I have a new piece in Slate that looks at the consequences of moving away from nuclear power. The climate policy analysis has attracted some thoughtful pushback. Let me address a few of the more important concerns.
I noted that DOE and EPA simulations of the next couple decades tend to find that a moderate carbon price boost nuclear significantly but does little for other sources. I also observed that we basically have three (not exclusive) options for near zero carbon power generation, which is what we’ll eventually need: nuclear; carbon capture and sequestration (for coal or gas); and renewables with storage. (Over the nearer term – lets say a couple decades – we also have gas without CCS.) I argued that if the United States eventually adopts a serious carbon constraint, the choice will ultimately be “between the devil [we] know and a technological prayer”.
The first line of opposition was to the DOE and EPA modeling of the Waxman-Markey bill (which I used as a proxy for a generic price on carbon). In particular, many argue that DOE modeling consistently underestimates plant construction costs. Alan Nogee of the Union of Concerned Scientists posted a helpful chart of DOE nuclear cost assumptions versus actual costs of real nuclear plants. I have a couple thoughts. First, the chart suggests that real costs could be lower or higher than the ones used in EIA 2010, though I’d agree that the high-end risk seems more substantial. Second, as Alan notes elsewhere, the EIA 2011 model is better. Yet the early release of the 2011 AEO suggests that the nuclear projections are similar to those from 2010. (The modelers see five new nuclear plants by 2035 rather than six, though it’s impossible to tell how much of this is due to higher capital costs, and how much is due to the assumption of cheap natural gas.) To really dig into the full consequences of the capital cost updates, of course, you’d need to remodel a carbon price using the 2011 NEMS code.
The second line of skepticism came from George Hoberg at the University of British Columbia. (Yes, for those of you who are clicking on these links, I now conduct all of my serious technical debates over twitter.) He flagged a paper from Energy Policy that suggested ways to get all global energy from wind, water, and solar by 2050. That paper does not rely on centralized energy storage to match supply to demand. Instead, it argues that a combination of demand management, hydropower for gap filling, long-distance grid integration, and the use of surplus power to produce electrolytic hydrogen would probably suffice. It also notes that the use of electric vehicle batteries for storage might be necessary – and that centralized energy storage might even be required. Relying on a hydrogen economy to materialize, or betting that centralized energy storage either will be possible or won’t be necessary, strikes me as qualifying for the label “technological prayer”. That is not to say that I dismiss the possibility – far from it – but it’s far from something we can count on.
The last challenge comes from David Roberts at Grist, who asks whether, in a world of limited resources, “variety” and “all of the above” are actually real options. (He points me to a thoughtful recent article in which he fleshes out his ideas.) I have mixed feelings about this. My guess is that the answer depends on a few things. If we have a hard, black and white emissions goal, where if we hit it all will be ok, but if we miss it we’re doomed, we might want to concentrate all our (limited) firepower on one technological bet and hope that it turns out ok. If, on the other hand, various degrees of success all count, we might want to adopt a more resilient investment and innovation strategy, even at the expense to lower odds of delivering a big bang. The answer also depends on exactly how constrained the innovation budget is. And it depends on the timelines we have in mind for steering money into the system: yes, there are increasing returns to scale up to some point, but beyond that, simply pumping more money into one area will yield declining results, and perhaps even backfire through cost inflation. (For more on all that, this paper by Varun Rai and colleagues is great.)
Bottom line? I certainly wouldn’t put all my bets on nuclear, but I’d be wary of putting all my bets on some other technology too.