In shaping a picture of the NRC for Congress, the General Accounting Office, in its 1997 report: GAO/RCED-97-145, May 30, 1997, stated: Determining the safety of plants is difficult because NRC does not precisely define it. Instead, NRC presumes that nuclear plants are safe if they operate within their approved designs (design basis) and meet NRC's regulations. However, NRC's regulations and other guidance do not provide either the licensees or the public with the specific definitions and conditions that define the safety of a plant. As a result, NRC does not have an effective way to quantify the safety of plants that deviate from their approved designs or violate regulations. Determining a plant's safety condition is, therefore, a subjective judgment.
The GAO's 1997 statement about NRC and the definition of safety is off the mark, because it ignores the very basis of safety in the nuclear industry.
Rather than embody safety in the agency, or in a single static model imposed upon 103 plants, the licensing system has imposed a vast creative task individually on each licensee, prior to the granting of each license. The task involves the writing of a detailed Safety Analysis Report. The definition of safety, for that plant, is embodied in the Safety Analysis Report, a huge document running to ten or more volumes, with internal references to thousands of calculations stored elsewhere, and hundreds (if not thousands) of detailed design drawings, also stored elsewhere.It is known as the SAR, or (final) FSAR. So, defacto, the engineering designers who wrote the FSARs and the technical specifications have also written the safety standards, 103 differing safety standards for 103 plants. Each standard is extremely concrete, there is no vagueness. However each standard is huge, and there are 103 separate versions.
This kind of a document cannot be inspected casually, or cursorily by GAO, nor can operator compliance with it be casually determined by observing the NRC. To wrongly imply that NRC methods were lax, or "subjective" is a misleading and self serving statement, designed to lift the onus of comprehending the FSAR system, off the shoulders of the GAO team, and deposit GAO's unreadiness to prepare its inspectors onto NRC's doorstep, as a vague accusation of "subjectivity". With 103 versions of law, residing in 103 FSARs at 103 Nuclear plants, the amount and difficulty of material is just too great for GAO to assess, much less sum up. GAO failed to adequately comprehend this system, and wrongly reported it to Congress as an NRC shortfall. It is in point of fact, the defacto status of present regulatory law. As law, as a sitting legal structure, it ought not be mischaracterized as an administrative shortfall.
Perhaps if it understood its intended mission more completely, GAO might have proposed a new legal structure, complete with a general unified FSAR, but of course, it lacked the technical competency to even determine the nature of what it was assessing, and so could not have successfully replaced it with a more comprehensive upgrade. As it is, GAO has shuffled its feet unknowingly, at the periphery, accusing NRC of not safeguarding the public, when in fact it was GAO failing its mission, the mission to understand just where the concrete jot and tittle of nuclear safety was embodied--- in the FSARs and the tech specs, and not within NRC. This kind of a safety standard demands the dedication of a qualified set of resident inspectors, tasked with climbing the extremely steep learning curve in each FSAR, as a preparation for understanding how each individual plant is fulfilling its specific commitments to each FSAR. Once the subject matter is mastered, then the individual inspector, be he an NRC resident inspector, or a GAO inspector, can be ready to realistically compare plant performance parameters to the mammonth compendium of promised performance parameters, that is the FASR and The Technical Specifications. With such knowlege in hand, the judgement is not subjective. It is extremely objective. Meet tech specs=pass. Not meet tech specs= fail.
This is the American system. If it is monumentally complex, and thus not amenable to easy GAO mastery, that fact just makes any casual GAO suggestions made after a cursory look-see a lot less than enlightening. Therefore must Congress remain in the dark, and simply trust NRC? Perhaps, but better that they understand their own inpectors' blind spots. Therefore allow me to analyze the GAO assertion,line by line, in the light of what I've revealed above.
Determining the safety of plants is difficult because NRC does not precisely define it.
This is not true. Determining plant safety is difficult, because it is precisely defined 103 separate ways in 103 FSARs, and because each FSAR , with its accompanying references may take a year or more for a talented individual to comprehend.
NRC presumes that nuclear plants are safe if they operate within their approved designs and meet NRC's regulations.
This is true, but is not a shortfall. The vast system of redundant safeguards embodied in each FSAR provides large margins of safety, and its initial approval came only after detailed critical evaluation to the best scientific/engineering standards. Such trust is not ill-founded trust.
However, NRC's regulations and other guidance do not provide either the licensees or the public with the specific definitions and conditions that define the safety of a plant.
This is not true. The Technical specifications provide an absolutely precise and objective standard to the licensees, and to NRC. Perhaps GAO is suggesting a tech spec primer series be prepared for public consumption along the lines of "A nuclear plant is safe, when its tech specs are met", with explanations. I doubt if the public would be interested. GAO, on a mission to find the tech specs, missed them entirely, and now it reports that NRC has none. Would the public do any better?
As a result, NRC does not have an effective way to quantify the safety of plants that deviate from their approved designs or violate regulations.
This might have been germaine in 1997. In 2007 it is not true. The Reactor Oversight Process is now in place, giving objective banded scoring to each plant, in all major areas. Note that this statement is not about safety per se, but rather it is about the reporting of safety conditions to the general public.
Determining a plant's safety condition is, therefore, a subjective judgment.
This was not true in 1997, and it is most certainly nonsense in 2007.
In missing the absolute inflexibility of the tech specs, and by looking in the wrong place for the exactitude (looking within NRC, rather than in the license), GAO overlooked the very concrete methodology for maintaining plant safety, as being non-existent. This failure has now propagated itself outward through the Congress, and the public, as a tribal myth, wrongly accusing NRC of laxity.
