Jump to content

435 Miles On 1 Gallon Of Fuel?!

  • Please log in to reply
10 replies to this topic

#1 Vapor Trails

Vapor Trails

    In a world where I feel so small, I can't stop thinking big.

  • Awaiting Authorisation
  • 16,523 posts

Posted 20 September 2009 - 08:31 PM

So, I've seen these commercials about CSX trains traveling 435 miles on 1 gallon of fuel. How can a train get such great gas mileage, and cars get such paltry mileage-even as hybrids?? :blink: The new Chevy Volt supposedly can get 230 mpg. F**k that, get me a train. :p~
Posted Image

Politicians are like bananas; they hang together, they're all yellow, and there's not a straight one among them.

"We're relevant for $ and a vote once every two years. Beyond that, we're completely irrelevant, except of course to consume, and preach the gospel according to [insert political demigod here]."--Cait

#2 Orpheus


    Get my agent! I'm supposed to be Castathan, not Indogene

  • Administrator
  • 16,868 posts

Posted 20 September 2009 - 08:47 PM

Trains have terrible acceleration and braking, and no steering. They also have their own private roadways with lower grades (inclines) than automobile roadways, essentially *never* hit a stoplight, and don't have to maneuver into parking spaces. You too, might get great mileage, if your car was designed to drive in a straight line for hundreds of miles without stopping -- but I suspect you'd miss the ability to stop at say, your house, office, store, etc. Moreover, most of the train is ... well, the bicycling term is "training" .. aerodynamically on the car ahead of it. To get the same benefit, you'd have to tailgate in one lane, start to stop, as the car ahead of you

Subway trains, which have better maintained, gentler tracks but must accelerate more rapidly and brake to a stop on the order of every km (give or take) don't get nearly the efficiency of a freight train. Even so, they are so underpowered that the MBTA Red Line (to name one) can have difficulty climbing the gentle incline of the Charles Street bridge in icy conditions.

It's also worth noting that they say CSX can move a ton of goods up to 423 miles on a gallon of fuel (Sept 2008 commercial), but that's a cruising average. Considering how many tons of good are in a train, and that accelerating (especially from a standing start) consumes many times the fuel of cruising the same distance, a train probably consumes several hundred gallons, just maneuvering out of the train yard at 3 mph, if the switches aren't set up properly for a clean run. The equivalent 240 trucks would burn much less over the same distance and speed.

Edited by Orpheus, 20 September 2009 - 08:59 PM.

#3 Nick



  • Islander
  • 7,121 posts

Posted 20 September 2009 - 09:03 PM

They're doing a bit of math-fu.  They aren't claiming to have trains that get 435 miles per gallon.  Rather, they're claiming to be able to haul one ton of freight 400+ miles per gallon under ideal conditions.  A train as a whole has mileage on the order of a few tens of feet per gallon.

Trains are the most energy efficient and economical way to transport a lot of stuff to and from the same place.  Of course, once you stop talking about a fully loaded freight train, the economics drop off very very quickly.

And the Chevy volt's mileage isn't really 200+ mpg.  On gas alone, it's probably more like 50 mpg.  However, when you assume a typical driving patter of under 30 miles per day, and the car is recharged overnight it's running mostly on stored electricity and treating that electricity as "free."  By that logic, an all-electric car gets unlimited miles per gallon, since it doesn't burn any gas at all.

FWIW, with current technology and typical utility rates, accounting for charging inefficiencies and losses--from the wall socket to rubber-on-the-road, a "gallon" of electricity costs about $1.00.

Orpheus said:

It's also worth noting that they say CSX can move a ton of goods up to 423 miles on a gallon of fuel (Sept 2008 commercial), but that's a cruising average.

Okay, that's almost comical.  So they're essentially just claiming to be able to keep a ton of cargo moving at that mileage.  I wonder how that compares to a truck on a highway's cruising average per ton of cargo.  Of course the train will probably do better, economies of scale and all.  I just wonder by how much.

Edited to add: I found a thread in an Amtrak forum where one poster got a written response from CSX about how they arrived at that figure:


Holy moly, I actually received a snail mail response from CSX:

Thank you for your inquiry about our advertising.

Our ad states "trains can move a ton of freight 423 miles on a single gallon of fuel," and it is correct. The 423 is calculated by dividing the total tons hauled by a single freight train by the total gallons of fuel used. So, on a per mile, per ton basis, the amount of fuel needed is extremely low. In fact, it is so low that trains are between three and four times more fuel efficient than trucks.

So, even though it takes more than one gallon of fuel to power a train for 423 miles, the train actually uses less fuel than the more than 280 trucks it would take to haul the same amount of freight.

