We’re now on our third week of posts seriously looking at Amory Lovins’ and the Rocky Mountain Institute’s latest paper that bashes nuclear energy. Today’s post deals with the paper’s claim that nuclear energy’s “true competitors” (according to RMI) are cheaper and therefore “produce” more “climate solution” than nuclear. I will show that RMI relies on weak sources, no sources, and cherry-picked data for their cost assumptions to exaggerate their claims.
From page 19 in RMI’s paper (pdf):
Every dollar spent on new nuclear power produces 1.4-11+ times less climate solution than spending the same dollar on its cheaper competitors. For a power source merely to emit no carbon isn’t good enough; it must also produce the least carbon per dollar…To come up with the above statement, RMI’s paper takes the cost assumptions for each technology from their graph below, inverts them to get kWh per dollar, finds each technology’s “CO2 emissions displaced relative to coal,” multiplies the kWh per dollar with the “CO2 emissions displaced relative to coal,” and then compares each technology’s results to nuclear to come up with the above statement. If you understood the first time what I just wrote in the previous sentence, then you’re a genius. It took quite awhile for me to make sense of this, but whatever. My post deals with the cost assumptions in the graph below.Initial Thoughts
I’ve gone through RMI’s Excel spreadsheet, methodology and 52-page paper and what they have basically done is picked and chosen many different data points from many different sources to compile the above graph. The primary cost assumptions are found in rows 20-38 in the worksheet titled “Climate Data” from RMI’s Excel spreadsheet. When digging into the numbers, I found the worksheet was extremely hard to follow, it doesn’t explain certain calculations, and calculates practically everything differently. It’s one big mess in my opinion.
One of the problems with the way RMI put the worksheet together is that the data comes from numerous sources published in different years. RMI compares data from a 2003 MIT study, a 2007 MIT study, a 2006 one-page WADE source, and a 2005 “personal communications” data exchange (will explain below) just to name a few. Picking and choosing certain data points from many different sources just screams the word “cherry-picking.” As I’ll show below, that’s exactly what they did.
Cogeneration Cost Data
The RMI worksheet assumes that the delivered costs (aka levelized costs) of “combined-cycle industrial” and “building-scale” cogeneration plants (aka combined heat and power plants - CHP) are 5.47-5.91 cents per kWh. One component (O&M) of RMI’s levelized costs comes from WADE’s one-pager on gas turbines. What’s interesting is that the WADE one-pager provides a link to the International Energy Agency’s 2008 Combined Heat and Power paper (pdf). According to page 24 in the IEA report, the delivered electricity costs for an “Accelerated CHP” plant are above 10 cents/kWh, nearly twice as high as RMI’s costs. Why didn’t RMI use this data from the IEA paper considering IEA is a more reliable, objective source than themselves or WADE? RMI uses IEA data for one of the cost components of nuclear (O&M). But apparently IEA data on co-gen plants are not good enough for RMI’s comparisons. Looks like cherry-picking to me.
Moving on. The RMI worksheet found that “Recovered-heat industrial cogen” plants are the cheapest power plants in their dataset. I went through a maze trying to find the capital cost assumptions for this type of power plant. In RMI’s “Climate Data” worksheet, the source of the capital costs for this plant was a paper titled “Mighty Mice” (pdf) that Amory Lovins submitted to the Nuclear Engineering International magazine. I found no such mention of the capital costs on “Recovered heat industrial cogen” plants in this paper. Instead, a link “for documentation” in the section titled Comparative Cost sent me to RMI’s page on energy efficiency. This page had nothing to do with the capital costs of “Recovered heat industrial cogen” plants. I went to RMI’s methodology next.
According to the third page of RMI’s methodology (pdf), I can supposedly find the “cost breakdown” of “Recovered heat industrial cogen” plants in a previous RMI document (pdf). It looks like the breakdown is on page 22 in the paragraph on Cogeneration. All the data in this paragraph is based on “personal communications” with Tom Casten, Chairman and CEO of Primary Energy. No capital costs were mentioned in the paragraph, instead, only an “all-in electricity price” was given. To me, relying on “personal communications” from 2005 for cost data is just plain weak. Especially since it is proprietary and there’s no way for me to verify it.
