Yesterday, I wrote about a new peer-reviewed paper from inventor Nathan Myhrvold and climate scientist Ken Caldeira. It found that, if there is to be any hope of staying in the zone of climate safety (or at least semi-safety), the transition to carbon-free energy must begin immediately and cannot include any merely “low carbon” sources like natural gas.
I sent Myhrvold a few follow-up questions. Here are his responses, lightly edited.
Q. Back when you were quoted in SuperFreakonomics, you seemed skeptical, if not contemptuous, toward the “immediate and precipitous anti-carbon initiatives” pushed by climate activists. But your paper seems to suggest that only such initiatives have any hope at all of keeping us in the zone of climate safety. Have you changed your mind about the need for immediate action? And/or the wisdom of activists?
A. Yes and no.
First, while I love the SuperFreakonomics book, it wasn’t my book and was never intended to fully represent my views on a complicated topic like this.
One thing we wanted to point out is how hard the problem is. There are many advocates who say that it will be easy to switch to a low-greenhouse-gas energy solution. They have a great “can do” attitude, but their schemes are hopelessly flawed. To beat one dead horse, that little book 50 Simple Things You Can Do to Save the Earth — well, it’s at best wishful thinking and at worse deeply disingenuous. Those “simple things” are not going to save us. At most they can let you foolishly think you are actually accomplishing something when you are not.
There are many more serious proposals to move to wind or solar energy — the so called “Pickens Plan,” for example, or a bunch of other people’s supposed “plans” to convert to solar or wind power — that simply don’t add up. I am contemptuous of “plans” that are not real and which at best fool people into thinking the problem is easy.
There is another set of people who have proposals that might work, but they are so draconian that no politician is going to implement them, and if they did, the population would revolt. Banning all new fossil fuel plants would surely help — try doing that. It is not realistic to think that you can get the world to change overnight. That is one reason why our paper models transitions lasting from one year (clearly an impossible lower bound) to 100 years (which is very slow).
The most galling examples are people who argue against researching geoengineering on the basis that we already know how to solve the climate problem. To them, the climate problem is a done deal, we just haven’t implemented yet. I think that is just absurd, but you see smart people make this argument.
There is a great quote attributed to Einstein: “Everything should be made as simple as possible, but not simpler.” Unfortunately, the clean-energy debate rarely avoids dipping into the “simpler” territory.
Q. More broadly, what do you see as the policy implications of your work? Is it, “We need to mobilize on the scale of WWII and build out clean power?” Or is it, “There’s no way to do this, so we need to spend money on adaptation and researching geoengineering?” Or some mix?
A. Some mix.
We need to invent new energy technologies. As one example, we can’t store energy worth a damn at the moment. Well, with pumped hydro we can store it in a dam, but only if we are lucky with geography. Without storage, wind and solar are very difficult to fully utilize.
We need higher efficiency in solar. We need new kinds of nuclear technology (I am working on this myself). We need lower cost for all of these things. It would be nice if carbon capture and sequestration would work — unclear how well it does, since it has never been tried at scale.
So there is a lot of energy invention to do.
I think we also need to investigate geoengineering. Right now, 2012 will have higher CO2 than 2011. Does anybody believe we are doing the things now so that 2013 or 2014 will be less than 2012? We are not over the hump yet, and until we are, I think that we need to understand that set of options — they may be real, they may not be, but it would be inexcusably irresponsible not to understand them.
We probably also need to investigate adaptation to a warmer world.
Q. Your finding with regard to natural gas flies directly in the face of enthusiasm (even among some greens) about fracking and the “100 year supply” of natural gas found in the U.S. Is the conclusion that natural gas can’t be a part of a clean energy solution at all? And have you heard from any of natural gas’s many fans, pushing back on your conclusion?
A. Natural gas is definitely cheap. It is definitely in big supply (100+ years). Those are true, and we don’t dispute them.
Prior to our study, nobody had looked in detail at the climate impact of converting to natural gas. People relied instead on the intuitive idea that emissions were “half of coal.” But climate change is nonlinear. It turns out that cutting by half is nowhere near enough to have an impact this century. In fact, it may take 200 years to get a measurable impact.
