10 CFR 50.59 – A history of the rule’s development

50.59 - a history of its evolution from 1961 to today

Prior to 1962, the 50.59 process for modifying a nuclear plant without requiring prior Commission approval did not exist. There was no UFSAR. There were only Technical Specifications, the Hazards Summary Report, and license amendments. This is a summary of the 50.59 rule’s evolution, from 1961 to today…

In the beginning of the commercial nuclear power age (which dates to 1956, when the first 10 CFR 50 regulations appeared in the Federal Register), the 50.59 process did not exist. The 50.59 process is how a nuclear plant determines if the NRC’s permission is required to make changes to a plant.

Many plant design changes do not require NRC approval to be implemented. They can be trivial and obviously have no impact on a plant’s ability to maintain the health and safety of the public. Or they can be substantial, but also be shown to not adversely affect the plant’s ability to protect the public.

The 50.59 process is how a plant determines whether prior NRC approval is required before it can implement a particular design change. It is located in Title 10, Part 50.59 titled “Changes, tests, and experiments.” Before 50.59, all plant design changes had to be approved by the Atomic Energy Commission (AEC), no matter how trivial.

The crucial summary text from the regulation states:

(c)(1) A licensee may make changes in the facility as described in the final safety analysis report (as updated), make changes in the procedures as described in the final safety analysis report (as updated), and conduct tests or experiments not described in the final safety analysis report (as updated) without obtaining a license amendment pursuant to §50.90 only if:

(i) A change to the technical specifications incorporated in the license is not required, and

(ii) The change, test, or experiment does not meet any of the criteria in paragraph (c)(2) of this section.

There are eight criteria which then follow, against which changes are evaluated. Those criteria determine whether a license amendment must be requested. But the rule has not always been so clear. It started out looking very different.

The History of the 10 CFR 50.59 Rule

The Atomic Energy Act of 1954 made it permissible for private companies to harness atomic energy to construct and operate nuclear power plants. The first rules and regulations governing commercial nuclear power operation appeared on January 19, 1956, in the Federal Register (Volume 21, Number 12).

About five years later, on April 8, 1961 (Federal Register, Volume 26, Number 67, Saturday, April 8), the AEC, under the chairmanship of Glenn T. Seaborg, published the draft 10 CFR 50.59 rule.

Seaborg’s major accomplishment had been leading the discovery of plutonium in 1941 and developing the production process for its use in atomic weapons during the Manhattan Project. A signer of the Franck Report, and like former AEC chairman Lewis Strauss, Seaborg had wanted the atomic bomb’s power to be displayed in an uninhabited area of Japan to convince the Japanese to surrender peacefully, without suffering a catastrophic loss of life.

Subsequently, he desired to see atomic energy harnessed for peaceful purposes, and the nuclear power industry expanded significantly in the decade that he chaired the AEC.

No doubt that the issuance of the 50.59 rule played a part in the industry’s advance, since the rule made it easier for plants to make modifications and upgrades that did not impact public health and safety. It reduced the number of reviews and approvals the AEC had to conduct, freeing up their time to approve the numerous licenses coming in for new plants.

In an NRC report that provides a short history of the 50.59 process, we are told exactly how bad things were before 50.59: “In 1960, the Commission recognized that the combination of incorporating the entire hazards summary report into a license as TS, together with the mandatory hearing requirement for all amendments, whether any person requested a hearing or not, no matter how minor, was a burdensome, unworkable process for controlling license requirements.”

The 1962 50.59 Rule

From the release of its first draft for public comment, it would then take about 14 months for the draft rule to be subjected to public scrutiny, evaluated, and revised. The final 50.59 rule was issued on June 9, 1962 (Federal Register, Volume 27, Number 112, Saturday, June 9).

The rule made three main changes to the 10 CFR 50 regulation:

  1. It revised Part 50.36 (on the designation of technical specifications);
  2. It added in new Part 50.59 in its entirety; and
  3. It created Appendix A to 50.59, which was a guide to the content expected to be contained in a plant’s technical specifications.

The rules on technical specifications needed revision to better fit and accommodate the new 50.59 philosophy. Previously, the rule stated that the Commission would identify which parts of the plant’s hazards summary report “will be deemed to be technical specifications that become part of the license or construction permit.” There was little formal guidance on the content of tech specs or which parts of the hazards report might become the tech specs.

