Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
Advancing Human Reliability Analysis Methods for External Events with a
Focus on Seismic

Jeffrey A Julius Jan Grobbelaar & Kaydee Kohlhepp
Scientech a CurtissWright Flow Control Company Tukwila WA USA


Abstract The reliability of operator actions following an external initiating event is a topic that has
increased importance following the 2011 seismicinduced tsunami at the Fukushima Daiichi site in
Japan This event has prompted licensees in the USA and internationally to reexamine their plant’s
risk profile and the plant’s ability to prevent andor mitigate damage following external initiating
events (external hazards) In support of the industry initiatives to evaluate and prepare for external
initiating events the Electric Power Research Institute and Scientech have developed a preliminary
approach to analyze the reliability of operator actions following external initiating events with a
specific focus on seismic events The preliminary approach has been published in EPRI 1025294 A
Preliminary Approach to Human Reliability Analysis for External Events with a Focus on Seismic in
December 2012 Since the development of the 2012 report the approach and methods suggested in the
report have been applied in the development and in the review of seismic PRAs that are currently in
development This paper summarizes the development of the current external events human reliability
analysis (HRA) methods and guidance and summarizes recent insights from applying this approach to
seismic PRAs

Keywords HRA Seismic HRA External Events External Hazards

1 INTRODUCTION

The purpose of EPRI report 1025294 [1] is to provide methods and guidance for the human reliability
analysis of external events PRAs based on the current stateoftheart in both PRA and in HRA
modeling Prior to the development of this report substantial research has been performed to develop
and improve Human Reliability Analysis (HRA) methods in support of Probabilistic Risk Assessment
(PRA) The development of existing HRA methods however was limited primarily to internal events
PRA specifically to initiating events that did not involve spatial impact These methods often contain
underlying assumptions that may or may not be applicable to areaspatial impacts especially those
affecting the plant site such as the regional impact following a seismic event external flood or
hurricane (external initiating event or external hazard) Recent HRA advances that culminated in the
publication of Fire HRA methods and guidance in NUREG1921 [2] were considered in the
development of EPRI report 1025294

Additionally the stateoftheart in seismic and external events PRA models and issues was surveyed
in order to understand existing external events HRA guidance [3 4] The results of this review were
not surprising As is common in HRA there was a wide variation in existing methods for external
events HRA Variation existed between methods used for different hazard types as well as plantto
plant variation for evaluation of a given hazard type In addition to reviewing current external events
PRA models and methods a review of historical operating experience was conducted The relevant
insights from the review of operational experience were incorporated into the development of the
various steps of the 2012 external events HRA process The operating experience review task
primarily built upon previous published work conducted by EPRI (PostEarthquake Investigation
Program from 1985 to 2012) as well as a review of utility presentations and LERs Additionally
interviews were conducted with personnel from nuclear plants impacted by recent seismic events The
review of historical data focused on realworld seismic events at nuclear power plants and other
industrial facilities and it was performed in order to identify potential failure modes and performance
shaping factors (PSFs) that should be considered when developing external events HRA
Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
EPRI report 1025294 provides a framework for external events HRA a general screening approach
and a detailed quantification approach which can be applied consistently across a variety of external
events were developed The report was written to provide methods and guidance for all external
events but included specific guidance for HRA in a seismic PRA including operator actions to
recovery from relay chatter The detailed quantification approach provided in EPRI report 1025294 is
an adaptation of the EPRI HRA Methodology also known as the EPRI HRA Approach for
internal events [5 6 7 and 8] The specific objectives of EPRI report 1025294 are listed below

 Provide a consistent framework for analysts to perform HRA for all external hazards

 Provide hazardspecific guidance for consideration of relevant performance shaping factors
(PSFs) based on operational experience and existing research

 Provide a general screening approach and detailed quantification method that can be applied
consistently to a variety of external events

 Provide seismicspecific guidance that reflects to the extent possible current research and
relevant operational experience
Paper organization Section 2 of this paper describes the HRA process as it supports external hazards
PRA Sections 3 through 8 summarize the treatment of external events HRA in EPRI report 1025294
Additional information on the external events HRA approach of EPRI report 1025294 has been
described in earlier conference papers [9 10 and 11] Section 9 summarizes insights including areas
of potential future research and conclusions

2 HRA PROCESS

As with recent HRA guidance such as NUREG 1921 [2] the external events HRA process is often
appears as a linear process with the following elements

