Thursday, August 14, 2003

previous entry | main | next entry | TrackBack (0)

John Deutch and M.I.T. flunk energy economics

The former CIA director, along with Ernest Moniz, professor of physics at MIT and former Department of Energy official, open their op-ed in today's New York Times with the following two grafs:

The world needs both more electricity and less pollution. The goals are not incompatible, but the solution will require better management of demand, smarter use of coal as well as renewable energy sources, and increased use of nuclear power.

As Congress considers an energy bill when it returns from recess, it will be under pressure to expand or limit the use of nuclear power. The issue, however, is not simple. More nuclear power will be necessary — but more nuclear plants will be built only if more safeguards and incentives are put in place. The challenge is to make nuclear energy safer, cleaner and more economical.

OK, sounds reasonable so far. Then I read the next three grafs:

We built a model to compare the costs of producing electricity from new nuclear, coal and natural gas plants. The model focuses on economic cost, not regulated or subsidized cost. According to our study, the baseline cost of new nuclear power is 6.7 cents per kilowatt-hour, compared to 4.2 cents for coal and natural gas (when the price of gas is $4.50 per thousand cubic feet). Plausible, but unproved, technology could reduce nuclear costs to those of coal and gas.

However, if a cost is assigned to carbon emissions — either through a tax or some other way, as in a current Congressional proposal that would limit emissions but allow companies to buy and sell the right to discharge more pollutants — nuclear power could become an attractive economic option. For example, a $50 per ton carbon value, about the cost of capturing and separating the carbon dioxide product of coal and natural gas combustion, raises the cost of coal to 5.4 cents and natural gas to 4.8 cents.

Even under these favorable circumstances, the regulatory uncertainty threatening the large-scale investment needed for a nuclear plant will require some government assistance. A production tax credit, similar to that extended to wind power, is a good idea. It would give private investors an incentive to complete a plant. If no plant is built and operated, no public money is spent. If the first plants are indeed built and operated competitively, more will follow and the possibility of reducing greenhouse gases increases. (emphasis added)

So, in other words, even after one factors environmental externalities into the cost of energy production, coal and gas are still more efficient energy choices than nuclear power. Bear in mind that the op-ed suggests that this calculation does not include the cost of disposing nuclear waste, so in all likelihood the gap in efficiency is even greater.

The conclusion I draw from this cost-benefit analysis is that compared to coal and gas, nuclear power is an inefficient substitute and should not be taken seriously. Deutch and Moniz argue that the government should just subsidize nuclear power and make vague allusions to reducing greenhouse gases. They ignore that their own analysis suggests nuclear power should be rejected in any comprehensive energy plan.

This op-ed was borne from a cross-disciplinary MIT-sponsored study. Having just read the chapter on "Nuclear Power Economics," I'm even more skeptical of the boosterism for nuclear power. Here's an optimistic assessment on p. 41 of the report:

These results suggest that with significant improvements in the costs of building, operating, and financing nuclear power plants, and continued excellent operating performance (85% capacity factor), nuclear power could be quite competitive with natural gas if gas prices turn out to be higher than what most analysts now appear to believe and would be only slightly more costly than coal within the range of assumptions identified.

Let me rewrite this a bit:

If all of our optimistic assumptions about nuclear power are true and all of our pessimistic conclusions about alternative energy sources are also true, nuclear power will still be less energy efficient than coal and barely on par with gas.

I certainly could be missing something here, but I don't think so.

MIT has one of the best economics departments in the country. How did they sign off on this?

posted by Dan on 08.14.03 at 11:47 AM


I believe that the French, with more sensible and less paranoid safety rules than we have in their nuclear plants, are able to operate their plants cheaper than we can. But I don't have data to back that up. Do you?

posted by: Registered Independent Joel on 08.14.03 at 11:47 AM [permalink]

Since my uncle Paul seems to have been involved in writing the MIT report on which the article is based, I think I should defend at least that work. If you read the Executive Summary, you'll see that the committee simply takes it as a given that America should reduce carbon dioxide emissions and then dedicates itself to determining what the best way to do this would be. The answer is nuclear power. How it is that Deutsch and Moniz then wound up pricing the cost of emissions at less than the amount that makes nuclear the efficient option is a bit beyond me.

posted by: Matthew Yglesias on 08.14.03 at 11:47 AM [permalink]

My query too: is nuclear power too costly only for the US? or in France and Japan are subsidies required?

posted by: Wondering on 08.14.03 at 11:47 AM [permalink]

This is from the executive summary:

In deregulated markets, nuclear power is not now cost competitive with coal and natural gas. However, plausible reductions by industry in capital cost, operation and maintenance costs, and construction time could reduce the gap. Carbon emission credits, if enacted by government, can give nuclear power a cost advantage.

So, I don't think disparate regulations matter.

Also, looking at the chapter I mentioned in the post, it looks like those carbon taxes have to be pretty steep for things to pan out.