But the change philosophy introduced by 50.59 placed the technical specifications at the heart of the process. Whether a change impacted them played a major role in determining whether a plant could implement the change without first being granted authorization by the Commission.

Since new plants were under construction, and soon even more would be under construction, their owners would need to incorporate the latest guidance into their plans. The tech-spec-related rule changes were intended to communicate the expectations to future applicants and also establish universal objective criteria to make the 50.59 change process more predictable and consistent across the industry.

The Main Differences Between The Proposed Rules and Final Rules

One of the major differences between the draft 50.59 rule and the final rule was the drastic reduction in reporting requirements.

This would greatly alleviate the administrative burden on both the plants and the regulator. The draft rule would have made a plant “promptly” file “a report of each change, test or experiment carried out” in accordance with the 50.59 process. So, even though the AEC’s permission was not required (meaning a license amendment did not need to be issued), the plants would still have to immediately report each change to the regulator that “screened out” of the 50.59 process (to use the modern parlance).

The language and requirement in the final rule was significantly softened. The final version of the 50.59 rule required only that the plants keep records of changes in the facility, procedures, or any tests and experiments carried out according to 50.59, and then submit to the AEC an annual report that summarized them. This is quite a reduction in effort compared to “promptly” submitting reports detailing every change made.

In both the draft and final rule, however, the substance and intent were largely the same. Some clarifications were made, but here’s the final version of 50.59(a) from 1962:

The holder of a license authorizing construction or operation of a production or utilization facility may (1) make changes in the facility as described in the hazards summary report, (2) make changes in the procedures as described in the hazards summary report, and (3) conduct tests or experiments not described in the hazards summary report, unless the proposed change, test or experiment involves a change in the technical specifications incorporated in the license or an unreviewed safety question, as defined in paragraph (c) of this section. If the proposed change, test or experiment involves a change in the technical specifications or an unreviewed safety question, it shall not be carried out unless authorized by the Commission pursuant to the procedures set forth in this section.

10 CFR 50.59(a), from Volume 27, Number 112 of the Federal Register

Note that this paragraph is just two sentences–bureaucratic writing at its finest!

The result was that plants were allowed to modify the facility, revise procedures, or conduct certain tests or experiments, but only if:

  1. The change did not modify the technical specifications or
  2. Involve an unreviewed safety question.

The Unreviewed Safety Question

The regulations introduced an important concept–that of the “unreviewed safety question.”

The AEC recognized that while many changes could be safely implemented without prior Commission review and approval, some could not. Some changes would not impact the tech specs but still involve uncertainty, especially at this early stage in the development of such a modern and potentially dangerous technology. For changes of this nature, the AEC would need to get involved. It would need to explicitly grant permission for a plant to proceed.

The 50.59 regulation basically drew a boundary around the core technical specifications and certain aspects of a plant’s hazards analysis that were crucial for ensuring that “the health and safety of the public will not be endangered.” These elements were at the heart of a plant’s operating license, and the AEC roped off these parameters and hung a “No Trespassing” sign around them. “Thus far, and no further,” was the regulator’s commandment. “At least, not without first being granted our permission to proceed.”

Three 50.59 Criteria

Besides an obvious encroachment upon the technical specifications, the AEC rulemakers established three criteria to help the utility determine if they were proposing a change that “involved an unreviewed safety question.” The three criteria will sound vaguely familiar to anyone who has done any 50.59 work of their own. The rule stated that a proposed change, test, or experiment shall be deemed to involve an unreviewed safety question if:

  1. The probability of occurrence of an accident previously analyzed in the hazards summary report may be increased; or
  2. If consequences of an accident previously analyzed in the hazards summary report may be increased; or
  3. If a possibility for a nuclear accident of a different type than any analyzed in the hazards summary report may be created.

The rule required that any change involving an unreviewed safety question be authorized by the Commission. The utility (licensee) had to “file a request for authorization of a change in technical specifications or of any change, test or experiment” that did not screen out of the three criteria.

The plant had to submit a hazards analysis with their request (essentially a “license amendment request”). The Commission would then either review the hazards analysis and decide the plant could proceed, or else they would refer it to the ACRS and hold a public hearing on the matter.

The key criteria provided for deciding whether to refer it to the Advisory Committee on Reactor Safeguards (ACRS) was if the Commission could determine “that the proposed change, test or experiment presents significant hazards considerations not described or implicit in the hazards summary report.” If it did not, then the Commission may authorize the change without submitting it to the ACRS and without a prior public hearing.