1 Identification and Definition

2 Qualitative Analysis

3 Quantification

4 Model Integration

a Cut set Review and HEP Reasonableness Check
b Recovery
c Dependency
d Uncertainty

Although this process is often depicted as sequential steps in the practical application to developing an
HRA these steps are iterative EPRI 1025294 [1] presents the guidance in the order which an HRA
analyst is likely to use the various elements of the guidance accounting for the iteration between
screening and detailed assessments Figure 1 provides a mapping between the external events HRA
process and the sections of EPRI 1025294 This figure shows the iterative relationship between the
PRA process the HRA process and allows both tasks to proceed in parallel
Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
EPRI report 1025294 provides two approaches for quantification first a screening quantification and
then a detailed quantification The screening approach is intended to require fewer resources and be
more conservative than the detailed quantification


PRA Model
(eg Initiating Events
Event Trees Fault Trees
Internal Events PRAHRA)
PRA Scope & Regulatory
Requirements Identification & Definition (Ch 3)
 From Internal Events PRA
 Response to External Events
 Not previously modeled in
Internal Events
List of Defined HFEs
 Preventive (New to External Events)
 Preinitiators
 Initiators
 Postinitiators
Feasibility (Ch 42)
Perform preliminary qualitative
analysis using screeninglevel
feasibility requirements
Does the HFE meet the
feasibility requirements for
Screening
Perform Screening Quantification
(Ch 43)
Use Screening Tables to produce a
screening HEP
Yes
Is PRA quantification okay
with a 10 for that HFE
No
HEP 10
Yes
Detailed Qualitative Analysis
 Detailed assessment of operational
narrative and PSFs (Ch 5)
 Reassess Feasibility using detailed
level feasibility requirements (Ch 42)
Quantitative Analysis (Ch 6)
Use Screening Tables or EPRI HRA
Approach to produce a HEP
Integrate with PRA (Ch 7)
 Recovery
 Uncertainty
 Review of CutsetsDependency
& Reasonableness check
Is PRA quantification okay
using the Screening HEP
No
No
Review of
Relevant
Operational
Experience
Yes
Model Refinement As
Needed
Plant Information
(eg EOPs Event
Response Procedures
Deterministic Analyses)
INFORMATION COLLECTION (Ch 52)

Figure 1
Mapping of HRA Steps To Sections Within EPRI 1025294 [1]


3 IDENTIFICATION AND CLASSIFICATION

The identification process in an external events PRA follows the same approach as in an internal
events PRA – to understand the plant response (including the procedures to be implemented) and to
understand how the plant response is captured in the PRA model Within the PRA operator actions
typically come from one of two sources 1) HFEs already existing in the PRA (usually from the Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
internal events PRA) or 2) procedure review in conjunction with modeled accident sequence review to
identify new operator actions

In general the following groups of procedures are reviewed for applicability to external events

 Preventive procedures –procedures for preparing for high winds hurricane or other external
events where the onset of the event is known beforehand typically when the hazard is
imminent

 Response procedures – Those procedures used in response to an initiating event Response
procedures include Emergency Operating Procedures (EOP) Abnormal Operating Procedures
(AOPs) Alarm Response Procedures (ARP) Severe Accident Management Guidelines
(SAMGs) fire procedures and seismic or other external event procedures (including FLEX
procedures)

 Normal operating procedures (NOP also known as operating procedures) Those procedures
used in day to day plant operation These procedures include normal shutdown and start up
procedures system alignment procedures and test and maintenance procedures These plant
specific procedures are well trained on and the wording is standardized across the complete
procedure set

For external events HRA there are three types of postinitiating event operator actions

 Internal events operator actions

 Preventive operator actions

 External event response operator actions

The internal events operator actions associated with these HFEs are actions required in response to a
plant initiating event andor reactor trip typically directed by the EOPs ARPs AOPs andor NOPs
Because internal events operator actions have been identified their HFEs defined and their HEPs
quantified as part of the internal events HRA it is not necessary to repeat the internal events HRA
identification process All that is required for the external events PRA identification process is to
determine which of these HFEs could occur in external events scenarios

Preventive actions would be plant and external event specific and the identification of these actions
would be performed by a review of procedures and discussions with plant operations These actions
would typically be included in the external events PRA on asneeded bases Preventive operator
actions are an area of ongoing study and while they are not explicitly within the scope of EPRI
1025294 [1] they are subject to the same feasibility criteria described in that report Example of
preventive actions could include

 Closing doors or placing flood barriers such as sand bags or drain plugs prior to flood
damage
 Transporting additional diesel fuel on site prior to an expected prolonged loss of offsite power
such as a hurricane
 Staging portable equipment (eg preparing to implement FLEX options)

