Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
CSA N289.2-10 (R2015): Ground Motion Determination for Seismic Qualification of Nuclear Power Plants
A Comprehensive Guide to the Canadian Standard for Establishing Seismic Design Parameters for Nuclear Facilities
1. Scope and Application
CSA N289.2-10 (R2015) — Ground motion determination for seismic qualification of nuclear power plants — is a key component of the Canadian nuclear standards suite developed by the CSA Group. It provides systematic procedures for establishing seismic ground motion parameters at nuclear power plant (NPP) sites in Canada. The standard applies to both new plant designs and periodic reassessments of existing facilities, ensuring that the design basis earthquake (DBE) and safe shutdown earthquake (SSE) are defined with appropriate rigour.
1.1 Primary Objectives
- To define the required seismic hazard inputs for the seismic qualification of structures, systems, and components (SSCs).
- To establish a consistent methodology for probabilistic seismic hazard analysis (PSHA) tailored to Canadian geologic and tectonic settings.
- To provide guidance on the development of design response spectra and acceleration time histories.
1.2 Relationship to Other Standards
This standard is designed to be used in conjunction with:
- CSA N289.1 — General requirements for seismic design and qualification of nuclear power plants.
- CSA N289.3 — Design procedures for seismic qualification of nuclear power plants.
- CSA N286 — Management system requirements for nuclear facilities.
Tip: CSA N289.2-10 (R2015) replaces the earlier N289.2-M1981 edition and incorporates modern PSHA practices, including the treatment of epistemic uncertainties. Users should also refer to the latest CNSC regulatory document REGDOC-2.5.2 for site evaluation.
2. Key Technical Requirements
2.1 Probabilistic Seismic Hazard Analysis (PSHA)
The standard mandates a comprehensive PSHA that includes:
- Seismic source characterization: Identification and characterization of all capable tectonic sources (faults and areal source zones) within a 500 km radius of the site. Sources must be classified by their maximum magnitudes, recurrence relationships (e.g., Gutenberg‑Richter parameters), and geometry.
- Ground‑motion prediction equations (GMPEs): Use of multiple GMPEs appropriate for eastern and western Canadian tectonic environments. Weighting of GMPEs must be justified via logic trees to capture epistemic uncertainties.
- Site response analysis: Incorporation of local soil and rock conditions through a site‑specific amplification study. Both linear and nonlinear effects must be considered, especially for soft‑soil sites.
2.2 Design Basis Earthquake (DBE) and Safe Shutdown Earthquake (SSE)
Two seismic hazard levels are defined:
| Parameter | Design Basis Earthquake (DBE) | Safe Shutdown Earthquake (SSE) |
|---|---|---|
| Annual exceedance probability | 10⁻³ / year (≈ 1 000‑year return period) | 10⁻⁴ / year (≈ 10 000‑year return period) |
| Purpose | Design of SSCs with moderate safety significance | Design of SSCs required to maintain safe shutdown |
| Response spectrum shape | Uniform hazard spectrum (UHS) at the DBE probability | UHS at the SSE probability (median or 84th percentile, as specified) |
| Minimum peak ground acceleration (PGA) | 0.1 g (regulatory minimum, unless lower hazard can be justified) | 0.2 g (or deterministic cap, whichever is larger) |
Important: The SSE must also be compared with a deterministic scenario (e.g., the maximum credible earthquake from the nearest capable fault). The larger of the probabilistic and deterministic values at each frequency is used for the final SSE spectrum.
2.3 Development of Design Spectra and Time Histories
Once the hazard curves are computed, the standard requires:
- Construction of uniform hazard spectra (UHS) for horizontal and vertical components at the specified annual exceedance probabilities.
- Selection and scaling of recorded or simulated acceleration time histories that match the target UHS over a period range of 0.1 s to 10 s. At least three pairs of horizontal motions and one vertical motion are needed for time‑history analysis.
- Incorporation of a safety margin: for the SSE, an 84th‑percentile spectrum may be used instead of the median if justified by the risk significance.
3. Implementation Highlights
3.1 Treatment of Uncertainties
The standard strongly emphasizes the use of logic trees to represent epistemic uncertainties in source models, GMPEs, and site response parameters. The resulting hazard curves are displayed with fractiles (e.g., 5th, 50th, 95th), and the selection of the fractile for design (e.g., median or 84th) is part of the qualification process.
3.2 Site‑Specific Investigations
Applicants must perform a geologic and geotechnical investigation program adequate to:
- Identify capable faults (displacement within the last 130 000 years for surface faults, 50 000 years for buried faults).
- Determine shear‑wave velocity profiles to 30 m or greater depth (Vs₃₀).
- Classify the site according to the National Building Code of Canada (NBCC) site classes (A–F).
Best Practice: Many Canadian NPP projects now perform a full “site‑specific PSHA” using an expert elicitation process, as recommended by Annex A of the standard. This ensures traceability and regulatory acceptance.
3.3 Periodic Reassessment
For existing plants, CSA N289.2 provides guidance on updating ground motion parameters when new scientific information becomes available (e.g., revised GMPEs, updated fault databases). The standard recommends reassessment every 10 years or after significant seismic events in the region.
