Diatribes of Jay

This is a blog of essays on public policy. It shuns ideology and applies facts, logic and math to economic, social and political problems. It has a subject-matter index, a list of recent posts, and permalinks at the ends of posts. Comments are moderated and may take time to appear. Note: Profile updated 4/7/12

17 December 2013

Comparative Energy Costs of Driving, Updated: Cents per Mile

[For my personal award to Apple Computer, click here. For recent posts on upheavals in Ukraine, click here.]

Almost two years ago, I published a table giving the comparative costs per mile of driving on various sources of energy. It was a popular post, but it’s time to update it.

As investors and drivers both know, energy prices can change. To quote legal boilerplate, “past performance is no assurance of future results.” Some things get cheaper, and some get dearer.

So I’ve revised my table to show late-2013 figures, side by side with early 2012’s. (The precise month in late 2013 varies with the type of energy. See the notes below.) Juxtaposing the two figures provides a sense of change, trends and volatility for each source.

The comments to this table also make a point I should have made in early 2012. Unless you generate your own electricity—for example, using a solar array—the technology used to generate it doesn’t make much difference in the price you pay as a consumer, and therefore in your energy cost per mile if you drive an electric car. Electricity is fungible, and power companies don’t provide discounts when they use cheaper means of generating it in your area. So you pay the market price, which today is still set mainly by the cost of burning natural gas or coal, even if your power company generates much of its energy portfolio with the sun or the atom.

When you put this point together with the high cost of energy from new nuclear plants and the now-lower cost of solar PV energy, you get some surprising results. Read on.

Here are the comparative results for March 2012 and December 2013, with the categories re-ordered to reflect decreasing per-mile costs today:

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

Energy SourceUnderlying Price ParameterCents per Mile Driven
 March 2012Late 2013March 2012Late 2013
Gasoline$3.78 per gallon$3.35 per gallon12.611.2
Natural Gas (Residential)$1.28 per gal. equiv.$1.89 per gal. equiv.4.36.3
Nuclear Electricity4.4 ¢ per kWh15 ¢ per kWh1.55.1*
Conventional Electricity
11.6 ¢ per kWh12.5 ¢ per kWh44.3
Conventional Electricity
6.8 ¢ per kWh7.23 ¢ per kWh2.32.4
Natural Gas (Industrial)$0.55 per gal. equiv.$0.53 per gal. equiv.2.2**2.0**
Solar Photovoltaic
Electricity (Residential)
N/A4.6 ¢ per kWhN/A1.5
Solar Photovoltaic
Electricity (Commercial)
5.1 ¢ per kWh2.4 ¢ per kWh1.80.8

* For new plants
** Assuming that service station’s retail sale of industrial gas would add only 20% for operating expenses and profit.

Comments: Three points jump out from this table. First, the cost per mile of driving on nuclear electricity increased dramatically from early 2012 to late 2013, by more than a factor of three. The reason is that the 2013 number is for new plants and the 2012 factor for old, fully depreciated plants, which in my view are obsolete and therefore dangerous.

The 2012 figure came from a Morninstar investor report on Excelon Corp., a partially-nuclear power company. That report is no longer available. The figure for new nuclear plants comes from a statement by David Crane, as reported in The Economist [subscription required]. Crane is the CEO of NRG Energy, which dumped plans to build two new nuclear plants in Texas after spending $331 million. His reported figure was 10 cents per kWh for generation cost, which I increased by 40% for distribution and 10% for profit (the average in early 2012 over all generation technologies) to get a residential retail price.

The moral of this story is that we can have cheap nuclear electricity, but only if we risk another Chernobyl or Fukushima. If we build new, safer nuclear plants, the cost of nuclear electricity will go up, higher even than for conventional (non-nuclear, non-renewable) electricity today.

In any event, I’m not aware of any power company that actually charges anything like 5 ¢ per kWh or below at residential retail. One could, in theory, if it had only obsolete nuclear plants. But in practice power companies use a variety of generation technologies and tend to charge the going market price for output, which today is north of 12 ¢ per kWh at residential retail. So insofar as concerns consumers, nuclear power differs little from the conventional electricity reported lower in the table.

This second salient point is that, in a bit more than a year and a half, the per-mile driving cost of all sources of energy went up, except for three: (1) gasoline, (2) natural gas at industrial rates (not at residential rates), and solar photovoltaic energy. Except for the last, the cost per-mile decreases were within the margin of error of this table: 11% for gasoline, and 6% for industrial natural gas.

