
190
ENSI Erfahrungs- und Forschungsbericht 2011
utility, using its full-scope training simulator. The
OECD Halden Project is also supporting the U.S.
study with its staff as part of the Project’s Joint Pro-
gramme, to which Switzerland (ENSI) contributes.
Additionally, a diverse group of partners including
industry, regulators, and research institutes, are
contributing on the assessment group as well as
in the HRA analyst teams. In addition to PSI, these
include EPRI, Sandia National Laboratories, Idaho
National Laboratory, SAIC, all US; NRI, Czech Rep.
PSI is participating in a task group within the OECD
NEA / Committee for the Safety of Nuclear Installa-
tions (CSNI), «Establishing Desirable Attributes of
Current Human Reliability Assessment Techniques
In Nuclear Risk Assessment». This task, led by the
U.K. Nuclear Installations Inspectorate, is a joint
task of the Working Group on Human and Orga-
nizational Factors and the Working Group on Risk
Assessment. Its aim is to perform an international
technical evaluation of HRA methods, considering
criteria shared by the member countries and oth-
er common criteria. The task group held its first
meeting in Sept. 2011, with the next meetings
planned for March and September/October 2012.
The work related to crew-plant simulation for safe-
ty assessment is supported by a cooperation with
the University of Maryland (Prof. A. Mosleh), on
the ADS software tool for dynamic event tree anal-
ysis [11, 12].
Assessment 2011
and Perspectives for 2012
EOC plant-specific pilot study III
1. Organization of EOC Pilot Study III with a Swiss
plant.
The pilot study will be initiated in 2012, based on
the agreement reached in 2011 with a Swiss plant.
Quantification of EOCs
2. Revision of CESA-Q guidance, focusing on
traceability and scope.
In the developed model-based quantification, the
human error probability is the direct output of the
model. This decreases the element of judgment
(which is typically leading to low traceability and
repeatability) required for the analyst: once ap-
propriate evaluations of the CESA-Q factors are
made, the probability derivation does not need
any additional judgment from the analyst. A re-
port on the developed quantification approach
will be completed in early 2012.
While the 2011 revision of CESA-Q focused on
the quantification element, the main objective for
2012 is the revision of the method guidance for
qualitative error analysis, focusing on additional
reference analyses that support identification of
challenging conditions inducing more likely errors.
This would extend the scope of the method to sit-
uations outside the current coverage of the CESA-
Q database.
Simulator Data for HRA
and Method Assessment
3. Evaluation of HRA predictions and method as-
sessment in the U.S. HRA Empirical Study.
The preliminary evaluation in the U.S. study was
completed and reviewed with the analysis teams
at a workshop hosted by the US NRC in Rock-
ville, MD, USA, in June 2011. The results will be
presented publically in two contributions to the
PSAM11 conference in June 2012. The final re-
port on the U.S. study is scheduled to be complet-
ed in 2012. In parallel, the study assessment team
is completing a final overall report on the Interna-
tional Study, also to be published in 2012. The U.S.
study confirmed many of the earlier international
study results, its final report focuses on the consis-
tency of predictions and of method performance
across analysis teams.
4. Derivation of scenario-specific lessons for HRA
from the data collected in the simulator studies of
the HRA benchmarks.
These lessons are presented in the Internation-
al HRA Empirical Study final report, to appear in
2012.
Technical Basis for Seismic HRA
5. Detailed overview of earthquake experiences at
selected nuclear power plants.
The survey of earthquake experiences at nuclear
power plants, focused on the human performance
elements, highlighted several issues relevant for
seismic HRA: interaction of non-safety-related
damage with emergency response, weaknesses
in limited aspects of emergency plans, the impact
of those weaknesses on personnel workload. The
survey included only 1 event approaching the de-
sign basis safe shutdown earthquake level (Suru-
ga Bay, 2009, for Hamaoka Unit 5), which means
that there was no or very limited damage to safe-
Comentarios a estos manuales