Comparative Energy Costs of Driving: Cents per Mile

[For an update to late 2013, with important new analysis from the consumer’s perspective, click here.]

[To jump to the table, click here. To jump to the conclusions, click here. For comment on the President’s interview on Iran, click here.]

Our most crying need for better energy policy is in transportation. That’s where we spend $372 billion per year on foreign oil at present prices and importation rates.

That waste will only get worse. As gasoline and oil prices go up—and they will (1 and 2)—that expense will increase. Every year, we will waste at least one-third of a trillion dollars (1) increasing our trade deficit, (2) enriching others’ economies, rather than our own, and (3) supporting regimes that we would rather not support, like Saudi Arabia, Iran, and Venezuela.

In all the heated discussions over how to bring that awful number down, I have never seen a comparison of greatest interest to every driver: the cost of driving per mile. So I’ve compiled the following table, using simple arithmetic and publicly available information (linked in or through this post).

The table shows the cost of fuel or electrical power, per mile driven, for various automotive energy sources. It focuses on the fuel or power source alone, ignoring the capital cost of the vehicle, depreciation, and other practically important expenses, such as insurance and maintenance. Its numbers are therefore much lower, for example, than the “mileage rates” at which the federal and state governments quite properly reimburse their employees for each mile of official travel.

The table shows energy cost alone. It lists energy sources in order of decreasing per-mile cost. Its sole purpose is to let readers judge for themselves the validity of competing claims about the relative “economy” of various sources of automotive energy. Notes following the table explain how each number was calculated.

Energy Cost of Driving, in Cents per Mile,
for Various Automotive Energy Sources

Energy Source	Underlying Price Parameter	Cents per Mile Driven
Gasoline	$3.78 per gallon	12.6
Natural Gas (Residential)	$1.28 per gallon equivalent	4.3
Conventional Electricity (Residential)	11.6 cents per kWh	4
Conventional Electricity (Industrial)	6.8 cents per kWh	2.3
Natural Gas (Industrial)	$0.55 per gallon equivalent	1.8
Solar Photovoltaic Electricity	5.1 cents per kWh	1.8
Nuclear Electricity	4.4 cents per kWh	1.5

The most important finding from this table relates to gasoline. You could increase to 80 MPG the mileage (30 MPG) used to calculate the per-mile energy cost of gasoline and still not match any other per-mile cost in the table. And the price of gasoline is only going to rise, as hundreds of millions of consumers in the developing world enter the middle class and cause massive increases in global demand for oil. Gasoline is obsolete.

Nuclear electricity, solar photovoltaic electricity, and natural gas at industrial prices all offer the lowest energy costs per mile. But the results in the table for the first two are, at best, accurate only within about 30%, and natural-gas prices are subject to much wider change. Therefore the table cannot distinguish among these three energy sources. Insofar as concerns cost, they should be seen as roughly equivalent. Conventional electricity (largely from coal) and natural gas at residential prices would roughly double their cost per mile.

Another implication of this table is that our system for distributing natural gas to residences is much more costly and inefficient than our similar system for electricity. Consider the ratio of energy cost per mile for residential service to that for industrial service. For electricity, that ratio is 4/2.3 = 1.74. For natural gas, it’s 4.3/1.8 = 2.39, 37% higher. As cars and light trucks switch to natural gas for cheaper fuel, one goal of energy policy might be to reduce this discrepancy.

On the other hand, relatively high residential gas prices might encourage the development of independent natural-gas service stations. With current natural-gas distribution schemes, consumers could nearly double their savings over gasoline by buying compressed natural gas from a natural-gas station, which presumably would pay industrial prices, rather than by installing a compressor in their homes and paying residential retail prices. Consumers would have to pay the natural gas station’s amortized operating cost and a reasonable profit, but together those expenses should not increase the per-mile cost of energy by more than about 20%.

The other important finding from this table is that all alternative forms of energy would lower gasoline’s current per-mile energy cost by nearly a factor of three. Natural gas at residential prices would lower it by more than factor of four, after recovering the cost of a home compressor. The lowest-cost alternatives—solar photovoltaic electricity, nuclear electricity and natural gas at industrial prices—would reduce the per-mile energy cost of driving to between one-seventh and one-eighth that with gasoline.

An upcoming essay will discuss the more general implications of these conclusions for short- and medium-term energy policy.

Notes:

General: The table’s cost per mile is rounded to the nearest tenth of a cent.

I have been unable to find enough publicly-available, authoritative data on windmills or solar thermal plants to compute the relative costs of their electricity. They are omitted for that reason, and no other.

Gasoline: This cost figure begins with the US Energy Information Administration’s national weekly average price of gasoline, all grades. The number shown is for the week of February 27, 2012, namely $3.78 per gallon. The cost per mile (12.6 cents) is that number, divided by 30 MPG—a respectable but not stellar mileage.

Natural Gas (Residential): The price per mile is based on energy equivalence, using the ratio of the residential retail price of an amount of natural gas containing the same energy as a gallon of gasoline to the retail price of that gallon.

