Risk Pilotin toimintatavalle on luonteenomaista
Tutkimus ja kehitys on siten luonnollinen osa toimintaamme.
T&K-projektit ovat meille strategisesti tärkeitä verkottuaksemme turvallisuus- ja riskianalyysien osaajien kanssa ympäri maailmaa.
Risk Pilotilla on pitkä kokemus T&K-hankkeista erityisesti pohjoismaisten ydinvoiman turvallisuustutkimusohjelmien myötä. Olemme lisäksi osallistuneet kansainvälisiin tutkimusohjelmiin kuten Euratomin tutkimusohjelmiin sekä IAEA:n ja OECD/NEA:n aktiviteetteihin. Asiantuntijoitamme kutsutaan usein mukaan alan työryhmiin.
Ohessa esimerkkejä T&K-projekteistamme
Risk Pilot has internationally top ranked experts in the field of digital I&C hardware and software reliability analysis within PRA context. R&D has been performed within NKS/NPSAG/SAFIR project DIGREL, and in the PSA for Ringhals 1 Risk Pilot has introduced a detailed reliability model on digital I&C.
The DIGREL project (2010-14) has been financed by Nordic Nuclear Safety Reseach (NKS), The Finnish Nuclear Safety Research Programme SAFIR and the Nordic PSA Group. The project was performed by Risk Pilot AB (project leader), VTT and Lloyd’s Register Consulting. As a result, guidelines to analyse and model digital systems in a PSA context have been prepared.
Together with a task group of OECD/NEA Working Group RISK a failure modes taxonomy for digital I&C systems has been developed. The taxonomy support perfomance and review of reliability analyses of digital I&C systems.
Software reliability analysis is still an open issue. In DIGREL, a practical approach on software modelling and quantification has been developed by a Finnish-Swedish-German collaboration.
To support the development work and to demonstrate the analysis approaches a generic digital I&C system example has been developed. DIGREL example model a simplified yet rather comprehensive PSA model representing a nuclear power plant with four-redundant safety systems and diversfied reactor protection system.
Read more about DIGREL and its example model using the links below:
DIGREL example architecture
HiDep is a software tool used for the analysis of common cause failures (CCF), based on the Extended Common Load Model (ECLM). ECLM is primarily suited for the CCF quantification in highly redundant systems with the number of components ranging more than four components.
The theory behind ECLM and the original HiDep toolbox are developments of Tuomas Mankamo from Avaplan Oy. Since 2015, the software product has been under copyright by Forsmark Kraftgrupp AB, Ringhals AB, Teollisuuden Voima Oyj and OKG AB. The HiDep toolbox may be obtained free of charge from the software owners under certain conditions.
Risk Pilot has updated the HiDep toolbox and prepared manuals for the use of HiDep. The toolbox is a set of Microsoft Excel applications including several modules needed for the manipulation and quantification of the CCF models and data. The current version is 2.7 (2015) and it uses Excel 97-2003 Workbook as the platform. Risk Pilot can provide services in CCF analyses and in the use of HiDep.
For more information: SSM report 2017:11
Human reliability analysis (HRA) is one of Risk Pilot’s key competence areas. For Risk Pilot, HRA is an elementary part of risk and reliability studies as well as a supporting method for human factors engineering applications.
Risk Pilot applies state-of-the-art methods for HRA, and our experts have been involved in international HRA method development projects since 1990’s, e.g., related to HRA methods comparisons, analysis of errors of commission, effectiveness of operability verification, assessment of human error dependences.
In the PSA studies for Forsmark and Ringhals NPPs, we use several methods dedicated to the assessment of various types of human interactions, e.g., the enhanced Bayesian THERP (Technique for human error rate prediction [NUREG/CR-1278]) for the assessment of so called post-initiator operator action errors. The enhanced Bayesian THERP method begins with a qualitative assessment supplemented with a quantification procedure that integrates generic human error probability scales with performance shaping factors and expert judgements in a Bayesian manner. It has been evaluated internationally in