External Hazards

Purpose

Taking External Hazards into account is an important part of quantifying the reliability of advanced reactor designs. While extreme external hazards can be very rare, it is important that they are still considered as one extreme external event can quickly lead to total plant failure, as seen in the Fukashima Daiichi accident.

Scope

External events usually encompass a variety of weather and geographical phenomenon, such as seismic events, flooding events, and high winds, but can also be related to non-geographical factors such as proximity to another plant. When considering external hazards, it is important that no stone goes unturned in identifying possible failure scenarios.

External Hazards

New Research

Collocated Industrial Facilities

Colocation of nuclear power plants with industrial facilities may introduce new external hazards, such as combustion loads and heat or electrical load rejection from the end user. In existing assessments, potential events have been identified and discussed from a very specific focus, such as risk from electrolytic hydrogen production and storage, but no overarching assessment has been performed that covers a comprehensive set of interactions.

The objective of future research regarding collocated industrial facilities is to perform a comprehensive assessment of potential effects on the technical elements of a nuclear plant PRA and assess the potential changes in magnitude, frequency, or consequences of accident scenarios affected by the type of collocated industrial facility. The assessment plans to select a representative set of advanced reactor designs and industrial applications as the basis for the evaluation to produce insights for a broad range of reactor designs and industrial facilities.

Rare Seismic Events

While the Licensing Modernization Project (LMP) aims to provide flexibility and cost efficiency through performance-based criteria, its frequency-consequence (F-C) target of 5E-7/yr for high-consequence Beyond Design Basis Events (BDBEs) can lead to overly conservative seismic requirements for structural design. Analyses show that meeting this target would require increasing design-basis ground motions by up to three times current U.S. Nuclear Regulatory Commission (USNRC) criteria, significantly raising costs without proportional safety benefits.

The following path is recommend for RIPB seismic design of SSCs in advanced reactors, consistent with LMP from NEI 18-04 [8] and based on EPRI 3002032031 [72]:

• For safety-related SSCs, the DBHL earthquake need not exceed the GMRS defined according to RG 1.208.
• High-consequence seismic BDBEs are allowed to exceed the F-C target provided that the union of these BDBEs demonstrates a seismic margin (HCLPF capacity) equal to at least 1.67 times the GMRS; and the cumulative risk metrics acceptance criteria or USNRC-endorsed surrogates (such as being developed in ANL/NSE-24/42 [9]) are met.
• The seismic design basis for non-safety related SSCs may follow commercial design codes and standards with special treatment as warranted to risk-significant SSCs.

Historical

One of the main conclusions of TECDOC 1487 [42] is that “the development of an external event PRA in parallel with the early plant design may help to identify the vulnerabilities as well as potentially overly conservative design features at an early stage, leading to a well-balanced and cost-effective improvement in safety”. Therefore a core issue that needs to be clearly addressed moving forward is the balance between the cost and benefit of developing detailed External Events PRA models with insights that can be used to adequately and robustly screen external events that may otherwise be expected to be fully included (e.g., is such an approach feasible, justifiable; can a graded approach be provided that accounts for increased margin to safety goals with sufficient consideration of uncertainty and degree of confidence). In relation to the section describing passive system reliability, a large part of determining the uncertainty of a passive system is being able to determine the risk associated with certain external events and how it relates to the designed passive safety systems. For example, external events have the potential to change assumed system boundary conditions which could result in the failure of the passive system to successfully perform its intended safety function.

The impacts of external hazards will likely provide a substantially greater fraction of the contribution to risk to the public than is the experience from the current fleet of water reactors. Because the current state of knowledge of potentially dominant very low frequency external hazards (phenomenology, event occurrence frequencies) is less developed than for internal events, the larger fractional contribution to overall risk will have a greater level of uncertainty. This situation also may require modifications to processes and criteria that have been used for the existing fleet of plants to assess and screen external hazards from consideration as providing a negligible contribution to PRA results. Some external events may be screened out through a deterministic and probabilistic process.

Research Roadmap (EPRI 3002026495) actions supported for passive system reliability include:

• Develop Enhancements to Licensing Process
• Establish Decoupling Framework for Nuclear Beyond Electricity (NBE) Users
• Demonstrate Risk-Informed and Performance Based Approach

A higher level of detail for these actions can be found in the report.