Software

RiskSpectrum Multi-Unit PSA

Quantify your multi-unit risk using your single unit PSA models

RiskSpectrum Multi-Unit PSA fully employ and utilize existing PSA models and their analyses of individual nuclear power plant units located at the same site.

The multi-unit scenarios are defined separately, by specifying initiating events affecting multiple units, failures of shared equipment, shared parts of safety systems, common cause failures of components across units, and possible human failure events with consequences exceeding a single unit.

Schematics of the PSA oxygen concentrator. | Download Scientific Diagram

Software In Action

 

One Multi-Unit Event Tree for All

RiskSpectrum Multi-Unit PSA automatically creates a so-called multi-unit event tree, in which the multi-unit events you have defined are arranged.

The failure and success of the multi-unit events form sequences.

At the end of each sequence, RiskSpectrum Multi-Unit PSA pass the information from the multi-unit event tree to the individual units PSA models and quantify their conditional consequence probabilities, e.g., core damage, for each unit separately.

Finally, RiskSpectrum Multi-Unit PSA compose the quantification results of the multi-unit sequences and individual units to the overall multi-unit failure frequency.

Quantification

The scalability of the application is ensured using existing MCS lists for the single units.

As the failure combinations that are relevant for a single unit are represented by the single unit MCS list, the failure of several reactors should also be contained within those MCS lists (unless the frequency of the multi unit failures is negligible).

The quantitative results will also provide quantitative and qualitative insights regarding which dependencies/shared systems or operator actions that are the most import dependencies from a multi unit perspective.

What is the Impact?

Two large size PSA model were used simulating multi-unit CCFs and multi-unit human error dependencies.

As expected, an increase in the multi-unit release probability can be seen when multi-unit dependency is considered.

The only inputs necessary for the quantification were the MCS of the single units and a list of the multi-unit CCF events and human error dependencies.