Lastly, the initial 50.59 rule attached a guide, called Appendix A, to help illustrate what the AEC expected to see in the technical specifications for nuclear reactors.

Becoming More Modern – The Major 50.59 Revision of 1968

By the mid-1960s, reactor applications were pouring in, and the AEC was overwhelmed. It was during this time that it initiated a proposed revision to the 50.59 rules to streamline the process.

In 1968, the AEC acknowledged that the level of detail contained in the Technical Specifications was unnecessary for ensuring public safety. It was also contributing to the regulatory bottleneck.

It was at this point that certain concepts came into existence that remain with us to this day. The 50.59 philosophy began to more clearly crystalize around allowing changes as long as they preserved the safety margins built into a plant’s design basis, its safety-related accident-mitigation systems, and the consequence analyses associated with its design.

With the 1968 revision, two general changes in terminology were introduced. First, the term hazards summary report was replaced with safety analysis report based on previous rulemaking. The Commission felt the term better characterized the document. (See page 832 of the January 21, 1966 issue of the Federal Register.)

Also, the phrase “systems, structures, and components,” or SSCs, was introduced. It evolved during the period between the proposed rules in 1966 and the final rule published at the end of 1968. It replaced the somewhat awkward phrase used in the proposed rule, “systems and major components of the facility.”

The term “Design Bases” was added to the definitions section in 50.2. Several changes were made to the licensing (50.34) and Technical Specifications (50.36) sections. These changes were intended to work together with the more streamlined 50.59 process envisioned by the AEC.

Section 50.34 was heavily revised to introduce the concepts of a PSAR (preliminary safety analysis report) during the construction stage and the FSAR (final safety analysis report) during the operating license application stage. The content of each was spelled out in greater detail. Emphasis was placed on the need to include analysis, evaluation, and identification of the proposed tech specs during the construction stage.

The concept of a quality assurance program was introduced. A license application would have to include within its PSAR a description and evaluation of the QA program which would control “the design, fabrication, construction, and testing of the structures, systems and components of the facility.”

Analysis Emphasized In Tech Specs

The Tech Spec (TS) requirements received a major overhaul. Now, TS bases would need to be included, though they would not become a part of the TS.

Second, the TS requirements were revised to emphasize that the Tech Specs would be derived from the analyses and evaluation included in the SAR. Previously, the TS were those “significant design features, operating procedures, and operating limitations” that were “considered” important for preserving public health and safety.

If it was unclear before, or in case others interpreted differently what “consideration” meant, the requirement was now made clear that the TS were to be derived from analyses. Rigorous evaluation would be necessary. Math and analyses would be relied upon to prove which features and parameters were important, as opposed to other, perhaps more subjective, methods.

Also in Section 50.36, key familiar concepts were introduced, like safety limits, limiting conditions for operation, and surveillance requirements. Unlike the TS bases, these were to become part of the TS.

Interestingly enough, the term “limiting conditions for operation” began life as “minimum conditions for operation” when initially proposed in 1966. Along with it were “maximum” safety system settings. Over the next 28 months, during the period of public comment and resolution, the AEC was persuaded to generalize these concepts and replace “minimum” and “maximum” with the ambiguous “limiting”.

Section a of 50.59 was preserved without change, but Section b was revised to clarify the content that the plants needed to include in their record of changes implemented without prior Commission approval. It became a requirement to document the reasoning behind their determination that a change did not involve an unreviewed safety question. A summary of these safety evaluations also had to be included in their periodic reports to the Commission.

Previously, the rule required only that a plant maintain records of the changes and submit a brief description of each in the periodic summary report. Documenting their reasoning was not a requirement. This changed in 1968.

Next, the concept of the “unreviewed safety question” was refined.

Updating the “Unreviewed Safety Question”

First, malfunctions of equipment important to safety were incorporated into the probability, consequences, and new-accident criteria of the unreviewed safety question. Also significantly, the definition was expanded to include the new concept of “safety limits” added to the TS. Now, reducing the safety margins as defined in the TS bases would involve an unreviewed safety question.

At this point, several of the eight modern 50.59 criteria were now present, but largely in condensed form as compound sentences.