The only driving cost that decreased substantially was that of commercial solar PV energy, which dropped by more than a factor of two. The reason was real, not mere price volatility. During that time both the purchase price per watt of solar cells and the “turnkey cost” of making a working array from them went down substantially. This post explains the phenomenon and its future prospects.

The final point surprised even me. The cost per mile of driving energy from our own small-scale residential solar array, installed this August (and described and pictured here), beat the cost of every other form of driving energy, except for commercial solar photovoltaic energy.

This point is real. It’s based on the actual price we paid for our array, before considering federal and state tax credits, but amortizing that cost over the array’s projected energy output. The resultant figure is only 1.5/11.2 = 13% of the cost of driving on gasoline. (There is no adjustment for distribution or profit because the array is ours; we paid for it and own, maintain and operate it.)

There is no comparable residential solar figure for 2012. Until we installed our array this last August, we had no accurate idea what residential solar PV energy would cost.

Unfortunately, since electricity is fungible, regardless of source, no consumer is likely to see the benefit of the extremely low cost of commercial solar photovoltaic energy until that source produces a lot higher fraction of our nation’s electric energy than it does today. So right now, today, installing your own solar array gives you the lowest cost of any means of moving a car forward, other than pushing it.

A close second is natural gas at industrial rates, resold at retail rates by commercial service stations. But unless you are a large organization that can negotiate industrial rates for your natural gas, you have to go to a service station to enjoy the substantial price advantage over residential rates. You can’t “gas up” at home, as you can if you have your own solar array.

Installing your own solar array also has a decisive medium-term (and long-term) advantage over converting your car to run on natural gas, or buying a new natural-gas or dual-fuel car. That’s price stability. Because solar PV energy requires no fuel and has very low maintenance expense, the energy cost of driving on the sun depends almost entirely on the cost of installing your solar array. That energy cost will therefore be stable for the array’s working lifetime, probably a century. But natural gas will get more and more expensive as power companies use more of it as a cleaner substitute for coal and drivers use more of it as a cheaper and cleaner substitute for gasoline. If you want assurance that your driving energy costs won’t jump in ten years, invest in driving on the sun.

My wife and I will be long dead before the solar panels in our array have to be replaced (barring accident or other insured loss). So we’ll also be long dead before the amortization of the array’s cost ends. But our heirs or future owners our house will have essentially free energy for driving, for about a century. (Our realtor tells us that we (or our estate) won’t recover the array’s full cost when the house sells; instead, the array will be a selling point making the property easier to sell, at least to educated buyers.)

So the key “takeaway” from the table is simple. Right now, today, you can cut your energy cost per mile of driving by installing a solar PV array and using it to power an electric car. In fact, you can cut that cost by a factor of 11.2/1.5 = 7, as compared to gasoline in a 30 MPG car. And once your array is paid off, you can drive for free, at least as far as concerns energy.

You’ll still have to buy an electric car and maintain it. But maintenance promises to be much cheaper for electrics than for gasoline cars, with the possible exception of the batteries. Once electric cars become true mass-market vehicles, car companies will no doubt reduce this expense with warranties, battery swaps, and/or battery replacement programs (with prorated expenses for use, as for used tires). (There are lots of ways for car makers to do this. For a complete analysis of one, click here.)

The driving cost of energy from your own solar array is more real than that from a commercial array. As noted above for nuclear energy, power companies mix and match energy sources and charge the going all-means rate for solar energy regardless of its low generating cost.

But if you have your own solar array, you can recharge your car’s batteries directly, an act that also solves the solar intermittency problem. If you use your car during the day, as most people do, you can avoid the entire cost of charging at night if your power company has “net metering,” which many do. (“Net metering” means charging you only for the difference between energy you use and energy you generate, regardless of timing and intermittency.)

Notes: The methods of calculation are precisely as described in my earlier post (after the table). The only differences are in the underlying cost parameters, which are mostly taken or derived from data from our own Energy Information Administrations website, as follows:

Gasoline: EIA Weekly Gasoline and Diesel Retail Prices, All Grades, for 12/09/13

Natural Gas (Residential and Industrial): EIA US Natural Gas Prices, Monthly, for September 2013

Conventional Electricity (Residential and Industrial): EIA Total Energy, Data, Monthly Energy Review, Table 9.8 Average Retail Prices of Electricity, August 2013

Solar Photovoltiac Electricity (Residential): The projected cost of energy from our own solar array, with a turnkey capacity price of $5.98 per Watt, based on this table, which converts turnkey array (capacity) prices to output energy prices. (There is no distribution or profit factor for residential arrays.)