4. Compliance Notes
4.1 Canadian Regulatory Context
The Canadian Nuclear Safety Commission (CNSC) requires compliance with CSA N289.2 for all new build applications and for periodic safety reviews. REGDOC‑2.5.2 (Site Evaluation – Seismic Hazard) explicitly references CSA N289.2 as an acceptable means of demonstrating compliance. Licensees should submit a full PSHA report, including logic trees, sensitivity analyses, and comparison with the deterministic evaluation.
4.2 Documentation and Quality Assurance
All input data, assumptions, and results must be documented in a seismic hazard report that includes:
- Source characterisation tables with references.
- Logic tree framework with weights and justification.
- GMPE selection and applicability analysis.
- Site response analysis results (e.g., amplification factors for 5% damping).
- Final UHS and time‑history sets.
Critical: Failure to properly account for nearby faults — especially in seismically active regions such as the St. Lawrence Valley or offshore British Columbia — has led to rejection of seismic hazard submissions by the CNSC. The standard’s requirement to consider “capable faults” must be strictly followed.
4.3 Peer Review Requirements
The standard recommends (and the CNSC typically mandates) an independent peer review of the PSHA by experts not involved in the project. The review should examine source models, GMPEs, logic tree weights, and site response methods. Documentation of the peer review must be submitted as part of the licensing basis.
4.4 Updates Since 2015 Reaffirmation
Although reaffirmed in 2015, users should be aware that the underlying scientific methods (e.g., GMPEs for eastern Canada) have evolved. The standard allows for the use of more recent, state‑of‑practice models if they are properly justified and compared with the 2015‑era models. The CSA N289.2 committee is currently working on a new edition (expected 2026–2027) that will incorporate updated seismic source models from the Geological Survey of Canada and modern GMPEs.
5. Frequently Asked Questions
Q: Can CSA N289.2-10 (R2015) be used for non‑nuclear critical facilities?
A: While developed specifically for nuclear power plants, the PSHA methodology described is applicable to other infrastructure (e.g., dams, LNG terminals, large bridges) where a high‑confidence seismic hazard assessment is needed. However, the specific DBE/SSE probability levels and deterministic criteria are tailored to nuclear safety requirements. Users should adapt the framework to their own risk targets and consult the relevant building codes (e.g., NBCC 2020).
A: While developed specifically for nuclear power plants, the PSHA methodology described is applicable to other infrastructure (e.g., dams, LNG terminals, large bridges) where a high‑confidence seismic hazard assessment is needed. However, the specific DBE/SSE probability levels and deterministic criteria are tailored to nuclear safety requirements. Users should adapt the framework to their own risk targets and consult the relevant building codes (e.g., NBCC 2020).
Q: What is the minimum magnitude considered in the PSHA?
A: The standard states that earthquake sources with moment magnitude Mw ≥ 4.5 should be included in the probabilistic hazard, as events below this threshold rarely cause damage to robust nuclear structures. However, regions with very high seismicity rates may justify a lower minimum magnitude. The logic tree should include sensitivity cases with Mmin = 4.0 and 5.0 to bracket the impact.
A: The standard states that earthquake sources with moment magnitude Mw ≥ 4.5 should be included in the probabilistic hazard, as events below this threshold rarely cause damage to robust nuclear structures. However, regions with very high seismicity rates may justify a lower minimum magnitude. The logic tree should include sensitivity cases with Mmin = 4.0 and 5.0 to bracket the impact.
Q: How does the standard address vertical ground motion?
A: Vertical design spectra are derived from the horizontal spectra using a vertical‑to‑horizontal (V/H) ratio. The standard suggests using a V/H ratio of 2/3 for periods less than 0.1 s and decreasing to 1/2 at longer periods, unless site‑specific data indicate otherwise. For the SSE, the ratio may be increased if the hazard analysis shows higher vertical components (e.g., near‑fault effects).
A: Vertical design spectra are derived from the horizontal spectra using a vertical‑to‑horizontal (V/H) ratio. The standard suggests using a V/H ratio of 2/3 for periods less than 0.1 s and decreasing to 1/2 at longer periods, unless site‑specific data indicate otherwise. For the SSE, the ratio may be increased if the hazard analysis shows higher vertical components (e.g., near‑fault effects).
Q: Are there separate values for eastern and western Canada?
A: Yes. The standard does not prescribe a single set of ground motion values; instead it requires a site‑specific evaluation. For eastern Canada (stable craton), the PSHA must use GMPEs appropriate for hard rock and moderate‑to‑low seismicity, while western Canada (active subduction and crustal faults) requires models that account for megathrust earthquakes, deep intraslab events, and crustal sources. The difference can lead to SSE PGAs in the west exceeding 0.5 g, whereas eastern sites may be around 0.2–0.4 g.
A: Yes. The standard does not prescribe a single set of ground motion values; instead it requires a site‑specific evaluation. For eastern Canada (stable craton), the PSHA must use GMPEs appropriate for hard rock and moderate‑to‑low seismicity, while western Canada (active subduction and crustal faults) requires models that account for megathrust earthquakes, deep intraslab events, and crustal sources. The difference can lead to SSE PGAs in the west exceeding 0.5 g, whereas eastern sites may be around 0.2–0.4 g.