I have nothing against nuclear power if it's a cost-viable option. But the MIT study and the NYT op-ed actually made me more skeptical of its utility.

posted by: Dan Drezner on 08.14.03 at 11:47 AM [permalink]

Small quibble- you said that nuclear power would come out less "energy efficient." Not sure what that means; in terms of converting potential energy (nuclear or chemical) into electricity, yes, nuclear does that less well. However, the energy source decision is fundamentally a political decision; not economic or even environmental. We need to decide what is the best overall choice given how we value environmental effects, cost of electricity, safety and health risks, etc. So a study (which I admittedly have not yet read but will) which tries to condense a decision into solely economic terms will not elucidate in any way the overall question of "what energy sources should we be using/investing in/researching?" Those questions need to consider other values as well.

posted by: JN on 08.14.03 at 11:47 AM [permalink]

Unless I missed something, they never claim that nuclear power will be more efficient on the bottom line. I inferred they were saying the cost difference was small enough to be negated by the envorinmental benefits (emissions, anyway) of nuclear power.

posted by: Justin McAleer on 08.14.03 at 11:47 AM [permalink]

Justin - The selections Dan quotes suggest that a $50 per ton tax on carbon emissions would achieve the same environmental benefits as switching to nuclear power, but at lower cost.

posted by: pj on 08.14.03 at 11:47 AM [permalink]

I'm hesitant to encourage any more NY Times bashing, but you also have to wonder why the op-ed page editor didn't at least raise an eyebrow when he or she read this submission. It has always struck me that these things get read and accepted with far more of an eye to the identity of the author and/or the timeliness of the subject than to anything else.

posted by: pblsh on 08.14.03 at 11:47 AM [permalink]

to answer the question "they have one of the best econ depts, why did they miss this?" the answer is: they had an idea and they decided to see what it would take to support the idea.

this is a typical MIT solution to a problem, and it's a typical way that science is done at the academic level. (I say that having attended MIT and UC Berkeley for ugrad and grad respectively.)

the physics and chemistry professors, Moniz and Deutsch respectively, said "we start at the point fossil fuels are bad. we then ask what can be done to improve the situation."

Professors and academic scientists, regardless of what you might think, often start with assumptions and don't test them. Then, given those assumptions, they then suppose a correct solution to a problem. They then design hypotheses and design experiments, or test meta data, or analyze data in order to arrive at a conclusion. But none of those things will ever challenge the underlying assumption, because none of those experiments were ever designed to.

Since the authors simply already agreed that carbon-dioxide emissions were bad, they didn't care about the fact that nuclear power is more expensive. They ignore their own analysis in reaching the point that economically, it's a bad decision. But they weren't economists. And an economist wouldn't fault their argument, necessarily. An economist would just say that if in fact you take it as a given that there is a viable reason to handle your limited resources in this fashion, this is the cost.

Note also that regardless of being economists, economists fall all over the political spectrum. This is because of their underlying assumptions of what is best. For example, slavery might be a much cheaper way of having labor than not having slaves (might not, i don't know.) Let's assume it is, for argument's sake. Economists don't argue slavery should be legal on economic grounds. They simply say "given we pay for labor, this is the result."

Basically, MIT professors are just as bad as the rest of academia at not challenging their base assumptions. Being bright doesn't make you a good phenomenologist. Virtually all science functions by not challenging base assumptions until you accidentally run the current direction of research aground --having exhausted all possible progress without challenging those assumptions.

posted by: on 08.14.03 at 11:47 AM [permalink]

"Bear in mind that the op-ed suggests that this calculation does not include the cost of disposing nuclear waste, so in all likelihood the gap in efficiency is even greater."

There's already a tax on nuclear power earmarked for waste disposal so that cost is included. But 6.7 cents/kW-hr? Yikes!

posted by: Bill Woods on 08.14.03 at 11:47 AM [permalink]

You're suggesting we decide a very complex issue based on simplistic and unreliable assumptions. In the first place, the technology for removing carbon dioxide from fossil fuel emissions has never been tested on a large scale in a working plant. Assumptions about its cost are purely speculative. You completely ignore the problem of particulates from coal and other fossil fueled plants. Estimates of the numbers of deaths they cause worldwide every year run into the hundreds of thousands. You are concerned about radioactive waste, but are apparently unaware that there are typically about two parts per million of radioactive uranium and thorium in coal. Both are emitters of alpha radiation, which can't penetrate your skin, but is potentially deadly if breathed in the form of particulates. One often hears of the need to store radioactive waste for thousands of years. In fact, the amount of residual radioactivity from operating a nuclear plant for 30 years is the same as that for operating a coal plant for the same period after about 550 years. The difference is that, in the case of a nuclear plant, one can control and store the waste. Even with particle precipitators and other emission controls this won't be the case with coal plants. The radioactivity they release will just be out there in the environment, and, with a half life of over a billion years, it won't be disappearing anytime soon.

I expect nuclear power will continue to cost more than fossil in the short term, although the gap will narrow considerably in a reasonable regulatory environment. The major difference between them is their environmental cost. Power from modern nuclear plants is not risk-free. It is, however, vastly safer and cleaner than that from coal burners. At the very least it's worth looking at this issue carefully and objectively. Instead we've let it become one of the demons the ideologues among us are always so busy saving us from. It deserves a closer look.

posted by: Doug Drake on 08.14.03 at 11:47 AM [permalink]

Post a Comment:


Email Address:



Remember your info?