It’s when journeys start getting longer than about 700 miles that planes start to gain an advantage on trains. Planes burn the most fuel when they take off and climb to altitude. That makes short flights very inefficient — you’re burning all that fuel only to travel a short distance. (Some countries, like France and Spain, have tried to ban the shortest flights when rail alternatives are available.)

Longer flights also tend to use larger aircraft, which provide economies of scale. And aircraft have become more fuel-efficient over the years. But choosing flights with several connections, for example, can quickly add to your footprint, because you’re taking off and landing multiple times.

This is where the big difference is. The author's trip from NY to Stanford (SF Bay Area) is serviced by many non-stop flights daily, from 737s up to 777s (United in particular has four 777s from EWR to SFO tomorrow, April 5). The 777s may or may not be particularly light on fuel (the airline may want to spend more flying it from Newark to SFO so they can buy as little fuel as possible at California prices), but a full plane that size doesn't burn much fuel per seat.

However, things change if you're one of the majority of Americans that don't get to fly non-stops, either for price optimization reasons (the cheapest trips NYC-SFO on April 27 involve connections in Boston, Los Angeles, Dallas, or Ft. Lauderdale) or simply because there's no market for non-stops from Jacksonville to SF and instead have to make multiple super-thirsty takeoffs on smaller, thirstier-per-seat planes.

 

[Willbridge]

It would be more realistic to compare long-distance sleeper travel with motor homes.

 

[MARC Rider]

The journalist based his estimates on a bunch of computer models that aren't actually backed up by actual fuel consumption test data and that don't account for variability from trip to trip. Probably the most robust way to compare emissions by mode would be to take the total passenger miles of the carrier and divide it by the total fuel consumption of the carrier. For Amtrak, you would have to disentangle the diesel-operated sections from the electrified sections, and then figure out emissions of the electric generation, though I think that the southern NEC, the electricity is generated by hydropower. This would include switchers and maintenance trains, too. Not sure how you would allocate emissions from maintenance trains on host railroads. For airlines, you'd need to include the fuel consumed by tugs and baggage carts. I'm not even sure that one could collect data that measures actual lifecycle emissions. All of that is done my modeling exercises that pull averages of the items with no measure of the variability of the averages.

There's very little published data showing actual fuel economy test results of trains with diesel engines, and I think that some of that might be a bit out of dates, as diesel technology has changed greatly over the past 10-15 years. The only "on-road" test data I've ever seen is what I posted here a few years ago:

https://www.amtraktrains.com/threads/how-green-are-trains-really-with-data.75161/

They were testing Piedmont trains in actual revenue service, and it seems they did 0.6 - 2 miles per gallon (1.6 to 0.5 gallons per mile), depending on which part of the route they were running. (A shorter section with lots of curves had fuel economy on the lower end.) These trains consisted of about 4 coaches and a cafe car with a locomotive at each end. The coaches hold 56 - 66 passengers each, so if the train were full, it would have about 225 passengers. At 1 mpg, that would mean that the train's performance would be 225 passenger miles per gallon. Unfortunately, I don't think the research I cited gave information of the passenger loads at the time of testing. More passengers, of course, add to the mass being pulled, which would decrease fuel economy, but given the low rolling resistance of steel wheels on rail, I wouldn't think the increase would be larger than the expected test to test variability, but I with no actual test data, I'm not sure.

A car getting 30 miles per gallon with 4 passengers would be getting 120 passenger miles per gallon. A highway coach that gets 5 miles per gallon with 50 passengers gets 250 passenger miles per gallon. What that suggests is that for short hauls and higher passenger loads, trains are clearly superior, even with older locomotives.

I haven't seen any data on airliners, and as has been pointed out, most of the fuel is consumed during takeoff and landing.

Anyway, getting an accurate comparison of the environmental impact of transportation modes requires a lot more data (not all of which is readily available) than what's available to a journalist working under a deadline.

 

[bonzoesc]

More passengers, of course, add to the mass being pulled, which would decrease fuel economy, but given the low rolling resistance of steel wheels on rail, I wouldn't think the increase would be larger than the expected test to test variability, but I with no actual test data, I'm not sure.

Wikipedia lists a Superliner coach as 75 tons and seating 78, and a sleeper as 80 tons sleeping 44. If everyone overpacks way worse than I do for a total weight of 400lb/pax, that’s about 16 tons of payload for a coach and 9 tons on a sleeper. More mass is gonna need more power to get going, but once you’re in motion it probably isn’t deforming the wheels enough to introduce more drag. It’s definitely not as significant as the rocket equation madness (where as your payload goes up your initial fuel burn needed to get it moving grows exponentially) involved in getting a payload off the ground in an airplane.

 

[jis]

The other factor to consider is that the airliners spew their bad stuff right in the layer of atmosphere where it can do most immediate damage. So all pollution even of the same sort is not necessarily the same in its final effect. Fuel burn is just a proxy and a rather rough one at that when it comes to determining actual greenness.

 

[GDRRiley]

I'll also note Amtrak is still mostly T0+ with the P42 but that will be T4 for most units in a few more years.

 

[dlagrua]

If the truth be known, train travel is FAR less polluting than air travel. Its not about miles per gallon it about how much fuel is used per passenger per mile. For instance if you have 300 people on the train and fuel consumption is only one mile per gallon, the fuel used to transport that passenger is 300 miles per gallon. Want to travel 3000 miles then only 10 gallons of fuel will be used. I do not buy into the statements made in that article. Even if the passenger fuel consumption is 200 miles per gallon the train is less polluting.
Consider that a Boeing 737 uses 2.8 Gallons of fuel per passenger mile. Do the math.

 

[Devil's Advocate]

There was a pilot who came on the forum a while back making the claim that flying was better for the environment than taking Amtrak. While this is demonstrably untrue in an apples-to-apples sense it's possible that you can find some examples where it will be close. During the back and forth with the anti-rail pilot the more it looked at least somewhat plausible (but also cherry-picked). It was genuinely surprising to see that some flying was even in the ballpark of riding a train, but Amtrak's T0 locomotives, old indirect track routes, slow and ever-changing speeds, and stationary HEP running did seem to make it possible.