So basically the spreadsheet and methodology pointed me to two different documents which provided no information on the capital costs for a “Recovered heat industrial cogen” plant. Not only that, the source of some of the data is a CEO. Apparently, RMI thinks it’s appropriate to use the info from the CEO of Primary Energy, but when NEI’s CEO says nuclear plants are competitive in a climate-constrained world, Amory Lovins complains that it’s false (as discussed on page 5 in RMI’s paper (pdf)). One word again comes to mind: cherry-picking.
Energy Efficiency Cost Data
RMI assumes, without references to any sources, that efficiency costs 1-4 cents/kWh. How can RMI claim these numbers without any sources?
Here’s what I’ve found when researching efficiency costs. According to the EIA’s 2006 Annual Energy Review, the costs of electric efficiency from utilities have remained between 3-4 cents/kWh since 1996 (adjusting to 2007 dollars, see graph below). The energy savings have also remained flat since 1996 as well. If utilities were to save more with efficiency, I could easily argue that it will cost much more than 3-4 cents/kWh.RMI’s claim that nuclear’s “cheaper competitors” produce “1.4-11+ more climate solution” is grossly exaggerated. Their “11+” number is based on the assumed cost of efficiency of one cent/kWh. Yet, according to EIA data, one cent/kWh is too low. I find it stunning that RMI advocates so much for efficiency, yet they provide no sources on the actual costs! I am not going to get into the RMI’s cost assumptions for coal, combined cycle gas, nuclear and wind because, as I’ll show below, RMI’s assumptions are irrelevant.
So do nuclear plants provide “more climate solution” per dollar than what RMI claims?
Yes they do, at least according to several electric utilities that are planning to build them.
Over the past several years, the capital costs of building different types of power plants have increased substantially. This trend is documented by Cambridge Energy Research Associate’s Power Capital Costs Index. Here’s their press release:
The IHS CERA PCCI – which tracks the costs of building coal, gas, wind and nuclear power plants indexed to the year 2000 – is a proprietary measure of project cost inflation similar in concept to the Consumer Price Index (CPI). The IHS CERA PCCI now registers 231 index points, indicating a power plant that cost $1 billion in 2000 would, on average, cost $2.31 billion today.The RMI paper only discusses CERA’s cost increases for new nuclear plants (from pages 7-10) and fails to note that the “supply constraints, increasing wages and rising materials costs” are affecting all types of power plants, including RMI’s co-generation plants. This trend is important to note because cost estimates older than a year ago are outdated. This further invalidates RMI’s cost estimates that are based on data from older studies.
“The fundamentals that have driven costs upward for the past eight years—supply constraints, increasing wages and rising materials costs—remain in place and will continue during 2008“ [said Candida Scott].
What are the latest new nuclear plant cost estimates then?
I can't give you exact new plant cost estimates because they vary among different sources. I can, however, provide the overall findings from several utilities who have made their own cost estimates.
In October 2007, Florida Power and Light submitted a Petition to Determine Need for Turkey Point Nuclear Units 6 and 7 (pdf) to Florida's Public Service Commission. Here's what page 11 states:
FPL’s analysis shows that for all of the scenarios evaluated (eight of nine), the addition of new nuclear capacity is economically superior versus the corresponding addition of new [combined cycle] units required to provide the same power output, yielding large direct economic benefits to customers as well as effectively addressing the criteria of section 403.519(4)(b). In fact, in the only scenario in which nuclear is not clearly superior, the natural gas prices are significantly lower than they are today and there are zero future economic compliance costs for CO2 emissions. Of all the scenarios evaluated, FPL believes these two to be the most unlikely. Moreover, even in these two unlikely scenarios, the results of the analysis show nuclear to be competitive or only slightly disadvantaged economically, while retaining the non-quantified advantages of fuel diversity, fuel supply reliability, and energy independence. Based on all the information available today, it is clearly desirable to take the steps and make the expenditures necessary to retain the option of new nuclear capacity coming on line in 2018.This is a pretty significant statement for FPL to find that a new nuclear plant is “economically superior” considering they are the largest owner of wind capacity in the U.S. and 42% of their generation comes from “state-of-the-art” combined-cycle gas plants.