Lastly, the 1968 revision deleted Appendix A and made the involvement of the ACRS permissible instead of mandatory if a change presents significant hazards considerations not described or implicit in the SAR. This would further reduce administrative burden when processing the multitude of plant changes that would inevitably continue to roll in.

All of these changes together were designed to ultimately give plants greater freedom to carry out plant modifications that did not encroach upon the nuclear plant’s ability to maintain the health and safety of the public. Beginning with the PSAR, for example, plant designers were cautioned to focus on what may become the technical specifications in order to expedite the licensing process and reduce long-term operating burdens imposed by design changes.

Having well-defined, lean tech specs and a uniform, rigorous analytical process by which they were determined was intended to streamline the licensing of the numerous plants under construction and also keep an eye towards the future. Having focused tech specs would prevent unnecessary future licensing reviews because routine or trivial design changes would no longer encroach upon bloated tech specs that needlessly contained more detail than was necessary to maintain public safety.

Other additions to Part 50 were made in the following years, including the attachment of the general design criteria of Appendix A and the quality assurance criteria of Appendix B. But these 50.59 rules were largely set in stone for the next three decades.

Three years later, in 1971, after a decade of service, Glenn T. Seaborg’s chairmanship of the Atomic Energy Commission ended.

Three Decades Later, A Major 50.59 Revision Is Proposed in 1998

The proposed 50.59 rule revision of 1998 contains substantially more text than the rules published in the old era. Much of that text is background and explanatory, which can be helpful for understanding the intent behind the revisions.

One major industry change that occurred over the intervening 30 years was that the Atomic Energy Commission was abolished and replaced with the Nuclear Regulatory Commission. This happened with the Energy Reorganization Act of 1974. The AEC’s responsibility for regulating the public health and safety of nuclear reactors was transferred to the NRC, which began operation in January of 1975. Its remaining responsibility was split across other regulatory bodies.

Vanquishing the Unreviewed Safety Question

After a relatively brief period of public comment and review (less than a year), the final rule was published in October of 1999.

The first innovation the NRC made to the 50.59 rule was to remove all references to the “unreviewed safety question,” or USQ. They felt that it created confusion around the purpose of 50.59. The purpose of 50.59 is to determine whether a plant needs to request a license amendment before implementing a change, but the NRC concluded that the concept of the USQ was leading some plants to make unauthorized changes because they concluded they were safe:

The Commission proposes to remove the reference in the rule to the term “unreviewed safety question” and instead to refer to the need to obtain a license amendment. The Commission believes that the terminology of “USQ” has sometimes led to confusion about the purpose of the evaluation required by 50.59. Some licensees have concluded that if they determined a change was safe, there could be no need for NRC approval.

Federal Register, Volume 63, No. 203, page 56,101

The problem is that a change that poses no safety hazard can still violate the current licensing basis, triggering the need to request a license amendment.

Defining Terms

A number of terms were given formal definitions to further reduce confusion. The definitions were:

  • Change – Public comments helped shaped the proposed definition into its final form, which incorporated qualifications to prevent requiring full 50.59 evaluations for trivial additions to the facility or to a procedure that would have no impact on safety. This facilitated the use of 50.59 screenings, which allows a plant to show that there is no adverse impact to any design function, thereby allowing them skip the full evaluation.
  • Facility as described in the FSAR (as updated) – Added to emphasize that the “facility” is not only just the physical systems, structures, and components (SSCs) of a plant, but also their underlying design requirements and evaluations or methods of evaluation that prove they’ll perform as intended. In other words, the facility also consists of design information.
  • Procedures as described in the safety analysis report – emphasizes that how SSCs are controlled is also crucial to their design and licensing basis.
  • Tests and experiments not described in the final safety analysis report (as updated) – intended to convey that activities that potentially place a facility outside its licensing basis must be evaluated against the 50.59 criteria before being performed.
  • Final safety analysis report (as updated) – The definition used the parenthetical phrase, “as updated,” instead of saying “updated FSAR,” to convey the requirement that “evaluations need to account for changes made through other processes that have not yet been included in an update to the FSAR.” This is because non-power reactors are not required to submit updates to their FSAR. The language also covers young plants that have not yet made it to their first required FSAR update. Any new change needs to be evaluated against the latest design information–not just the latest formally published design information–in order to be accurate.

