Let me start out by saying I'm not necessarily pro nuclear energy - I consider myself primarily anti-fossil fuels and I've come to believe over the last year or so that nuclear energy will be key if we want to get rid of coal and other fossil based fuels. Nuclear energy diaries may be the most contentious ones that are written on this site and since I started to comment in diaries on this topic, I've seen many of the same arguments come up over and over again, so I've decided to start a diary series to attempt to address some of these issues.
The first one tries to examine whether nuclear energy is really so much more expensive than wind or solar based solutions. If you're interested in seeing how much electricity produced from an overbudget nuclear plant in Finland, T Boone Pickens' wind farm, or a planned desert solar plant would cost, please read on.
Myth? Nuclear Power is More Expensive than Renewables
As no new nuclear plants have been built in the US since 1979, figuring out how much it will cost to construct one is somewhat difficult. The first new nuclear plants built in the US would almost certainly be a generation III+ reactor. One such reactor is currently being built in Finland, Olkiluoto 3, and is an example of how things can go wrong when constructing a new plant; its two years behind schedule and between 25% and 50% over budget. On the other hand, reactors in Japan as well as China and Korea have been built on time and on budget.
How much, without any time or budget overruns, are new nuclear plants supposed to cost? Westinghouse is planning on constructing two AP1000 reactors in Georgia, at an estimated cost of href="3 Billion. The AP1000 is a 1100 MW generation III+ reactor. A useful way of comparing the costs of different forms of electricity generation is to determine the cost per watt ($/W). The cost of these two reactors would be $5.9/W. Joseph Romm recently wrote an article in Salon entitled Nuclear Bomb where he came up with numbers of between $5.5 and $8.1/W, so the cost of the two reactors in GA is a reasonable ball park number. The NEI believes credible values are quite a bit lower, between $3.5 and $4.5/W (pdf).
How does this compare to either wind or PV solar? We'll look at wind first. Jerome a Paris played a role in financing the Princess Amalia Wind Farm, an offshore wind farm in the Netherlands (read his diary about it here). According to wikipedia, it is a 120 MW wind farm that cost €380 million, or €3.16/W (the conversion rate today would put it around $4.64/W). This is either equal to the cost of new nuclear (if you believe NEI) or about half the cost (if you believe Joe Romm). The big news in wind, and the largest wind farm I know of being built in the US, is an enormous (4000 MW, or a nameplate capacity equal to three or four new nuclear plants) being built by T Boone Pickens in Northern Texas. The cost, a paltry $10 billion for the turbines, or $2.5/W. However, its going to cost the state of Texas somewhere between $3 and $6 billion to build the transmission lines, though Mr Pickens has said he will pay as much as $2 billion to have the line built himself. If we add in $4 billion for the transmission line we're at $3.5/W. Note that this is a higher number than usually given for wind (a normal cost on shore or $2 to 2.5$/W), and the higher cost is almost entirely due to the cost of transmission. While the US has a large amount of wind available for harvest, a significant amount may lie outside of economically recoverable areas due to the cost of transmission, hence the idea of stranded wind.
How does PV solar look at the moment? Nanosolarhas been mentioned frequently here and elsewhere, with a goal of producing PV solar panels for $0.99/W. They are currently only selling to industrial scale companies, and we don't know what their current $/W is. A fantastic website to get current information is Solarbuzz, which has the current average price in the US at $4.83/W. Neither of these prices include the costs of installation. Its difficult to find costs for installation alone, but if we base our PV panel price on solarbuzz, and total installed price on numbers I found at Solar Power Authority installation may add between $2 to $4/W. That would put nanosolar at $3/W (there is reason to believe installation will fall to roughly $1/W here though - the general rule is installation costs about the same as the panels) and the average current price at $6.83/W.
The other form of solar is concentrated solar thermal. Andasol 1 is being built in Spain, and will be the world's largest (in terms of size) CSP. It is a 50 MW plant and cost €310 million or $9.3/W. Nevada Solar 1 is a 75 MW 75 MW plant that cost $266 million or $3.54/W. Finally, Solel Solar Systems will be building the newest largest CSP in the mjove desert. Mojave Solar Park 1 will be a huge 553 MW CST plant costing about $2 billion or $3.61/W.