Thank you again for contacting us.


TellCSX Team

So they aren't being as deceptive as I thought--they're not just talking about a cruising average, but a single start-to-finish haul they've actually acheived.  The acceleration costs are in there, and just averaged out over the whole trip.

Edited by Nick, 20 September 2009 - 09:25 PM.

#4 Nick



  • Islander
  • 7,121 posts

Posted 20 September 2009 - 09:29 PM

And a few more comparisons from a post in the thread I linked to above:


A Ford Excursion, getting 10 mpg, when loaded to its capacity of eight passengers, gets 80 passenger miles per gallon. Ton/miles per gallon is not a measure of consumption, but of efficiency.

I need to move 400 tons of (800,000 lbs) of stuff 423 miles. A a truck hauling 80,000 lbs gets about 4 miles to the gallon, IIRC last time I was looking at peterbilts. So to move that 400 tons, I need 10 trucks. So I am going to move these items .4 miles for every gallon consumed, so to travel the distance of 423 miles, I need to us 1057 gallons of diesel fuel. Alternatively, I can tack it onto the back of a freight train (CSX isn't all that much more efficient than its competitors, btw), and it will take 400 gallons. So this 18-wheeler moves one ton of freight 162 miles.

It ain't a ridiculous increase in efficiency, you understand. About 2.5 times. But its a big solid demonstration of the advantage of moving things in bulk.

To further the comparison, let us move these 400 tons of stuff via a Ford F350 with single rear wheels, which can carry 4000 lbs, or 2 tons worth of stuff. It gets around 10mpg fully loaded. To carry 400 tons, you need 200 F350s. It will take 42.3 gallons of fuel per truck to do the move, or a total of 8460 gallons of fuel. This F350 moves 1 ton of frieght 20 miles per gallon of fuel.

Its the best way to measure haulage efficiency of bulk freight. And in this regard, the railroad is clearly superior.

#5 Orpheus


    Get my agent! I'm supposed to be Castathan, not Indogene

  • Administrator
  • 16,868 posts

Posted 20 September 2009 - 11:58 PM

It looks like I owe them a bit of an apology -- they were indeed fairer than I thought.

In their 4th Quarter 2007 report, CSX says they moved 253 billion revenue ton-miles of goods in 2007, and consumed 569 million gallons of diesel #2 fuel. That's 444.6 ton-miles/gallon.

I'd have been impressed it it was just the cruise efficiency of their best trains, but you're right, Nick: it looks like it's total average efficiency for their entire operational fleet!

However, for the sake of completeness, there's a little ambiguity in what it means for a train company to move X tons Y miles. On most transportation networks, loads are shared where possible. If you fly from one city to another, you may be handed off to another airline for part of your trip, either by plan or due to unexpected circumstances. Trucks don't do this too often because a "load" is usually a pretty discrete unit withthe same origina and destination, so swapping one trucking company for another just means swapping the rig/driver, not the trailer -- and that rarely carries benefits over having the original rig/driver complete the trip (as opposed to a train, cargo freighter or airline, where each rail car, shipping container or passenger could easily have a different origin, destination and commitment schedule). In other words, for *some* of those ton-miles, they probably don't burn any fuel at all. Instead they burn money, by paying another rail company to do part of the route (it doesn't make sense to send a train to Sault Ste. Marie, if you only have five cars  headed that way; it'd be smarter to pay another rail company to tack those cars onto their rain. I've certainly often seen trains that had a few "mismatched" cars in the livery of another rail company.

You are still the carrier of record for the cars you hand off, still responsible to the customer--and they're probably *your cars.  Naturally, it's better to deliver as much of your own cargo as practical, to keep all the profit from your contract, than to pay another railco part of your profits to do one leg. Since the 423 ton-miles/gal they claim is within 5% of the 444 tin-miles per gallon from their report, I'd guess they manage to keep it within 5%. In fact, that's a lot more efficient scheduling/logistics than I remember from my days when I noticed such things -- trains in the 1970s were a LOT more mismatched, but in the 1980s they seems much closer to "unicultures". Maybe they used to rely more on a mile/ton swapping arrangement, and settle among each other in cash periodically That would have made it apparently cheaper to swap cars at will -- until efficiency became paramount, and hauling your competitor's crappy old cars started costing you appreciably more than hauling your own spiffy new cars. [Railcos that were hard pressed for resources and railstock would likely swap cargo off to others more than railcos with ample assets]

I don't know as much about [largely domestic] rail as I do about international marine freight, but it occurs to me that there are some substantial differences in the underlying technology that I, as a consumer, never considered.