External events response actions are new postinitiating event operator actions used to mitigate the
effects of an external event This category of HFEs is typically not included in the EOPAOP network Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
of procedures These operator actions are identified by review of the external event response
procedures in conjunction with the modeled PRA functions and sequences

Response actions are sometimes called recovery actions and may appear in event trees or the fault tree
portions of the PRA Response actions consist of the following types of actions

 Terminating the impact of the external initiating event – actions taken to identify and protect
components that are operating in an undesired state or are threatened after the external event
has occurred These are somewhat analogous to preventive actions described above but often
have a shorter time window For example if a poweroperated relief valve spuriously opens
due the initiating event the failure may be able to be recovered by deenergizing the valve

 Mitigation of external initiating event consequences using the affected SSC – actions taken to
recover failed SSCs by providing an alternate success path For example actions taken in
response to a seismicallyinduced LOSP and SBO due to relay chatter preventing load
sequencers from loading EDGs and equipment loads onto vital AC buses The HFE models
operators resetting circuitsrelays from the control room or in the switchgear rooms restarting
the EDGs and loading equipment manually if the load sequencer remains unavailable Note –
human reliability analysis does not address repair of failed components

 Mitigation of external initiating event consequences using alternate components – actions
taken to recover failed SSCs by providing an alternate success path For example restoration
of power to an electrical bus by aligning an alternate component such as a standby
rectifierinverter or a source (such as a skidmounted diesel generator used for FLEX) Note –
human reliability analysis does not address repair of failed components

Regardless of how the operator action is identified the corresponding HFE must be defined for use in
the external events PRA The human failures are defined to represent the impact of the human failures
at the function system train or component level as appropriate

For new actions the definition should start with the collection of information from PRA and
engineering analyses For actions carried over from the internal events the existing definition should
be reviewed and modified asneeded to account for the new context of the external event

4 FEASIBILITY

The HRA for most spatial analyses is typically performed in conjunction with the PRA development
Because tasks of the PRA are typically developed concurrently not all of the information required to
perform a detailed HFE quantification will be known initially and the PRA will need screening HEP
values initially to develop and quantify the risk model

For a screening analysis the HFE definition and feasibility assessment are conducted simultaneously
as part of the initial qualitative analysis If a more detailed analysis is needed then the initial
qualitative analysis should be further developed Prior to performing the qualitative analysis if the
operator action did not pass the screeninglevel feasibility assessment the feasibility should be
reassessed after gathering more details

Regardless of when the feasibility assessment is conducted or the level of detail of the current PRA
the feasibly assessment needs to consider the following at a minimum
 Timing
 Manpower
 Cues Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
 Procedures and training
 Accessible Location & Environmental Factors
 Tools and equipment operability

If the operator action is feasible the analyst can proceed to perform either a screening or a detailed
quantification If the analyst finds the screening to be too conservative or limiting the analyst is
encouraged to apply the detailed HRA method EPRI 1025294 [1] provides additional detail on
considerations for each of these feasibility criteria


5 SCREENING ANALYSIS & QUANTIFICATION

The screening process is optional but it provides a set of HEPs for the initial PRA model
quantification and helps identify the important sequences The ranking can be used to determine
which sequences might be further analyzed to reduce the calculated risk by detailed modeling

The screening method provided was initially developed specifically for application in developing a
seismic risk assessment However with the current stateofknowledge it is reasonable to use the
described screening approach in EPRI 1025294 [1] as a screening method for other external events
with the caveat that future research on other external events may require this approach to be modified
to incorporate relevant operating experience

51 Step 1 Identify Damage State of the Plant Following the External Initiating Event

Both the screening HRA method and the detailed HRA method start by asking the analysts to identify
the damage state of the plant following the external initiating event The damage state is intended to
account for the overall context resulting from the external event beyond the specific failures dictated
by the cut set including impact to local infrastructure and nonsafety related systems level of
heightened stress general increase in level of coordination and workload and quality of working
environment These damage states described in Table 1 were selected based on the definitions
provided in EPRI NP6695 and its update EPRI 1025288 [4] but have been adjusted here to correlate
more closely with the impact of the context on operator performance Because the design basis for the
range of external events can vary substantially from plant to plant the bins selected here reflect the
effect of the external event on the plant rather than providing absolute values (eg PGA values) Bin
definition is generic for seismic because the seismic hazard is not straightforward and there is not a
direct correlation between hazard level and damage state Recommendations for seismic HRA
provided in Table 1 may not fit the damage state definition appropriately for every plant and is
provided only as a starting place when no other information is available it is expected that the HRA
analyst will have to interface with the PRA analyst to correlate the damage states provided here with
the hazard bins (eg ground motion intervals) used in the PRA Note The SSE is a convenient but
generally very conservative value higher values could be justified HCLPF recommendations here
may also be overly conservative
Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
Table 1
Damage State Definitions For Screening

