On August 20, 2014, Remy DeVoe, a graduate student in nuclear engineering at the University of Tennessee, published an earthshaking piece on ANS Nuclear Cafe titled Unintended Anti-Nuclear Consequences Lurking in the EPA Clean Power Plan. Unfortunately, there has been a bit of a delayed reaction; so far, only the most carefully tuned instruments have noticed any movement.
In his article, Remy described how he and Justin Knowles, another UT nuclear engineering graduate student, chose to spend part of their summer in the very “nuclear” activity of digging deeply into a mathematical model to find out the basic assumptions underlying its results. The model they chose was the formula that the EPA has described as a “consistent national formula” for calculating each state’s existing CO2 intensity (CO2 mass per MW-hr) — using 2012 data.
Aside: Please forgive my use of the underline format for a phrase that is not a link. Though underlines generally indicate the existence of a link here on Atomic Insights, I wanted to reproduce the phrase exactly as it appears — including text formatting — on EPA Connect: The Official Blog of EPA’s Leadership in a June 4, 2014 blog post by Janet McCabe titled Understanding State Goals under the Clean Power Plan End Aside.
Remy and Justin entered a hypothetical scenario into the model in which each state with nuclear generating stations shut down all of their nuclear plants and replaced their zero emission electricity generation with natural gas fired generation. Remy and Justin uncovered a starting fact. In 15 states the carbon intensity in pounds per MWhr as calculated by the EPA’s spreadsheet model decreased.
That result made no sense to them. They thought that someone had to have made a mistake somewhere in the formulas used in the model that had not been caught by any of the reviews that must have taken place inside the EPA before the rule was issued for comment.
They dug into the spreadsheet and realized that the formula for calculating each state’s initial emission intensity gave every power source except nuclear credit for 100% of its actual generation in 2012, which was the year used for calculating the starting point from which each state’s carbon intensity reduction plan will be required a certain percentage reduction by 2030. For nuclear electricity production, the EPA’s “consistent national formula” applied a factor of .058 to its 2012 generation and then that far smaller number was added to the rest of the state’s 2012 power output to compute the overall carbon intensity.
That means that states with a substantial portion of power being produced by zero emission nuclear energy today have what appears to be a much higher carbon intensity than they actually have. If they close one of their operating nuclear plants, their EPA calculated carbon intensity number increases by a very small number because only 5.8% of the plant’s zero emission output was in that number in the first place. If the state chooses to replace the output of the nuclear plant with natural gas, there is a possibility — dependent on the state’s mix of other plants — that the carbon intensity number will actually decrease from what it was with the nuclear plant running.
Remy and Justin then met with the people who devised the formula and showed them what they had found. Those EPA staff members did not admit an error; they simply explained their methodology and assumptions in creating the model. They encouraged the pair to provide a comment to the proposed rule and said it would be considered during the process of finalizing the rule.
As young graduate students studying hard to become professionals in a field where integrity is highly valued and reinforced, Remy and Justin have graciously projected their personal honesty onto the EPA decision makers and believe that the formula was the result of an inadvertent misunderstanding that has “unintended” consequences.
I’m a little more experienced in the ways of Washington staffs and quite a bit more skeptical about the actions taken under the cover of pleasant sounding phrases. There is only a small chance that the formula’s results are unintentional. I suspect that it was created with the conscious goal of minimizing the CO2 benefits of nuclear energy.
During my 33-year career as a US naval officer, I learned to trust, but verify, and to stop trusting the work of anyone who would not admit an error, especially when confronted with proof.
Here are links to the spreadsheets documenting Remy”s and Justin’s analytical work.
State Goal Data Calculations:
Plant Level Data Calculations:
These documents can be edited so you can view the equations. The pair has requested feedback.
It is possible that the person who made the decision to bury the policy choice deep inside of a spreadsheet is not in an executive leadership position, but it is unlikely that a lowly analyst made the decision without managerial or executive acceptance that it was the right thing to do.
If the economic performance of nuclear energy production in a state is already acceptably competitive, it is unlikely that any plant owner will decide to shut down a plant merely to gain a little credit in the CO2 intensity reduction effort. However, if the economics of the plant are marginal because of temporarily low natural gas prices or other factors, this perverse incentive might be one more factor incentivizing a premature closure decision.
On Tuesday, August 26, I recorded a phone conversation with Remy DeVoe, Justin Knowles and Lane Carasik, who was the leader of the Nuclear Engineering Student Delegation this summer. Please have a listen, get a little angry and take some effective action by producing your own comment.
This is a call to action. Don’t let the EPA get away with imposing this formula. Follow the process provided before October 16, 2014 deadline. This is a high interest proposed rule; there are already more than 1900 comments so attention to this specific issue might require a noticeable volume mentioning the same feature of the formula.
I accept the court decision that the EPA has the authority to regulate CO2 emissions, but that decision did not give the EPA the authority regulate badly. It seems unlikely that a court would support an agency that decided to issue a regulation with a known flaw resulting in the potential for rewarding an increase in the production of a pollutant.
Note: During the show, we briefly discussed the fact that nuclear energy capacity factors in 2012 were in a temporary valley and the lowest in the past 12 years. Here is the capacity factor history graph copied from NEI’s excellent nuclear energy statistics library.