#6 Woodmansee

  • Demigod
  • 129 posts

Posted 21 September 2009 - 12:06 PM

Thanks for the investigation. I was similarly intrigued by the advertising claim.

It would be interesting to get the data and do the same sort of mean cargo ton/mile gallon math for a container ship across the ocean, and for a long range air flight (UPS or FEDEX cross country cargo flight).

Anyone up for this research challenge?


#7 Orpheus


    Get my agent! I'm supposed to be Castathan, not Indogene

  • Administrator
  • 16,868 posts

Posted 21 September 2009 - 10:20 PM

Well, it's not as simple as with trains. Weather, routes, currents and most particularly the size and technology of the ship all play a big role. I decided to go with "best case" fleet operating numbers, and chose the Maersk Line (just one of many shipping companies owned by Maersk) as my basis.

Note: "TEU" means "Twenty-foot Equivalent Unit": an intermodal shipping container that is 20' x 8'x8'6". "FEU" means "Forty-foot Equivalent Unit", but these terms are approximate: a 45' High Cube (45'x8x9.5') is still counted as an FEU

By itself, the Maersk Line is the largest container shipping company in the world, with over 450 container ships totaling 1.9 TEU (avg: 4200+ TEU/ship) and including the 8 largest container ships in the world (Emma Maersk-class; 11,000 TEU @ a standard 14 tons/TEU; 15,200 TEU by full carrying capacity) as well as 27 of the top 51 largest container ships (this figure may be outdated, they took delivery on some more this year--with some reticence). Big ships are more efficient, so other companies would likely be less efficient.

The Maersk Line website offers the following chart comparing fuel usage:

The units are unfamiliar, but are easily converted, using this conversion table, and the value they give for diesel trains agrees precisely with the value I calculated for the CSX fleet (440 ton-mi/gal) to the decimal places given. [As always: check my math]

.067 kWh/tonne-km = 14.9 tonne-km/kWh = 9.27 tonne-mi/ kWh = 10.2 ton-mi/kWh (tonne = metric ton = 1.1 US tons)
9.27 ton-mi/kWh / (3410 BTU/kWh) = 0.00299 ton-mi/BTU

0.00299 ton-mi/BTU * (1,000,000 BTU/7.19 gal #2 Fuel oil) = 416 ton-mi/gal of #2 fuel oil (aka standard diesel)

0.00299 ton-mi/BTU * (1,000,000 BTU/6.67 gal #6 Fuel oil) = 448 ton-mi/gal of #6 fuel oil (aka bunker C fuel)

I wasn't sure if locomotives use standard diesel fuel or bunker oil, but I know that the primary diesels on big ships use bunker oil, so I'm assuming diesel locomotives do, too. They are similar sizes and have similar corporate imperatives. Bunker oil is by far the largest fuel oil market globally, Grades #5/#6 are commonly called FFO (Furnace Fuel Oil) HFO (Heavy Fuel Oil) or "residual fuel oil" IFO-180/IFO-380 (for a viscosity of 180/380 centipoise, respectively)

[Well, so the diesels on *normal* big ship are comparable to a locomotive. Each individual engine on an Emma Maersk-class is five decks tall, and makes a locomotive look like a parts trolley . This video, and the many similar videos on that YouTube page will tell you more than you want to know. Just be careful with the multi-parts: a few different video series are posted under similar names]

Given the agreement between Maersk's train figure and my own direct CSX calculation, let's work from there:

Maersk S-type (6600 TEU)  0.067/0.018 = 372% of a diesel locomotive's ton-mi/gal (26.9% of its fuel consumption)
Maersk PS-type (11,000 TEU)  0.067/0.014 = 479% of a diesel locomotive's ton-mi/gal (20.9% of its fuel consumption)

CONCLUSION: Assuming locomotives use bunker oil C (#6 grade fuel oil) and get 448 ton-mi/gal
Maersk S-type = 1666 ton-mi/gal
Maersk PS-type = 2146 ton-mi/gal

Now for the sanity check:
The Maersk Line website currently offers to ship 1 TEU Qingdao/Xingang → Bremerhaven for $625. Assuming they use a PS-class for this highly-specific high-traffic run, and with bunker C (IFO 180-380) spot prices running $440/tonne in East Asia, where they would be fueling up:

$440/tonne (spot price, Bunker C) * .991  tonne/m  (specific gravity of Bunker C) = $436/m =  $0.436/L
$0.436/L * 3.79 L/gal = $1.65/gal