Bin
#
External Event Damage State Description Recommended Link to Seismic Hazard
1 No damage to the plant safetyrelated SSCs or non
safety SSCs required for operation Limited damage
to nonsafety nonseismic designed SSCs like
residences and office buildings

Below the SSE
2 No expected damage to the plant safetyrelated SSCs
or to rugged industrial type nonsafety SSCs required
for operation Damage may be expected to non
safety SSCs not important to plant operations and to
the switchyard (eg LOOP expected) Falling of
suspended ceiling panels

At or above the SSE up to HCLPF of most
fragile safetyrelated SSC
(eg 2011 North Anna event)
3 Widespread damage to nonsafety related SSCs
andor some damage expected to safety related SSCs
Significant number of vibration trips and alarms
requiring resetting
Above the HCLPF of most fragile safety
related SSC to HCLPF of critical
instrumentation or HCLPF level of 25th
percentile component whichever is lower
(eg 2007 KashiwazakiKariwa
2011 Onagawa events)
4 Substantial damage to safety related and nonsafety
SSCs The threshold of this damage state is such
that it produces a cliffedge effect in the likelihood
of operator response

Widespread damage to critical
instrumentation
(2011 Fukushima Daiichi and Daini events)


52 Step 2 Plant Damage Assessment

Another consideration that appears in the screening trees is based on the plant damage assessment
Following an external initiating event often times the entire site is affected and the effects (such as
flood water obstructing access high radiation areas andor damaged equipment) can impact human
performance far after the event is over Thus the external initiating event can have impacts on both
cognition and execution that last a considerable amount of time There is expected to be an overall
reduction in workload and complexity once the site has been assessed and the extent of the damage
understood
The damage assessment has been defined in the external events HRA as a breakpoint for both
cognition where not only is the damage to the plant known (after the breakpoint) but also the
workload and distractions associated with determining the impact of the event are reduced The
cognitive load on the operators is reduced because they have a clearer picture of the damage inflicted
on the plant due to the initiating event including an understanding of what equipment is damaged but
may not have failed yet By this point in the scenario it is also expected that the crew have had the
opportunity to implement basic working solutions to compensate for issues caused by the external
event (eg determining usable pathways establishing alternate means of communication etc) For
seismic events a detailed plant damage assessment consists of a postevent walkdown and is usually
performed within 4 to 8 hours of the event This initial walkdown should not be confused with the
more involved formal damage assessment required prior to restart to identify issues which might
degrade longterm reliability of components typically called postshutdown inspections and tests
Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
53 Step 3 – Assessment of Time Margin

A review of the operating experience suggests that the PSFs associated with seismic events manifest
themselves most often as a delay rather than direct failure of the operator action Therefore the level
of credit assigned at the screening level is dependent upon the amount of time margin – or tolerance
for unexpected delays – available

54 Step 4a Quantify Screening HEP For HFEs From Internal Events PRA

For operator actions carried over from the internal events PRA into the external events PRA the
internal events qualitative and quantitative analysis can be used as the starting point for the external
events PRA quantification A simple decision tree has been developed to show how to determine a
multiplier to apply to the internal events HEP The event tree considers the following headings in the
development of the HEP The end state of the decision tree branches are either a screening HEP or a
multiplier for the internal events HEP Multipliers range from 2 to 50 and screening HEPs range from
the internal events HEP value to 1

 Immediate memorized action (or not)
 Action location
 Damage state
 Time margin consideration
 Cue before or after plant damage assessment

55 Step 4b Quantify Screening HEP For New External Events HFEs

For new operator response actions that were not carried over from the internal events PRA into the
external events PRA multipliers are not applicable but the same factors are used to determine a
screening HEP Screening HEPs were developed by selecting a base human error probability (BHEP)
then applying the same multipliers used in the internal events HFEs described above the screening
values were given no more credit than 10E2