People are surprised by this and some people have pushed back, saying, “how can this be so?” Well, the detailed answer is all of our equations, but the simple answer is that due to the time lags in the system, you can’t get a near-term benefit until the emissions cut back is dramatic — like 10X.
The value of work like we have done is precisely that it highlights counterintuitive results. Those results may not fit with somebody’s first guess, or to the gas industry’s entrenched position. It is unfortunate that gas does not help, but that is where the science leads us.
One can try CCS with gas (our CCS cases include both coal and natural gas-based CCS), but even there you need to be careful, because the LCA estimates for CCS don’t show as much benefit as people think. Most CCS claims to eliminate 90 percent of CO2 emission from combustion, but the total LCA estimate must include leaks of natural gas during production and the reagents used in the CCS process. The net result is that CCS is far less impressive than you would think on an LCA emission basis.
One can also hypothesize better technology in the future (our paper has a whole section modeling what technological improvements could do), which is great, but it is important to realize when you are making up a fairy tale and when you are making a prediction based on current understanding. Future technology may be magical, but if it does not happen soon, it can’t matter all that much to climate change this century. It is already getting to be too late. That is the trouble with the time lags.
Right now, the best available LCA estimates on natural gas show it has at most very modest benefit over coal, so little that it should really not be viewed as being better for the climate than coal. Even a big reduction in this — as occurs for gas-based CCS — would still not make it a great solution.
I wish it were otherwise.
Q. The “carbon debt” incurred by the construction of power plants plays a big role in your paper, partially explaining why the carbon benefits of clean power take so long to manifest. Where did you get your info on carbon debt? Can you say a little about it and the role it plays?
A. There is a whole field of life-cycle analysis (LCA) wherein people study the total emissions required both to make a power plant and to operate it. This study by Sergio Pacca and Arpad Horvath is a good example and pretty readable (they are not all that way!). These studies tell you how much emissions come from each technology.
We surveyed the field and picked a high and a low LCA value for each technology we studied. That is where the basic emission data come from.
Ideally, the LCA studies would all agree. They don’t, because they have different assumptions. Solar photovoltaic cells require a lot of electricity to make. If you assume that your solar cells are built in Europe with a European mix of electricity generation, which includes French nuclear and some wind and other sources, you get one answer. If you build the identical solar cell in China, which has mostly coal-generated electricity, you get a different answer.
Now that our paper is out, it would be good for the people who make LCA studies to revisit their assumptions and try to narrow the gap between the high and low estimates that we used.
Q. Some energy analysts predict an increasingly tight oil supply and volatile/rising oil prices, which could stifle economic recovery and suppress growth. Have you thought about running scenarios where resource shortages (oil, water, etc.) play a role?
A. We wanted to isolate the climate effects alone of shifting a fixed amount of electrical power generation (1 TWe as base case). Since that is today’s coal-powered electricity load and we are currently growing that load, it would take a very severe downturn to keep demand such that coal is only 1 TW for the 40 to 60 or so years that a transition takes. So we already modeled a very conservative scenario.
More realistically, over a 40-to-60-year transition, demand will grow, so you need to convert a lot more than 1 TWe of coal.
Total per-capita energy use in China is about one-fourth to one-fifth that in the U.S. In the next 40 to 50 years, they will be a lot closer to the U.S. (and U.S. levels will also rise). That isn’t in our study — we assume fixed demand of 1 TWe.
The best thing that could happen for climate is for the price of fossil fuels to rise. Unfortunately, the development of cheap natural gas through fracking has done the opposite. There seems to be no shortage of cheap gas.
Q. Given the alarming results of your modeling, do you plan on getting more personally involved in the climate/energy area? If so, what might that look like?
A. Yes, if there is an opportunity. I am not a political guy, so I am not interested in advocacy or policy persuasion. It’s just not what I do, nor am I good at it. I can invent, and I like science. We have done a lot of invention around trying to solve clean energy problems. This paper with Ken was an effort to make a direct contribution to the science. If I see more opportunities — either in science or invention — I will definitely work on them.