Development of multi-unit probabilistic safety assessment using the Khmelnytskyi NPP site as an example

The safety assessment of nuclear power plants (NPPs) has traditionally focused on individual reactor units, without fully considering the interactions between multiple units at a single site. However, recent international studies, particularly following the Fukushima Daiichi accident, have emphasized the importance of integrated risk assessment for multi-unit (MU) sites. This study presents the initial stages of developing a Multi-Unit Probabilistic Safety Assessment (MU PSA) for the Khmelnytskyi NPP, addressing both regulatory and methodological aspects. The research follows a six-stage framework, with this paper covering the first two stages. The first stage involved a comprehensive review of the Ukrainian nuclear safety regulatory framework to determine its applicability to MU risk assessment. The second stage focused on updating the integral probabilistic models of Units 1 and 2 of the Khmelnytskyi NPP to a unified “freeze date”. This process involved standardizing failure rate data, refining initiating event frequencies, and ensuring consistency in system reliability parameters. The PSA models were updated using RiskSpectrum PSA software, incorporating operational reliability data, failure statistics, and maintenance records. The regulatory review highlighted that Ukrainian nuclear safety regulations only partially account for MU risk and do not establish specific risk thresholds for sitewide indicators such as Core Damage Frequency (CDF) or Large Early Release Frequency (LERF). This underscores the need for further regulatory refinement to align with international best practices. The model update provided a harmonized basis for MU PSA by: standardizing equipment failure data and initiating event frequencies; accounting for shared system dependencies and common-cause failures; updating human reliability assessments and accident sequences. These refinements ensure that subsequent research phases can build on a robust foundation for evaluating integrated site-wide risk. The findings from these initial stages lay the groundwork for a comprehensive MU PSA methodology applicable to Ukrainian NPPs. Further research will focus on integrating the individual unit models into a full-site probabilistic model and performing sensitivity and uncertainty analyses. The results will contribute to enhancing safety assessment approaches and supporting regulatory improvements for multi-unit sites.