A thousand cubic feet of natural gas provides 1.027 million BTU, which is energy-equivalent to 8.27 gallons of gasoline. So to calculate the energy-equivalent retail price of natural gas, we just divide its price per thousand cubic feet by 8.27.

As of November 2011 (the latest data available) the all-US average residential retail price of natural gas was $10.59 per thousand cubic feet. So, as of November 2011, at residential retail prices, an amount of natural gas providing the same energy as a gallon of gasoline cost $10.59/8.27 = $1.28.

The cost of that natural gas per mile is therefore the just cost for gasoline, multiplied by the price ratio for equivalent energies of natural gas and gasoline, thus: 12.6 cents per mile x (1.28/3.78) = 4.3 cents per mile.

Conventional Electricity (Residential): The price of electricity per kilowatt-hour is the EIA’s nationwide average price of residential electricity for 2010. The cost per mile is that number divided by a calculated mileage (miles per kilowatt-hour) of the Nissan Leaf.

We derive the Leaf’s electrical mileage by dividing the Leaf’s range (70 miles) for highway driving in hot summer by its battery capacity, 24 kilowatt-hours. Since the Leaf’s range for driving in hot summer is one of its lowest (its nominal range is 100 miles), this mileage factor is conservative (on the low side), so the cost per mile errs on the high side.

Conventional Electricity (Industrial): We use precisely the same method as for residential conventional electricity. But we use the EIA’s nationwide average price of industrial electricity for 2010. The resulting cost per mile is conservative on the high side, as above.

Natural Gas (Industrial): This calculation proceeds as for natural gas at residential prices. The only difference is the differing retail price of natural gas for industry. For the same month (November 2011) as in that calculation, the all-US average industrial price of natural gas was $4.53 per thousand cubic feet. So the industrial retail price of an amount of natural gas providing the same energy as a gallon of gasoline was $4.53/8.27 = $0.55. And the cost per mile for industrial natural gas was 12.6 cents per mile x $0.55/$3.78 = 1.8 cents per mile.

Solar Photovoltaic Electricity: We come to this number in the same way as for other forms of electric power. But we use the calculated cost of solar photovoltaic power (3.42 cents/kWh) taken from this post, based on the amortized capital cost of the solar plant and the time value of money for financing it. Then we add 50% for retail cost, as follows: 40% for the average cost of transmitting and distributing electric power today, and 10% for the producer’s profit. The result is the underlying retail price parameter, 5.13 cents per kWh.

This calculation omits fuel costs, as solar power needs no fuel, and external costs, as solar plants require no fuel extraction and have no effluent. It ignores maintenance costs because they are much lower than for coal plants, which generate a plurality of conventional electricity, and because there is insufficient experience with large photovoltaic solar arrays to estimate maintenance cost. Since the relevant maintenance cost is the cost per kilowatt-hour delivered, it is unlikely to change the figures shown significantly. (For more on how to calculate renewable energy costs, see this post.)

For four reasons, the table’s per-mile cost figure for solar electricity is the least certain of all and is likely to be the most conservative, i.e., unrealistically high. First, it is based on a solar-cell production cost of $1 per Watt. Yet Morningstar’s investment report of 12/20/11 [subscription required] expects the industry’s leading low-cost producer, First Solar, to reach 65 cents this year—a 35% reduction. In time, that cost is expected to drop as low as 50 cents. Second, the assumed capital cost of the rest of the plant (an additional dollar per Watt) will come down as experience in building and maintaining solar photovoltaic plants grows.

Third, the table’s cost of solar photovoltaic power assumes unrealistically high financing costs: a forty-year loan at 4% interest, which is more than homeowners pay today for financing their homes. With lower interest rates and shorter payback periods, let alone government subsidies and/or loan guarantees, that cost would drop considerably.

Finally, like all the table’s entries for electric power, the cost of solar photovoltaic power is based on the Nissan Leaf’s current electrical mileage. The Leaf is the very first all-electric production car, and that number is likely to go up with time and experience. There is no reason, in principle, why the cost of driving a mile on solar power cannot drop to half a penny or even below.

Nuclear electricity: This number is computed just like the number for conventional electricity. But it uses a lower cost for nuclear electric power, derived from Morningstar’s 1/13/12 investment report [subscription required] on Exelon Corp., a nuclear power company. The report says that Exelon’s nuclear power (no doubt at wholesale) costs $15 per megawatt-hour, as compared to conventional power, which costs $40. So the table reduces the price parameter for conventional electricity at retail by that ratio (15/40), thus: 11.6 cents/kWh x (15/40) = 4.35 cents/kWh.

Correction: An earlier version of this post put the three cheapest per-mile costs of driving at one-eighth the cost of gasoline. That ratio is correct for nuclear electricity, but the correct cost ratio for industrial-priced natural gas and solar photovoltaic energy is one-seventh that of gasoline. These differences are not significant because, as explained in the conclusions, the numbers in the table are not accurate enough to make such small differences reliable.

Obama on Iran: The Adult in the Room

Yesterday Atlantic Magazine published an exclusive interview of the President by crack reporter Jeffrey Goldberg on Iran’s nuclear program. Everyone who cares about foreign policy—let alone Iran’s nuclear threat—should read it.