14 tons / 2146 ton-mi/gal = .00652 gal/mi = 6.52 gal/1000mi
(Actually, 14 ton/TEU is an international standard, so it's probably metric tons -- make that 7.18 gal/1000 mi)
7.18 gal/1000mi * $1.65/gal = $11.85/1000 mi

($625/TEU)  / ($11.85/1000mi) = 52,750 mi
However, fuel is generally said to be 40% of the operating cost of a container ship run (sorry, no cite, but I read this in many places while looking up these figures. On shipping in general (smaller, non-containerized) it's 50-60%, but large container ships burn less fuel, driving the ratio down)

(0.40 * $625/TEU)  / ($11.85/1000mi) = 21,100 mi

Given that his is the public-offering single-TEU price (the site lets you counterbid for a cheaper price, especially in quantity) and that the return leg, Bremerhaven → Qingdao/Xingang, is likely to be mostly empty (The Chinese sell much more to the West than they buy) and the round trip is likely to be in the 16-20K mile range, I'd say that this passes the sanity check for the range of mileage (fuel consumption) that would enable them to make a reasonable profit with this price offer.

Certainly, I'd say that it would be impossible to ship a TEU Qingdao/Xingang → Bremerhaven if the container ship burned 4.79 times as much fuel as we calculated here (i.e. 448 ton-mi/gal, as a rail car does by these numbers)

$625 / ($11.85*4.79)/1000mi = 11,000 mi -- leaving nothing for other expenses (or profit)

I hope these approximations meet with your satisfaction, Paul. Sorry I couldn't find anything more direct (it's probably a competitive secret), but I know you're probably even mathier than I am, and will enjoy the read. You do the same type of calculations, professionally, after all ... just on an interplanetary scale.

-- Orpheus "Trucker to the stars? Where do I sign up?"

#8 Orpheus


    Get my agent! I'm supposed to be Castathan, not Indogene

  • Administrator
  • 16,868 posts

Posted 22 September 2009 - 01:45 AM

I found this page, which has some interesting information on residual fuels. It's a little dated but nonetheless notes an uptick in the return to residual fuels in large transport. However while an Emma Maersk-class container ship can output 109000 horsepower (82 MW) and burns  fuel much like a contemporary fixed-based, 100 MW capacity, city powerplant, and  the largest diesel locomotives ever built were the 6600 hp (5MW) Centennials in 1969-1971, comparable to many smaller city powerplants of that era, the trend in marine container ships is toward larger powerplants (better able to handle heavier, slower burning, fuels) while the Centennials were all but retired by the mid 1980s (only one remains operational).

It's possible that railroads may have converted smaller locomotives to handle heavier fuels: heavy fuels are cheaper, which would impact CSX's balance sheet, not the MPG or ton-mi/gal that I was asked about. I alluded that ton-mi/gal comparisons might be comparing apples to oranges [or #2 diesel to #6 bunker oil) to a certain degree, which is why I used a ratio comparison of the kWh cited in the Maersk table. kW are a unit of power [energy per unit time], which is a more meaningful number. Gallons are a unit of volume -- and, really, who cares what *volume* of fuel you burn? That's like saying ketoprofen must be a better, more powerful drug than aspirin or ibuprofen, because has smaller pills

Hidden: you don't really want to read me ranting on over-the-counter analgesics, do you? (click to hide/unhide)
Believe it or not, there was was a big advertising war on this very point in the late 1980s, which I couldn't fathom, since a bigger difference between 325mg aspirin, 200 mg ibuprofen, 100 mg naproxen and 12.5 mg ketoprofen tablets was how much substituents ("extra stuff") a manufacturer added to help a pill stick together in the bottle, but fall apart/dissolve in your stomach -- but the truly relevant difference is the relative painkilling ability of those various doses ("therapeutic effect" usually matters a lot, "potency' almost never does). Pharmcos didn't seem to worry about thereapeutic, because they couldn't market it when every painkiller claimed to have "the most". Why is a standard tablet of OTC (over-the-counter) acetominophen (e.g. Tylenol/Panadol) exactly 325mg? Clinical studies? No. It's because the standard tablet of aspirin had been set at 325mg in the late 1800s, and pharmcos thought buyers might compare mg like ounces in a packet of lunch meat,

Acetominophen is just a safer variant of phenacetin, which wasn't dosed in 325 mg increments--nor was acetominophen, when it was a prescription drug. In crude terms, acetominophen slowly breaks down to phenacetin in your body, so you get phenacetin benefits but with lower/safer serum levels than you'd endure if you took the phenacetin "straight".