$ Cost/W for aforementioned electrical generation |
Nuclear | $3.5-8.1/W |
On-shore wind | $2/W |
Pickens' wind farm | $3.5/W |
Solar PV | $6.83/W |
Nanosolar w/$2 installation | $3/W |
Nanosolar w/$1 installation | $2/W |
Andasol 1 | $9.3/W |
Nevada 1 | $3.54/W |
Mojave Solar 1 | $3.61/W |
The Catch - Capacity Factor
This table makes nuclear, unless it came in at the bottom end of the range, very expensive. There is, of course, a catch, and that is that each of these technologies has a varying capacity factor. The capacity factor of a power plant is the ratio of its energy production over a period of time to the energy it would have produced if it was operating at its nameplate capacity the entire time. The capacity factor for nuclear power plants is approximately 90%, coal is 70%, wind is 30%, and solar is 20% (table here, pro-nuc website but the numbers are generally accepted). What this means is that if we want to produce the same amount of electricity generated by a 1000 MW nuclear plant, we need to build about 3000 MW of wind turbines.
I don't know if there is an energy industry term for this, so I'm going to make one up now; $/We, or $ per watt nuclear equivalent. To get this number, we simply multiply the $/W by the capacity factor ratio, 3 for the case of wind, 4.5 for the case of solar. The CSP plant Andasol 1 has a CF of almost 41% (unfortantely this link breaks the diary, but its 179 GWh from a 50 MW plant) Nevada Solar 1 is about 20% and Mojave Solar Park 1 is about 30%. Capacity factor can be easily calculated by taking the amount of electricity generated and dividing it by 365 days a year, 24 hours per day, and finally by the nameplate capacity.
$ Cost/W for aforementioned electrical generation |
| $/W | $/We |
Nuclear | 3.5-8.1 | 3.5-8.1 |
On-shore wind | 2 | 6 |
Pickens' wind farm | 3.5 | 10.5 |
Solar PV | 6.83 | 30.74 |
Nanosolar w/$2 installation | 3 | 13.5 |
Nanosolar w/$1 installation | 2 | $9 |
Andasol 1 | 9.3 | 23.25 |
Nevada 1 | 3.54 | 15.93 |
Mojave Solar 1 | 3.61 | 10.83 |
According to Greenpeace Olkuluoto 3 that I mentioned at the beginning of the diary is going to cost at least €5 billion or $7.3 billion. At 1600 MW this puts this over budget reactor at about $4.6/W.
These are, at best, back of the envelope calculations, but I hope they give the impression that nuclear, though certainly expensive, is not any more expensive to build than any other available technology. In the end though, what matters is the price of electricity produced, and this is where renewables catch up, mainly due to operation and maintenance costs and financing and construction times. Now we're getting to the "pulling numbers almost out of my ass" part, but let's see what we get.
An estimated cost of electricity
The easiest part is operation and maintenance. From the Nuclear Energy Institute, we can see that O&M costs for nuclear are (in cents per kWh) 1.29 for O&M, 0.47 for fuel, 0.1 for spent fuel storage, and about 0.1 for decommissioning, for approximately 2 cents/kWh. Wind turbines and CSP plants do require some operation and maintenance, so we'll go with 0.2/kWh.
Financing and construction adds a significant cost to the price of nuclear, and much less to wind and solar (construction time to power on is much less, meaning less financing time). Here are the basic assumptions I'm going to work with. Construction time: wind 6 mo, CSP 1 year, nuclear 5 year. 80% financing amortized over the lifespan of operation which is 25 years for wind and solar and 40 years for nuclear.
$ Cost/W for aforementioned electrical generation |
| $/W | $/We | cents/kWh |
Nuclear | 3.5-8.1 | 3.5-8.1 | 6.4-12.2 |
On-shore wind | 2 | 6 | 6.1 |
Pickens' wind farm | 3.5 | 10.5 | 10.4 |
Solar PV | 6.83 | 30.74 | 30.2 |
Nanosolar w/$2 installation | 3 | 13.5 | 13.4 |
Nanosolar w/$1 installation | 2 | 9 | 9 |
Andasol 1 | 9.3 | 23.25 | 20.7 |
Nevada 1 | 3.54 | 15.93 | 15.5 |
Mojave Solar 1 | 3.61 | 10.83 | 10.81 |
Based on these numbers, Olkiluoto 3 is going to come out (assuming no further significant delays, which is certainly possible) around 8 cents/kWh, slightly more expensive than cheap on shore wind, but less expensive than nanosolar panels when they get to $2/W installed or any form of CSP.
Of course, take these numbers with a grain (or shakerfull) of salt. The $/W and capacity factor figures are good, but financing and O&M make trying to calculate the cost of a kWh difficult - I think these are reasonable approximations and I imagine we'll find out in the comments whether I'm wrong or not. I think, however, the idea that nuclear is far and away more expensive than other alternative forms of electrical generation is untrue and I hope I've convinced you of that fact as well.