Source-Term Assumption in a Design-Basis Accident Analysis
You are a licensing engineer reviewing a design-basis accident analysis for a small modular reactor. The applicant's safety analysis report uses a source term for a loss-of-coolant accident that assumes 100% of the volatile fission product inventory in the fuel-cladding gap is released, but assumes only 1% release from the fuel matrix for the same isotopes. The basis cites 'industry practice' and a 1990s study. The staff reviewer notes that the applicant's fuel has a higher burnup than the study's fuel and that the gap inventory fraction was measured on a different fuel design. The applicant argues the analysis is conservative because the gap release dominates and the matrix release is small. The review schedule has 30 days remaining before the licensing milestone.
1. Is the applicant's 'conservative' argument adequate? Explain.
Read the worked answer
Not necessarily. The claim that gap release dominates is only conservative if the gap inventory fraction is itself bounding for the applicant's fuel. The applicant used a fraction measured on a different fuel design and a study from the 1990s; higher burnup generally increases the gap inventory and can change the matrix release behavior. If the gap fraction is underestimated, the 'dominant' term is wrong and the conclusion fails. The 1% matrix release is a separate assumption that also needs a basis for the applicant's fuel. 'Industry practice' is not a technical justification; the basis must be traceable to data or models applicable to the specific fuel. The analysis may be conservative, but that must be demonstrated, not asserted.
2. What would you require from the applicant to close the review, and how do you handle the 30-day schedule?
Read the worked answer
Require: (1) a burnup-dependent gap inventory fraction for the applicant's fuel, with data or a validated model and uncertainty; (2) justification of the 1% matrix release for the applicant's fuel and burnup, including sensitivity to the assumption; (3) a sensitivity study showing that the conclusion (dose below the acceptance criterion) holds across the uncertainty range of both assumptions; (4) identification of the conservative bounding case and confirmation that the analysis uses it. On schedule: the 30-day milestone does not justify accepting an unsupported basis. I would issue a request for additional information with a clear response date, document the schedule risk, and escalate to management so the milestone decision is made with full knowledge. If the applicant cannot respond in time, the correct outcome is a schedule slip, not a licensing decision on an unverified basis.
3. The applicant offers a verbal commitment to 'address it in a future revision.' How do you respond?
Read the worked answer
A verbal commitment to fix it later is not a basis for a safety finding now. The analysis as submitted must support the licensing decision; a future revision is not part of the current basis. I would decline to close the review on that basis, document the offer and my response, and require either a written commitment with a binding schedule and an interim conservative assumption that is demonstrated bounding, or a revised submittal before the milestone. If management directs acceptance, I would document my technical objection in writing and ensure it is part of the record. The professional obligation is to the safety basis, and 'we'll fix it later' is a classic pathway to an unresolved safety issue.