In the late 1980s, some ketoprofen ad guru decided to turn the smaller number of mg in a safe OTC dose of ketoprofen from a marketing drawback into a selling point: Orudis KT's  "smaller, therefore more potent" pitch filled the airwaves for a few years, but it isn't a player the US OTC market today. Patients quit buying it after trying it and finding it more irritating to the stomach, more prone to cause bleeding, much more expensive -- but no better at reducing most pain than the cheaper/safer ibuprofen. It's sometimes better for certain kinds of pain, but not for general analgesia.

Heavier, thicker fuel oils are also more polluting [more smoke particulates, carbon dioxide, nitrogen oxides, and often more sulfur],, and are suited for use in ships on the open ocean than trains/cars running through urban centers and suburbs.

For a straight-up measure of energy efficiency, just take the inverse of the figures given in the Maersk table: you want ton-km/kWh (efficiency) not kWh/ton-kn (consumption). However, that wasn't what I was asked, so I didn't address it (except in passing). I'll say it now: for a straightforward comparison of energy efficency --

Diesel = 1/ (.067 kWh/tonne-km) = 14.9 tonne-km/kWH
S-class = 1/(.018 kWh/tonne-km) = 55.6 tonne-km/kWH
PS-class = 1/(.014 kWh/tonne-km) = 71.4 tonne-km/kWH

Though those may be the fairest figures, I didn't think those units would have any visceral meaning to most EtUers.

Also, to better answer GR's original question --"why can't cars match railroads?"-- Google books has "The Elements of Railroad Engineering, which has all sorts of specific numbers that I didn't expect to be able to find anywhere.

#9 Nick



  • Islander
  • 7,121 posts

Posted 22 September 2009 - 07:52 AM

View PostOrpheus, on Sep 22 2009, 02:45 AM, said:

Also, to better answer GR's original question --"why can't cars match railroads?"-- Google books has "The Elements of Railroad Engineering, which has all sorts of specific numbers that I didn't expect to be able to find anywhere.

Correct me if I'm wrong, but isn't it mostly due to friction/drag, with the rest being made up by the economy of scale in a larger engine?  Cars tend to travel at higher speeds, meaning even more friction/drag, and while trains take enormously more energy to START moving, their energy requirements to overcome friction and drag once at cruising speed is considerably less than a passenger car faces (per unit of mass).

#10 Orpheus


    Get my agent! I'm supposed to be Castathan, not Indogene

  • Administrator
  • 16,868 posts

Posted 23 September 2009 - 10:36 AM

Well, you're basically right. What I found interesting was the wealth of numerical detail (even if it may be a bit outdated. For example, the rolling resistance of a freight car reaches its minimum value at about 6-8 mph, and can be as low as 2 lbs/ton of gross weight in real life on an ideally lubed railcar in the summer. I'd forgotten that "back in the day" crews of up to 4 yardmen would move the fully laden cars around the yard manually -- and sometimes only two were actually pushing!

It's apparently harder to move most 1-ton automobiles around manually than a heavier empty railcar!

One advantage railcars have is that they can have parallel perpendicular wheels on simple axles. Since automobiles must maneuver large amounts over a wide range of speeds, their wheels are never really parallel: they have camber and are "toed in" at a small angle (and those angles are optimized for higher speeds, where the energy/force demands are greater; at low speeds the engine has ample torque to overcome a suboptimal angle) Most of us never think about what goes into the routine wheel alignments that we have done on our cars, but it's those small angles (important to fuel economy in roadway driving) that keep it from being a simple home adjustment with a laser pointer and a wrench.

#11 Zanthra

  • Islander
  • 1,841 posts

Posted 23 September 2009 - 08:24 PM

There is also the idea that the diesel generators of Diesel-Electric trains are years ahead of automotive hybrid technology.  In fact mention was made of the Chevy Volt, which happens to be the first I have heard of a Gasoline or Diesel fueled car with a solely electric drivetrain.  This is a very good thing because unlike electric motors, Internal Combustion Engines vary widely in how efficient they are at different RPMs, and disconnecting it from the drivetrain lets you run it at optimal speed any time it is running.

Also it would likely be even more efficient to run freight on electric rails as they often do in england, this lets the large efficient steam turbine generators at power plants generate the electricity, and eliminates the need to transport the fuel to the train, and the need to carry it, the engine, the generator, and the batteries with the train.

Edited by Zanthra, 23 September 2009 - 08:29 PM.

"Making something tedious will not stop players doing it if it's very clearly the best option. They'll do it, and they'll hate it." --CCP Games speaking of their Lessons learned in making EVE Online

0 user(s) are browsing this forum

0 members, 0 guests, 0 anonymous users