Probabilistic Risk Assessment

Delivering Confidence through Probabilistic Risk Assessments

As the power generation industry continues to evolve, plant safety is a required constant. Our Probabilistic Risk Assessments (PRAs) / Probabilistic Safety Assessments (PSAs) go beyond safety to drive down operating costs through a sustainable and reliable approach.

Optimizing applications both during online and outage seasons provides utilities with comprehensive risk mitigation solutions, supported by our proven PRA and PSA models that capture invaluable risk information. Our team of experts interprets and evaluates that information to develop and implement robust PRA applications to enhance safety and performance.

To learn more about our PRA support, click here.


Optimizing Risk-Informed Applications and Maintenance

Our solutions include reducing maintenance and operations resources, minimizing regulatory interactions, reducing outage scope and complexity, re-categorization of system risk-importance metrics potentially minimizing component replacement costs, and reliable and rapid real-time risk monitoring to identify and manage plant risk.

Learn More

Optimizing PRA Models and Operations

Our solutions include development and update of PRA models to meet RG 1.200 guidelines as a prerequisite to use risk-informed applications, and training and consultation needs. Specific PRA model support includes internal events (systems), internal fire, internal flooding, external hazards, severe accident analysis, model optimization and integration, and specialty risk products.

Learn More

Optimizing Reliability Data Using Natural Language Processing and Artificial Intelligence

FLEX strategies have been implemented throughout the US nuclear fleet to cope with postulated post-Fukushima type events. US nuclear plants have been crediting FLEX components in their PRA to reflect the as-built, as-operated conditions. Reflecting FLEX in the PRA in a way that is consistent with the ANS/ASME PRA standard has unique challenges. One such challenge is the use of appropriate reliability data for the portable equipment used in the FLEX strategies. Westinghouse is a leader in an industry-coordinated activity through the PWROG to develop sound estimates of the reliability data to be used in modelling of FLEX equipment in the PRA. The Westinghouse FLEX reliability data has been developed starting from raw data collected from the fleet. The reliability data was reviewed and processed to generate failure rates that are recognized by the NRC as the preferred data that meet the PRA standard as endorsed by RG 1.200 and as such technically adequate for risk informed applications.

Westinghouse innovation has developed Artificial Intelligence (AI) algorithms to facilitate the review of the condition reports from the sites and classify them appropriately as functional failure of the equipment. The Natural Language Processing (NLP) algorithm is designed to help the data analyst sift through hundreds of condition reports. NLP will help the Westinghouse engineers process more plant data in the upcoming review of the reliability data, to account for 5 more years of operating experience.

This work has put Westinghouse in the leading role for the development of FLEX reliability data, and Westinghouse innovation is pushing NLP/AI forward! The results of this effort are linked below.

FLEX Equipment Data Collection and Analysis

Contact our team for more information and support on creating FLEX PRA models that can support your risk informed applications!

  • Ken Kiper - Westinghouse reliability data lead
  • Kyle Hope - FLEX reliability data lead
  • Nick Zwiryk - Data scientist, NLP lead
  • Damian Mirizio - PWROG PMO
  • James Boatwright - Risk application product manager
  • Daniel Margotta – Sales Manager