Solar Photovoltaic Electricity (Commercial): The cost of generation using a large-scale commercial array with a state-of-the art turnkey price of $2 per Watt capacity, amortized over the lifetime of the array, plus a 50% distribution and profit adjustment, as explained here. [Scroll down to notes on “Solar Photovoltaic Electricity.”]

As noted above, no power company is likely actually to offer energy at this rate, for two reasons. First, power companies mix and match power sources and charge the going rate for energy generally, regardless of source. Second, solar power is intermittent, so power companies need a method of storage or other methods of generation to use it efficiently. This gives them a reasonable excuse never to sell solar PV energy, with its world-beating low generation cost, by itself. But if you have your own solar array, your car’s batteries, or your power company’s net metering, solves the intermittency problem for you, transparently and with no extra effort or additional investment on your part.

Erratum: An earlier version of this post calculated the advantage of our small-scale solar array over gasoline as 11.2/0.8 = 14, or only 7% of the energy operating cost. Those figures, of course, are for a large-scale commercial solar array, not for our small-scale residential retail array. Big firms that install their own large-scale arrays to run a fleet of electric cars could realize that advantage, but consumers like us cannot, yet. I regret the error.

An Award to Apple

Congratulations Apple! For 2013, you win my Gratuitous Customer Screwing Award. Hands down.

Whatever possessed your programmers to mangle your file-handling commands and demolish longstanding industry standards so? Whatever executive-cretin approved the mindless, makework changes?

Wasn’t it enough that the venerable “Save As” command disappeared from your recent O/S updates without a trace? In your new “Mavericks” operating system, did you really have to change “Close” to mean “save and close”?

Now inadvertent changes to a forgotten document sitting on a desktop for days or weeks become permanent on closing it. The next time a user opens it, he or she has to scan through and look for damage. What a colossal waste of time!

If you want to create a new version of an old document and keep the original, you have to open it, copy it, open a new blank page, and copy the document in. If you open it, change it, and close it, just out of habit, there goes the old document. Without a “Save As” command, there’s no simple way to preserve the original and put the modified file in a new place.

And God help you if you happen to cut, rather than copy, the original document and then overwrite the cut-page memory buffer. When you close the file, you get a blank page in storage, and your backup file disappears. Bye, bye hard work!

All this needless churning gives the term “maverick” a whole new meaning. For the personal-computer industry’s entire history, file-handling commands have had simple, intuitive, universally accepted meanings. “Close” meant, simply, close the file without saving. “Save”—not surprisingly—meant “save the file.” “Save as” meant save the file with another name and/or in another storage location.

So simple and intuitive. So longstanding. So venerable. So well established in the habits and cerebella of all who didn’t buy their first computing device yesterday.

But tradition and customers’ habits mean nothing to computer programmers who think they own the world, or at least its virtual reflection. Who do they think they are? bankers?

Now I look forward to equally destructive and gratuitous new initiatives by Apple. Why not have its corporate vehicles drive on the left hand side of the road, out of solidarity with the Brits and Japanese? Why not use “left-handed” nuts and bolts that open clockwise and close counter-clockwise? Why not put car doors on the roof of the car, so that the sides can be free of obtrusions and dings? Why not change every command on the menu? After a few years of total confusion, won’t that give Apple a lock on users’ habits?

Why not just go the Full Monty and make products out of anti-matter, then watch customers explode?

The cloud is the excuse, you say. Well, would “Save in Cloud” and “Save Locally” be too hard to understand? Must customers like me, who’ve used computers for decades, adapt to cell phone and tablet users who are too lazy or forgetful to save before closing a file? Why not add new commands, rather than change ones customers have been using for decades? Or at least why not give customers not raised on smart phones a choice of command regimes?

Apple’s programmers, we are told, are among the industry’s most “innovative.” So we can assume that they were thinking about something when they made this mess. Whatever it was, it certainly wasn’t existing customers.

It’s never too late to change back. Remember Coca Cola’s misshapen bottle!

New customers in their teens may laugh at an old geezer like me. But after a few years of constant, gratuitous turmoil and churning in their online work and social environments, they won’t be laughing any more. They’ll be looking for providers that change only what’s absolutely necessary to add useful new features but leave the basics alone.

I don’t know whether Apple will respond to this online complaint. But I do know one thing. Unless it does, I’ll be using Google, not Apple, for cloud services. Google’s programmers think of existing customers first, and they usually leave at least an optional path back.

Footnote: I know, I know. The title “Mavericks,” plural, comes from the name of a surfing spot in California, not the word. But there’s an obvious double entendre: Apple prides itself on being an industry maverick and therefore (in its mind) an industry leader. Since when does a leader slam existing customers to attract new ones or sell new things?



Post a Comment

<< Home