Valid points here, but I'm not sure I agree with all of it.  For regulated monopolies - which is how most of our generation mix was built - fuel is a pass through.  Clearly, there is a societal risk exposure with fuel price volatility, but the folks making the investment decisions have behaved rationally - and the problem is one of too much regulation rather than too little.  We don't need some enlightened board making decisions about risk-allocation nearly as much as we need a regulatory regime that makes individuals responsible for the risks they take on.  (A great pet peeve of mine is that utility commissioners have absolutely no accountability.  No matter how bad the decision they make, they cannot be held accountable for those actions - and since the utilities they regulate get guaranteed cost recovery, they don't bear the cost either - we simply stick it all on the consumer.)

This isn't to say that I disagree with their point about risk, just that I'm not sure I like where it seems to head.

On another point though, there are very valid comparison that can be made on a c/kWh basis - we just need to make sure we include all the cents.  There are an awful lot of decisions that get made where we assume that the transmission infrastructure has no cost (it actually adds 3 - 4 cents, primarily for capital amortization).  Thus, one hears the assertion that "coal costs 3 cents, you technology costs 5 cents, ergo your technology doesn't work".  Not really, if my technology can be sited at the load and avoid the 3 - 4 cents of transmission costs.  This isn't necessarily a risk issue, but simply a matter of making sure we do apples-to-apples comparisons.  And again, this is an artifact of regulatory blinders.

All that said, I do like the piece - it just strikes me as being a bit too complicated in the wrong places, and a bit too simplified in others.  But still a very interesting set of intellectual questions to be asking.  Thanks for sharing.

David, I'm afraid this argument sounds very much like the "manufactured uncertainty" that critics of climate change keep drumming up to convince us that scientists haven't actually come to a consensus on the causes.

That may seem harsh, but in the argument here (obviously I haven't read the book itself) I don't see any quantification, and quantification of the uncertainties claimed is key to understanding whether they're actually important or not. Take the "continually shifting non-linearity" issue for instance. For a given generator, say a coal-fired power plant, you can quantify exactly what fraction of capacity it's running at, or whether it's turned off, and you get a standard capacity factor. That's really all you need to know to combine capital and fuel/maintenance costs to get the actual cost per kWh. There's no deep mystery from "non-linearity", it's quantifiable.

Similarly attributed cost of externalities may be a political decision, but it cannot be entirely arbitrary. In fact finding a way to quantify the externalities of carbon is exactly what carbon cap-and-trade markets are all about. If politics sets that cost too high, then the people with clean energy solutions can do very well. If politics sets that cost too low, we end up with a system with a bias against clean energy solutions (as we have now). But there is a quantifiable measure of that bias you get from the carbon price if you had cap-and-trade...

Capital costs of renewable solutions with current technology are known, and in particular for solar are very large. Built in to those capital costs are underlying energy costs: a substantial fraction of the cost of a standard solar panel can be attributed to the energy input to create the silicon, glass, and metal components. If the cost of fuel increases a factor of 10, your capital costs have just shot up too, so you can't think of the capital costs of renewables as intrinsically separate from the costs of other energy sources.

Yes there are differences that make the typical "cents per kWh" estimates somewhat arbitrary - most importantly the cost of financing in estimates I've seen has been all over the map (perhaps partly because of different estimations of risk premium). That's because people try to compare with a single number that doesn't always make sense. But there are real hard numbers behind the cost estimates. The choice of generating technology could still be "essentially political" because you can legislate anything, but is that a wise use of political power?

Just a few years ago, when oil was $20/bbl., solar heat cost could simple payback (zero-interest capital) in 8 years.  Now, with oil at $80, the simple payback is 2 years, two orders of binary magnitude.  

Renewable energy is new and future cost reductions are expected.  Energy is something like 25% of solar materials cost.  As fossil fuel prices increase so should inflation and discount rates.  So, it is complicated.  Depending on assumptions of future fuel prices, some 25 year solar graphs leap off the page.  

The net present worth of solar using fixed utility discount rates with fuel price increases can become quite unbelievable.

And nobody should use roof top pv for the baseline of solar capital and energy content.  That type of solar economy is probably an unsustainable ceiling.

Along the lines of Sean's comment, this is a good reference, and probably one a lot of people ought to read.  But to make it a lot simpler, he's talking about the industry convention of describing energy price by the "average" cost of a "marginal" (meaning the last) kwh that has to be generated at a particular time to meet a particular load.  Peak times produce some stunning marginal costs, but most of the time isn't peak, and more to the point, we don't pay the marginal price.  What we pay is the combined marginal price of all moments plus the capital cost, and from our perspective, what drives CHANGE in price, is largely (not entirely) CAPITAL costs, which the author probably deals with in another chapter, but not in this section.

And when capital costs are considered, the heierarchy he identifies goes belly-up.  Efficiency is the cheapest, at 3 cents per KWH, followed by wind (I hear that wind has been affected by the price shocks hitting the coal industry, so let me say I think wind is 20% cheaper than coal, plus it has no fuel costs, but duck the price issue) coal seems to be pushing 9 cents or so, internal capital only cost, and natural gas is right there.  Nuclear is ironically, too expensive to meter while efficiency has now taken over the mantle of being too cheap to meter.

A lot of people in and out of the industry still don't get it.  But Wall Street does, which is why all those coal plants are being cancelled.  Let's help them along and educate some more people about the true costs of electricity.

- Ned