What’s hilarious is that the RMI paper (on pages 6 and 7) used FPL’s high cost estimates to imply that nuclear plants are becoming even more uneconomical. Yet they neglected to mention FPL’s key findings on page 11 as well as the whole point of the filing. Here’s FPL’s page 2:
While FPL continues to advance reduced electricity usage and load management techniques through industry-leading conservation efforts and demand side management (“DSM”) programs, and actively cultivates and pursues the development of additional renewable generating capacity within the state, by themselves these efforts are not enough. FPL must also at times construct large, baseload capacity additions if the Company is to continue “keeping the lights on.” The proposed Project is intended to help meet FPL’s growing need for additional baseload capacity, which is the essential foundation of any utility’s supply portfolio, because these plants run year-round to provide the continuous supply of electricity that customers require. The Project also will enhance the reliability of FPL’s system by reducing reliance on fossil fuels and diversifying the resource mix.Failure to note these statements is another clear example of RMI’s cherry-picking. Here’s what Progress Energy Florida had to say in their Petition to Florida’s PSC (pdf) back in March 2008 for two new nuclear units (page 4):
PEF made its determination that Levy Units 1 and 2 were needed and the most cost-effective source of power to customers after fully accounting for the express considerations for nuclear power plant need determinations that the Florida Legislature set forth. Levy Units 1 and 2 will meet a reliability need in 2016 and beyond, while capturing cost-saving efficiencies and economies of scale from the successive construction of two nuclear power plants. Levy Units 1 and 2 will help the Company achieve greater fuel diversity and will enhance fuel supply reliability and security. The Levy units will avoid 864 million tons of carbon dioxide (“COz”), 1.4 million tons of NOx, 5.8 million tons of SO*, and 28,800 pounds of mercury over a sixty-year time frame and will, accordingly, position the Company to better respond to existing fossil fuel environmental regulations and future greenhouse gas (“GHG) regulations.And here is SCE&G’s application to the South Carolina PSC (pdf, docketed in May 2008) to build two new nuclear units at its Summer nuclear plant station (page 6):
SCE&G’s total net reliable generation capacity, including its two-thirds share of the output of the VCSNS Unit No. 1, is 5,687 MW, compared to a 2007 peak demand of 4,998 MW. The Company’s peak demand continues to increase and is presently forecasted to be 5,791 MW by 2016 and 6,133 MW by 2019. SCE&G can efficiently meet as much as 209 MW of this increased demand through conservation, load-shifting, off-system purchases, renewable energy resources or through the installation of gas-fired peaking units. However, without the additional base load capacity represented by the proposed Facilities, SCE&G will not be able to meet the increasing need for efficient base load power in its electric service territory and assure reliable, reasonably priced electric supply to its customers and the State of South Carolina.At least three utilities tasked with providing reliable power to their customers contradict RMI’s findings on nuclear plant costs. I’m sure the companies planning to build new nuclear plants in this list would also agree with the statements made by FPL, Progress and SCE&G.
The Energy Tribune sums it up best:
Lovins has a number of critics, and among the most prominent is Paul Joskow, a professor at the Massachusetts Institute of Technology. “My rule of thumb,” Joskow wrote me in an e-mail, “is to double his [non-nuclear] cost estimate and divide his energy saving estimate in half to get something closer to reality.”Pretty much.
Here are links to my previous posts for this series: Amory Lovins and His Nuclear Illusion – Intro, Amory Lovins and His Nuclear Illusion – Part One (The Art of Deception), Amory Lovins and His Nuclear Illusion – Part Two (Big Plants vs. Small Plants), and Amory Lovins and His Nuclear Illusion – Part Three (Energy Efficiency and “Negawatts”). I have two more posts left to publish – one on nuclear and grid reliability (also incorporating thoughts on decentralization) and then my overall conclusion of RMI’s paper. Stay tuned.