In so many ways, the interview shows just how smart and skillful the President is. First, he gave his interview to a serious journalist in a serious publication (one one the few we have left). That alone shows how gravely he treats the issue. No one-liners or Fox rants for this President!

Second, he assuaged fears in Israel and here at home about how seriously he takes the threat of a nuclear-armed Iran to both Israel and the world. At least twice he pledged that we “have Israel’s back” and always will.

Third (and, for me, most important) the President’s substantial analysis showed how complex and nuanced the problem is, and how well he appreciates all the risks and possible unintended consequences of air strikes. Generously, he also gave credit to Israeli Prime Minister Benjamin Netanyahu for having similar understanding, thereby reassuring Netanyahu skeptics like me.

Fourth, the President offered strong reasons for a military option, but only if all else fails. His reasons included: (1) a nuclear arms race in the Middle East, (2) the risk of someone using nuclear weapons in sectarian passion, (3) the danger of utterly destroying the nonproliferation regime, (4) the risks of Iran giving weapons to terrorists, (5) the risks of further conventional war in the world’s most volatile region, (6) dangers to the global economy, and (7) the greater durability and reliability of a “solution” that Iran itself endorses.

Finally, the President made clear, in no uncertain terms, that he’s not bluffing in keeping the military option on the table. Here are his exact words:

“I think that the Israeli government recognizes that, as president of the United States, I don’t bluff. I also don’t, as a matter of sound policy, go around advertising exactly what our intentions are. But I think both the Iranian and the Israeli governments recognize that when the United States says it is unacceptable for Iran to have a nuclear weapon, we mean what we say.”

A man who has spent his entire political career practicing the diplomatic art of understatement need not say more. He certainly doesn’t have beat his chest or threaten. Coming from Barack Obama (let alone as President!), those sentences mean far more than any off-the-cuff, thoughtless frat-boy one-liner bubbling up on the Republican campaign trail.

Until reading his interview, I had been prepared to go all the way toward stopping Iran, except for war. But now I would support the President in an American—if not Iraeli—air strike to knock out Iran’s enrichment facilities if all else fails. Here, besides the ones the President outlined, are my own additional reasons:

An Israeli air strike might start a war with Iran. That’s especially so if it has to be repeated, as it would probably have to be to complete its mission. The US, being much stronger and more remote, is unlikely to suffer a real war. Iran might make reprisals against American forces, embassies and citizens in the region. But Iran has nothing like the global reach to make war on the US, let alone at home. A war like those in Afghanistan and Iraq is extremely unlikely because only our Air Force and Navy would participate. So an American air strike is better for both Israel and regional peace than one by Israel. (If one happens, I hope Israel will just sit on the sidelines and do no more than provide any necessary bases and logistical support. Plausible deniability would be a very good idea.)

Second, the Israeli Air Force simply doesn’t have the firepower that we do. After ten years of war in Afghanistan and Iraq, we have a huge arsenal of drones, cruise missiles, stealth fighters and stealth bombers, all with handlers and pilots having recent combat experience. We also have much bigger bunker-busting bombs. If even one of those penetrated Iran’s enrichment caverns while centrifuges were running, it would not only create absolute mechanical havoc. It would spew radioactive uranium compounds an all directions. Further enrichment work would require massive decontamination efforts, even if possible. Air strikes are not the best option, but if they become necessary, we can do the job.

The Israelis’ chance of completing the mission, let alone with a single strike, are much lower. Their chance of completing it without starting some sort of wider war are even less.

Finally, Bashar al-Assad has given us a more recent but equally potent reason not to shrink from air strikes. What would happen if a psychopath like him got nuclear weapons? Would he have used one, instead of conventional artillery, to massacre Homs?

A quick answer might seem “no.” But as you watch Syrian video of him strutting among his sycophants like a manic madman while his forces butcher his own people, the answer seems less clear. Would a man like him have any real understanding of what nuclear weapons can do? Even if so, would he have the insight and the modicum of empathy needed to understand why no one, anywhere in the world, has used nuclear weapons except to end history’s most terrible war? Psychopaths like him have none of those traits.

The risk of a tyrant using nuclear weapons to cow his own people and annihilate his own domestic opposition is something the world never really considered until recently. Now we should.

I personally think that Iran’s leaders are more rational and less psychopathic than Assad. But who really knows what they might do if their hold on power were threatened? If the regime had nuclear weapons, no external country would likely threaten to use—let alone actually use—its own nukes, for fear of starting a nuclear war.

Should we gamble the future of Iran’s own people, let alone Israel’s and the world’s, on untested hope? Might nuclear weapons make a gruesome tyranny durable, perhaps even perpetual? I think that’s a question no one wants to answer through sad practical experience.

So, for all these reasons, I would support the President in an American air strike against Iran’s nuclear facilities. But I’ll do so only if and when he, with his extraordinary intelligence, political skill, insight and empathy—plus secret, shared American and Israeli intelligence—says that all other options have failed and nothing else will work.

permalink