How do I ensure my reactor core modeling with simulation software is compliant with traceable assumption requirements?

To make your reactor core modeling traceably compliant with assumption requirements, you want every assumption in the simulation to be:

1. Explicitly stated
2. Justified
3. Linked to a source or basis
4. Version-controlled
5. Tested for impact
6. Reviewed and approved

A practical way to do this is to build an assumption traceability chain from requirements all the way to model output.

1) Start with the governing requirement set

Identify the exact standards, procedures, and licensing basis that apply to your work, such as:

• regulatory requirements
• internal QA procedures
• code of conduct or licensing basis documents
• software validation and verification procedures
• model development standards

For each requirement, determine:

• what it demands
• whether it applies to your model
• how compliance will be demonstrated

2) Create an Assumption Register

Maintain a controlled log for every assumption used in the model. Each entry should include:

• Assumption ID
• Statement of the assumption
• Reason for the assumption
• Source/basis
  • experimental data
  • design document
  • prior validated model
  • engineering judgment
  • expert elicitation
• Applicable requirement
• Model location
  • equation, boundary condition, material property, geometry, control logic, etc.
• Impact assessment
  • what result or safety margin it affects
• Validation status
  • confirmed, conservative, unverified, pending
• Sensitivity ranking
• Reviewer/approver
• Date and version

3) Trace each assumption to a defensible basis

A compliant assumption should never be “because it seemed reasonable.” Instead, tie it to one of the following:

• measured plant or test data
• benchmarked experimental evidence
• accepted correlations or codes
• vendor documentation
• prior qualified analyses
• conservative bounding assumptions supported by logic

If an assumption is based on expert judgment, document:

• who made the judgment
• their qualifications
• why no better basis existed
• why the assumption is conservative or acceptable

4) Link assumptions to model inputs and outputs

Show where each assumption enters the model:

• input parameter
• boundary condition
• mesh simplification
• fuel performance correlation
• thermal-hydraulic coupling
• reactivity feedback term
• uncertainty distribution

Then identify which outputs it influences:

• peak fuel temperature
• cladding stress
• neutron flux
• power distribution
• reactivity margins
• critical heat flux margin
• burnup predictions

This creates traceability from requirement → assumption → model implementation → result.

5) Perform sensitivity and uncertainty analysis

Traceable assumptions are not enough by themselves; you also need to show they are not hidden failure points.

For each important assumption:

• vary it over a credible range
• quantify its effect on key outputs
• identify whether it is dominant, minor, or negligible
• compare against acceptance criteria

If an assumption has large influence, consider:

• refining it with better data
• bounding it conservatively
• adding uncertainty treatment
• explicitly limiting the model’s domain of applicability

6) Control the software and model version

Use configuration management for:

• software version
• solver settings
• material libraries
• geometry files
• input decks
• scripts
• post-processing tools
• assumption register version

Every report should state exactly:

• software name and version
• model revision
• input deck checksum or controlled file ID
• date run
• analyst
• reviewer

7) Verify and validate the model

To support assumptions, document:

• Verification: the model is implemented correctly
• Validation: the model is adequate for the intended use

Useful evidence includes:

• unit tests
• benchmark comparisons
• regression tests
• comparison to experimental or plant data
• documented acceptance criteria

If the software has built-in models or correlations, ensure those are also validated for your application range.

8) Document limitations and applicability

Every assumption set should define:

• intended operating range
• excluded conditions
• known limitations
• conservatisms
• conditions requiring re-analysis

This prevents later misuse of the model outside its qualified domain.

9) Use formal review and signoff

Have qualified reviewers confirm:

• the assumptions are complete
• the basis is adequate
• the traceability is intact
• the sensitivity results are acceptable
• the conclusions match the evidence

Keep review comments, responses, and closure records.

10) Keep a traceability matrix

A simple traceability matrix can look like this:

Requirement	Assumption ID	Basis	Model Input	Output Impact	Verification/Validation	Status
Core thermal margin must be demonstrated	A-12	Vendor fuel data	Gap conductance	Peak fuel temp	Benchmark case XYZ	Approved
Reactivity feedback must be bounded	A-18	Conservative correlation	Doppler coefficient	Shutdown margin	Sensitivity study	Approved

Best practices

• Prefer measured data over judgment
• Prefer conservative assumptions where appropriate
• Avoid unstated “default” software settings
• Don’t reuse assumptions from old models without re-justifying them
• Capture rationale contemporaneously, not after the fact
• Make assumptions easy to audit by a third party

Bottom line

To meet traceable assumption requirements, you need a controlled system where every reactor core modeling assumption is: identified, justified, sourced, linked to requirements, tested for impact, and formally approved.

If you want, I can also provide:

1. a fillable assumption register template, or
2. a traceability matrix example tailored to reactor core simulation.