Wednesday, July 09, 2008

Burning coal is burning oil

I found some numbers for the oil cost of burning coal.

Freight trains in the United States burn 1 gallon of diesel to move a ton of frieght 436 miles.

Average distance coal travels in US: 628 miles from mine mouth to powerplant. At $4.03/gallon, that's $5.80 for the diesel to move a ton of coal from the mine mouth to the powerplant, on average. Wyoming coal costs $9 at the mine mouth. So, electric producers pay almost as much for the diesel to move the coal as for the coal itself. Since marginal petroleum is imported, it's fair to say that coal is not entirely a domestic fuel.

The average powerplant cost for coal in the U.S. in 2006 was $34.26/ton. That's because coal mined outside of the Powder River basin in Wyoming costs a lot more to dig out -- the average mine-mouth price across the U.S. in 2006 was $25.16/ton. The difference is $9.10/ton, which is the cost of transport. The cost of diesel was a bit lower in 2006, but it looks like around half the transport cost is the diesel.

If the coal is 22 MJ/kg, and the plant is 35% efficient, then for each kWh at the powerplant you spend on average 1.8 cents for the coal. Just the fuel cost of the coal plant is more than the total operating cost of the Palo Verde nuclear powerplant, per kWh. This result is entirely independent of subsidies or clean coal. The black stuff is apparently just really expensive.

A while back, I snarkily suggested that mine mouth coal powerplants were a way to keep the pollution away from rich people. Looks like I was wrong:
  • Transporting a kWh of electricity 1000 miles increases the cost by 19%.
  • Transporting the coal necessary to make that electricity 1000 miles costs $14.49/ton, assuming cost is linear with distance. That's a 58% increase in the cost of the fuel. Assuming the fuel cost is 70% of the cost of producing electricity, that's a 40% increase in the cost of the electricity.
  • 4000 miles (across the continent) by electricity: increase cost by 107%.
  • 4000 miles by coal train: 160% increase.
What about the extra carbon? Transporting 1000 miles as electricity means you must make an extra 8.7% more electricity which gets lost in the wires, which produces 8.7% more CO2. Transporting 1000 miles by coal train burns 6.3 kg of carbon in the diesel to deliver perhaps 800 kg of carbon, which increases the total carbon released by 0.8%. Clearly the diesel locomotive is the lower carbon, if much more expensive, alternative.

Average distance coal travels in China: 230 miles. They're burning a lot less diesel to take advantage of their domestic coal.


  1. I did not realize that such large amounts of oil were used in coal generation. Very alarming actually.

    The Chinese often like to build coal plants near cooling waters and coal mines rather than population centers, and then transport it across the country via transmission lines. Some of the are really long, 1000 miles or bigger. This makes sense for them already, and will make more sense with higher oil prices which are very likely with Peak Oil around the corner.

    I would like to point out that many modern supercritical coal plants are 45 percent net efficient or sometimes better, and it's still improving slightly.

    What about underground gasification burned in combined cycles? Latest tech gets 20-25 percent losses in gasification and more than 50% combined cycle efficiency.

  2. Great article, thanks. TP

  3. Coal here in australia is much more expensive - high quality coking coal is selling over $200 tonne I believe

  4. pnm is into palo verde nuclear, san juan and four corners coal fired electric generation plants.

    interesting computations you make about the amount of diesel required for coal fired plants.

    mercury contamination from san juan and four corners have badly polluted new mexico lakes.

  5. Maybe the Chinese are building their plants closer to the mines because their coal plants are newer, and so were planned based on higher voltage transmission.

  6. What happens if you replace the power transmission lines with DC instead of AC. My understanding from what I read is that they do not lose any power over huge distances.
    If so, please correct me if I'm wrong, then the east and west coasts could share powerstations and therefore even out the spikes in demand experienced by the power plants.
    If the us is paying £700B/yr to foreign powers, then it doesn't take much to consider spending some serious money upgrading the power transmission network.
    Just needs planing, foresight, political will ... oh bum;-))

  7. Andrew,

    DC suffers resistive losses just like AC. There is no magic.

    HVDC is typically configured as two lines, one at e.g. +500 kV, and one at -500 kV. If each wire carries 1000 amps, you're moving 500 MW per wire -- all the time.

    HVAC is typically configured as three lines, each oscillating between e.g. +500 kV and -500 kV. When any one line is at +500 kV or -500 kV, and carrying 1000 amps, then it's moving 500 MW at that instant. Current oscillates just as voltage does, and so if you take the average of current*voltage, the average power moved, per wire in this case, would be 354 MW.

    The difference between HVAC and HVDC is that the same amount of metal carrying HVDC can move a bit more power, on average, than if carrying HVAC. It's not a huge factor, but when you are talking about long distance transmission, even small factors are worth a lot of money.

    Getting the rights to string electric wires across the countryside is a matter of political will. But power transmission is physics. The existing grid was well designed, but not for a world of expensive oil. There will be no cheap and easy way to make it better.