In order to build and operate a nuclear plant in the United States, you must obtain a license from the Nuclear Regulatory Commission. This requires a multi-year process involving 100's of millions of dollars in paperwork. The Vogtle 3/4 license is a 782 page document that spells out in mind numbing detail the licensee's requirements. But it is a one way contract. The licensee must follow all the requirements precisely. The licensor can change those requirements at will.
Section 187 of the Atomic Energy Act makes this abundantly clear. Here it is in its entirety.
SEC. 187. MODIFICATION OF LICENSE.— The terms and conditions of all licenses shall be subject to amendment, revision, or modification, by reason of amendments of this Act or by reason of rules and regulations issued in accordance with the terms of this Act.
This ability to change the terms of the deal whenever the NRC feels like it is not hypothetical. The first major backfit was a new set of Emergency Core Cooling Systems rules which put Indian Point 1 out of business in 1974. This was followed by example after example. Here's just a few.
Sequoyah
A study by the General Accountability Office of the Sequoyah plant documented 23 changes ``where a structure or component had to be torn out and rebuilt or added because of required changes." The Sequoyah plant began construction in 1968, with a scheduled completion date of 1973 at a cost of $300 million. It actually went into operation in 1981 and cost $1700 million. This was a typical experience.
Yankee Rowe
Yankee Rowe was shut down 8 years early in 1992 on the basis of an unproven claim of loss of ductility. It's 40 year license was really a 32 year license.
AP1000
In 2006, the NRC certified the AP1000 design. On this basis, two utilities opted to build four AP1000 plants. In 2009, the NRC imposed a new aircraft strike rule which forced major changes in the design all the way down to the foundation.
Turkey Point
In February 2022, the NRC rescinded previously approved license extensions for the Turkey Point and Peach Bottom plants. The extensions had been approved after an 18 month long process costing the applicants tens of millions of dollars. To justify the reversal, the commissioners claimed the word ``initial" in the NRC regulations did not mean ``initial". There is no appeal from such arbitrary, inconsistent behavior.
Is there any rational investor who would play by these rules? In a normal market, the answer is no. Of course, not. But US nuclear power did not grow up in a normal market. Under the rules of a regulated utility market, capital cost increases imposed by changes in the NRC rules result in an increase in the rate base, and an increase in utility profits. The regulator covers his rear and expands his turf. The investor breaks even or better. The loser is the ratepayer, who had no say in the matter.
Along came deregulation. Deregulation was in part the response to the skyrocketing costs of nuclear power plants. The ratepayer was fed up with being the victim of this incestuous regulator-investor cost spiral. But in a deregulated market, no rational investor can play by NRC rules. There are only three choices:
1) A deregulated nuclear in a deregulated market. This will require a firm, automatic compensation scheme for radiation harm based on an individual's dose rate profile and nothing else. Underwriter Certification of Nuclear Power outlines how that would work. Properly implemented, this is the solution to the Gordian Knot.
2) A prohibitively expensive nuclear in a regulated market, the Vogtle model.
3) A socialized electricity grid run by a public monopoly, following the French model of the 1970's. That monopoly must be responsible for both providing electricity and nuclear safety. Otherwise this will simply be (2) on steroids as far as cost is concerned.
(2) is a non-starter. So we really only have two choices. Pretty simple.
Amazing to us that people don’t understand the tools we have or the monkey wrenching of over regulation fighting one of those tools (which happens to be the best for purpose).
Good piece, Jack!
This is my new favorite substack - thanks Jack!
Why not an option 4) a congressionally reformed NRC with a dual mandate that promulgates regulation based on the sigmoid threshold model you proposed?