This crosswalk maps POET outputs to specific Initiating Event Analysis (IE) and Data Analysis (DA) supporting requirements (SRs) of ASME/ANS RA-Sa-2009, the Level 1/LERF PRA Standard endorsed (with exceptions and clarifications) by NRC Regulatory Guide 1.200, Revision 3 (Appendix A). It identifies where POET supplies the data, estimation method, and documentation that help an analyst satisfy each SR.
Read this correctly. POET is a tool, not a PRA, and does not itself "conform" to
the Standard. Conformance and Capability Category are determined by a PRA peer review (NEI 17-07) of a
plant's PRA. The "CC" column indicates the capability category POET's data and method can support; the
analyst is responsible for applicability, integration, completeness, and documentation in the PRA of record,
and for reviewing the method as a newly developed method (see the Method Basis & V&V page). SR text is
paraphrased here; refer to the Standard for authoritative wording.
Initiating Event Analysis (IE) — Section 2-2.1
| SR | Requirement intent (paraphrased) | How POET supports it | Level | CC |
|---|---|---|---|---|
HLR-IE-A / IE-A3 | Review plant-specific initiating-event experience so the list of challenges reflects plant history. | Surfaces the plant's actual unplanned outages and scram events (event drill-down) and classifies their causes, giving the plant-experience review directly from operating data. | Supports | I-III (data) |
IE-A4 | Review generic analyses / industry operating experience of similar plants. | Fleet and peer-group benchmarking plus the NRC/INL industry-average baselines provide the industry-experience comparison. | Supports | I-III (data) |
IE-A7 | Incorporate events such as unplanned controlled shutdowns that include a scram prior to reaching low power. | POET's outage and scram datasets already include unplanned/forced shutdowns and scrams. | Supports | I-III (data) |
HLR-IE-B | Group initiating events by similar mitigation requirements. | POET groups forced events by duration (short/medium/long) and by attributed cause; mitigation-based grouping remains the analyst's. | Partially supports | — |
HLR-IE-C / IE-C1 | Calculate IE frequency accounting for relevant generic and plant-specific data. | Core POET output: empirical plant-specific IE frequencies combined with NRC/INL generic baselines. | Supports | II-III |
IE-C2 | Use the most recent applicable plant-specific data; justify exclusions. | User-selectable analysis window with a data-vintage stamp; the window and exposure basis are recorded in the traceability block. | Supports | II-III |
IE-C4 | Combine generic and plant-specific data with a Bayesian update; justify the prior. | POET performs the Bayesian update (NRC industry prior and/or empirical-Bayes fleet prior) and justifies prior applicability with the DA-D4(c) consistency check. | Supports | II-III |
IE-C5 | Calculate IE frequencies per reactor-year, weighted by the fraction of time at power. | POET normalizes per reactor-critical-year (days at power > 0 / 365.25) in addition to per reactor-calendar-year, directly implementing the at-power weighting. | Supports | II-III |
IE-C7 | Use time-trend analysis to account for established trends (e.g., declining trip rates). [CC-III] | POET reports the annual IE-frequency trend; acceptable trend methodologies cited (NUREG/CR-6928). | Supports | III |
IE-C12 | Compare results with generic data sources and explain differences (reasonableness check). | NRC comparison table with the plant/NRC ratio and the prior-data consistency verdict provides the reasonableness check. | Supports | I-III |
IE-C13 | For rare / extremely rare events, use industry generic data. | POET defers to the NRC/INL industry baselines for rare categories; cause attribution maps observed causes to NRC categories. | Supports | I-III (data) |
HLR-IE-D | Document the initiating-event analysis consistent with the SRs. | Per-result traceability block, full report (PDF/Word), and the Method Basis & V&V page. | Supports | — |
Data Analysis (DA) — Section 2-2.6
| SR | Requirement intent (paraphrased) | How POET supports it | Level | CC |
|---|---|---|---|---|
HLR-DA-A | Each parameter clearly defined (logic model, basic-event boundary, model used). | POET defines each output (IE-SHORT/MEDIUM/LONG-TRANSIENT, exposure, refuel, state fractions) with explicit definitions and units. | Supports | I-III |
HLR-DA-C / DA-C1 | Obtain generic parameter estimates from recognized sources. | NRC/INL SPAR industry-average IE frequencies, component reliability, CCF alpha-factors, and LOOP data are reproduced verbatim from the published source. | Supports | I-III |
HLR-DA-D / DA-D1 | Calculate realistic estimates from generic + plant-specific evidence via Bayes update; choose priors as noninformative or representative of industry variability; characterize uncertainty. | POET offers exactly these prior choices: Jeffreys noninformative, the NRC industry prior, and an empirical-Bayes fleet prior (industry variability), each producing a posterior with a credible interval. | Supports | II-III |
DA-D3 | Provide a mean value and a statistical representation of the uncertainty interval (Bayesian updating acceptable). | Every frequency and exposure carries a posterior mean and a 90% credible interval (5th-95th); state fractions carry Wilson intervals. | Supports | II-III |
DA-D4 item (c) | When using Bayes, check the posterior is reasonable and examine inconsistencies between the prior and the plant-specific evidence. | POET computes the prior-predictive (gamma-Poisson) consistency p-value and reports a consistent / not-consistent verdict against both the NRC and fleet priors, plus the posterior mean. | Supports | II-III |
DA-D5 | Use an accepted CCF model such as the Alpha Factor Model. | The reference library provides the SPAR alpha-factor CCF parameters (data side of the CCF estimate). | Supports | I-III (data) |
HLR-DA-E | Document the data analysis consistent with the SRs. | Traceability block, exports with embedded metadata, and the Method Basis & V&V page. | Supports | — |
Notes & scope
- Strongest fit: IE-C (frequency estimation) and DA-D (parameter estimation / Bayesian updating with uncertainty), where POET's empirical frequencies, dual normalization, Bayesian priors, credible intervals, and prior-data consistency check map almost one-to-one.
- POET supports, does not replace: complete IE identification and grouping (IE-A1, IE-A5/A6 FMEA, full HLR-IE-B), fault-tree-modeled initiators (IE-C8-C11), and component test/demand data collection (most of DA-C) remain the analyst's responsibility.
- Caveat carried through: POET's all-unplanned frequency is a daily-power-derived proxy for the reactor-trip rate; cause attribution (keyword-based, text-coverage floor) narrows it toward specific NRC categories. The analyst should confirm IE completeness independently.