Introducing the 50.59 Minimal Increase Principle

The prior version of the rule required Commission approval if the probability of an accident or a malfunction may be increased by a change. This was too stringent and subjective and created an onerous burden on the industry.

So the Commission revised the rule to allow the probabilities to increase minimally. Now, a change could be implemented without prior approval if it would not “result in more than a minimal increase.” Minimal was generally defined as less than or equal to 10 percent of the difference between the current calculated value and the regulatory guideline value.

Previously, if there were any uncertainty at all over whether a probability increased, then prior approval was required. This was a likely scenario when dealing with trivial probabilities. The new rule allowed for some increase and set a limit on how much increase was permissible before Commission approval was required.

However, in their explanatory material, the NRC emphasized that “no more than minimal” is still a qualitative standard, so quantitative analyses are not required to demonstrate conformance–though they may be offered.

To drive home this point, the NRC substituted “likelihood” for “probability” because “The Commission concludes that the word ‘likelihood’ is more generally understood to represent qualitative judgments.”

Defining the eight 50.59 criteria

To enhance clarity, the NRC separated the multiple criteria that were bound up within compound statements. They broke them out into one sentence per criteria. The first of these compound statements became criteria i through iv.

The same kind of simplifying change was made to the second compound sentence also, forming criteria v and vi.

The third criteria, “margin of safety as defined in the basis for any technical specification is reduced,” had often been interpreted different ways without universal agreement, so it was overhauled completely. In its place was put new criterion vii, which establishes the design basis limits for fission product barriers as a hard limit that cannot be exceeded without prior approval. Combined with new criterion viii, the NRC believed these two criteria together protect the safety margins included in the design limits and in the margins included in the design methods that demonstrate that requirements are met.

Criterion viii, identifying departures from an evaluation methodology used to establish the design bases or as part of the safety analysis, had not been present prior to the 1998 proposal. Along with the definition of facility that was included, it eliminated an area of confusion which had arisen over three decades of industry operation and change.

The confusion stemmed from lack of clarity over whether design standards, specifications, and analysis methods were considered a part of the facility as described in the SAR. They aren’t physical structures, systems, or components (SSCs). Some plants concluded that SSC design information and technical requirements were not considered part of the plant, meaning they could be changed without prior NRC approval.

The NRC concluded differently, however. They recognized–accurately–that changing the basis of a facility’s design now could result in future physical changes that really did warrant NRC review and approval. If the underlying design standards could be changed, future changes may not be seen as adverse to public safety.

The NRC believed this kind of change would unknowingly erode the safety margins built into the facility that they wanted to protect.

This is why criterion eight was included with the 1998 proposed rule: to clarify that evaluation and analysis methodologies are, indeed, part of the facility. No built-in conservatisms could be reduced without prior approval.


NEI 96-07 was developed in tandem with the 50.59 rule revisions of 1999. In fact, the NEI submitted several comments that helped formulate the final version of the rules.

The 50.59 rules have received minor revisions since 1999 (to correct omissions and improve the licensing process for future reactors). But generally, the rules and 50.59 process (informed by NEI 96-07) have remained in their present form for over 20 years.

The 50.59 rule first appeared in 1962 under the AEC chairmanship of Glenn T. Seaborg. It introduced the idea of protecting the Technical Specifications from any changes whatsoever unless approved by the Commission through the license amendment process. It also introduced the concept of the unreviewed safety question and provided three criteria for determining whether one existed. If a change introduced an unreviewed safety question, as evidenced by meeting any of the criteria (such as by introducing accidents of a new type, etc.), then Commission approval was needed to implement the change.

The 50.59 rule received a significant overhaul in 1968. It was then that several more long-term concepts took shape, such as safety margin, limiting conditions for operation, surveillance, safety limits, and the UFSAR. The criteria for determining if a USQ was introduced were expanded, and the and from there the rule remained intact for the next 30 years.

The revision of 1999 changed the rule into the modern form in which it exists today. It eliminated the concept of the “unreviewed safety question,” separated the compound criteria into eight distinct criteria, emphasized that design information is included in the concept of the facility, and clarified that 50.59 design changes must protect safety margin by preserving the conservatisms built into calculation methodologies used in the safety analysis.

The published rules omit the commentary included in the Federal Register. This commentary provides valuable insight into the purpose and intent of the written rules. It can aid anyone who is responsible for developing 50.59 screenings and 50.59 evaluations.