6 DETAILED ANALYSIS & QUANTIFICATION

By the time the analyst has reached the stage requiring a detailed analysis of the HFE the HFE has
been defined and the basic feasibility has been assessed The HFE definition and feasibility comprise
the foundation of the qualitative analysis The feasibility criteria for screening is more stringent than
that required for detailed analysis (eg the screening requirement for an action to be considered
feasibility is that the primary cue must be available whereas the detailed analysis stipulates that either
the primary or secondary cues must be available) Therefore additional data gathering and analysis
may need to be performed to satisfy the feasibility criteria for a detailed analysis if the feasibility
criteria for the screening analysis were not met

Qualitative analysis is an essential part of an HRA The objectives of qualitative analysis are to
understand the modeled PRA context for the HFE understand the actual asbuilt asoperated
response of the operators and plant and translate this information into factors data and elements used
in the quantification of human error probabilities

Recent experimental studies have shown that the quality of the quantitative analysis is strongly
impacted by the quality of the qualitative analysis even for fairly prescriptive methods such as the
EPRI HRA Methodology ERPI 1025294 [1] provides detailed guidance on performing a thorough
qualitative analysis using insights from seismic operating experience

The EPRI HRA Methodology (also known as the EPRI HRA Approach) is based on EPRI’s SHARP
and SHARP1 [5] HRA framework After the qualitative analysis has been performed a detailed Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
quantification is performed using methods recommended by EPRI within the HRA approach
Specifically one or more of the following methods

 Cognitive Methods The Human Cognitive ReliabilityOperator Reliability Experiment
(HCRORE) andor CauseBased Decision Tree Method (CBDTM) [67] for cognition
 Execution Technique for Human Error Rate Prediction (THERP) [8] for execution

One advantage of using existing methods for external events HRA is that at a minimum the same
fundamental aspects and factors affecting human performance apply to Level 1 internal events PRA as
well as external events PRA —therefore applying these methods to external events scenarios should
yield a good firstorder approximation of operator failure and would further be consistent with the
modeling for nonexternal events scenarios at many nuclear power plants Although the methods used
for external events HRA modeling are the same as those used for Level 1 internal initiating events
EPRI 1025294 provides guidance on how to make the relevant selections within the EPRI HRA
Methodology to appropriately account for the impacts of the external events defined in the qualitative
analysis

7 FLOOR HEP FOR HIGH DAMAGE STATES

For extremely high damage states the uncertainties dominate so EPRI report 1025294 [1]
recommends that a floor HEP to reflect the uncertainty associated with the plant damage The floor
HEP is treated as a lower bound If the external events HRA calculates HEPs below the lower bound
then they will not be used and the floor HEP will be used instead Based on accounts of historical
events operational experience data has shown that it is possible for operators to become confused or
distracted by multiple conflicting indications such as spurious instruments or alarms or many failures
caused by a highly damaging event In theory operators should be focused only on the safe shutdown
paths associated equipment and instruments and alarms as directed by the applicable procedures
However in a complicated scenario such as following a spatial event like a seismic event maintaining
this focus might be difficult In addition good reasons might exist for the operators to have a wider
scope of attention (eg secondaryside systems or equipment that is commonly important during
normal operations and systems or equipment of recent concern as a result of current plant
configurations and preexisting conditions)

Sensitivity studies could be conducted to identify whether the applied lower bound limit has little (or
no) effect a significant effect or perhaps a moderate effect Effects might be represented and
evaluated simply as different values of HEPs to represent the HEPs associated with different
conditions for the same HFE

8 MODEL INTEGRATION

Once the HEPs have been quantified at the appropriate level the operator actions and associated HEPs
must be appropriately integrated into the PRA model Model integration consists of various tasks
depending on the PRA model including cut set review HEP reasonableness check recovery
dependency and uncertainty EPRI 1025294 [1] provides guidance on these elements of model
integration in the context of external events

9 INSIGHTS AND CONCLUSIONS

EPRI report 1025294 [1] provides methods and guidance for the human reliability analysis of external
events PRAs EPRI report 1025294 was developed using insights from recent HRA advances from
Fire HRA [2] as well as considering the stateoftheart in seismic and external events PRA models
and issues [3 4] EPRI report 1025294 provides a framework for external events HRA a general
screening approach and a detailed quantification approach which can be applied consistently across a
variety of external events were developed The report was written to provide methods and guidance for Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
all external events but included specific guidance for HRA in a seismic PRA The detailed
quantification approach provided in EPRI report 1025294 is an adaptation of the EPRI HRA
Methodology also known as the EPRI HRA Approach for internal events [5 6 7 and 8]

In 2013 initial testing was conducted on the proposed guidance in EPRI report 1025294 The testing
was limited to pilot plants that were in the process of developing seismic PRA and also to testing the
concepts on plants that were conducting seismic PRA peer review The objective of the testing was to
obtain insights that would be used to refine the external events HRA methods and guidance As part
of the testing a gap analysis was conducted on the ability of current human reliability analysis (HRA)
methods to support current requirements of external flooding risk assessments [11] The gap analysis
started with a review of the requirements from the ASMEANS PRA Standard [12] and the
requirements of the Interim Staff Guidance (ISG) for the Flooding Integrated Assessment [13]
The insights from the gap analysis between the current external flood risk assessment requirements
and the current stateofpractice and the insights from the current testing are summarized below
 The external events HRA process (including external flood and seismic) is identical to the
Fire HRA process which includes a screening step that the internal events PRA (IEPRA)
does not typically require for the IEPRA postinitiator HRA

 External flooding requires a new category of HRA events for external flooding – those
operator actions taken as preventive measures This category includes preparatory
measures such as building isolation This new category of actions should have the same
engineering treatment and modelling considerations as postinitiating event actions

 Feasibility of operator actions applies to all types of operator actions in all hazard groups
including external flooding A strong qualitative analysis represents the best means for
supporting a detailed HRA evaluation in order to demonstrate compliance with PRA
Capability Category II [3] specifically supporting requirement HRG3 for the
incorporation of plantspecific and scenariospecific factors

 Quantitative methods for external flooding actions will likely have difficulty
demonstrating compliance with PRA Capability Category II due to limitations in existing
methods such that it can be difficult to tell if the resultant human error probability is
conservative (Capability Category I) or bestestimate (Capability Category II) for
supporting requirement HRG1

 Quantification of human error probabilities has the following issues related to each
category of operator action 1) Preventive actions – in general qualitative analysis and
quantitative methods need to be developed and refined 2) Postinitiating event actions – in
general qualitative and quantitative methods essentially follow the guidance of NUREG
1921 (Chapter 4) which is being updated in the preliminary EPRI External Events HRA
Guidance document [2]

 Uncertainty and dependency considerations are the same although new sources of
uncertainty are introduced

A pilot of the EPRI 1025294 approach is underway as part of a seismic PRA development At the
same time that this preliminary guidance is being tested and the pilot being conducted EPRI is
updating its methodology for seismic PRA This seismic HRA method will be updated and finalized
based on the lessons learned from this pilot and based on changes in the general seismic PRA
guidance The general objectives of the pilot study are listed below

Probabilistic Safety Assessment and Management PSAM 12 June 2014 Honolulu Hawaii
1 better understand the seismic HRA issues and how they interact with the SPRA
2 test the screening method guidance for usability and reasonableness
3 test the detailed method guidance for usability and reasonableness
4 identify any gaps in the method

EPRI 1025294 [1] is considered to be a preliminary draft as it is expected that this guidance will be
updated in the future The update is expected to include a review of operational experience and
additional guidance specific to other external events (eg External Flooding and High Winds)

10 REFERENCES

[1] A Preliminary Approach to Human Reliability Analysis for External Events with a Focus on
Seismic EPRI Palo Alto CA 2012 EPRI 1025294
[2] EPRINRCRES Fire Human Reliability Analysis Guidelines US NRC Washington DC 2012
NUREG1921EPRI 1023001
[3] PreEarthquake Planning and Immediate Nuclear Power Plant Operator PostEarthquake
Actions US Nuclear Regulatory Commission Regulatory Guide 1166 1997
[4] Guidelines for Nuclear Plant Response to an Earthquake EPRI Palo Alto CA 1989 NP6695
This reference includes the Technical Update published October 2012 under the same title (EPRI
Report 1025288)
[5] Operator Reliability Experiments Using Nuclear Power Plant Simulators EPRI Palo Alto CA
1990 NP6937 as supplemented by EPRI TR 100259 [6]
[6] An Approach to the Analysis of Operator Actions in Probabilistic Risk Assessment EPRI Palo
Alto CA 1992 EPRI TR100259
[7] Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications
(THERP) A D Swain and H E Guttman US NRC Washington DC 1983 NUREGCR1278
[8] Systematic Human Action Reliability Procedure (SHARP) Enhancement Project SHARP1
Methodology Report EPRI Palo Alto CA 1992 EPRI TR101711
[9] A Review of Seismic Operating Experience with Implications for Human Reliability Mary
Presley et al presented at PSA 2013 American Nuclear Society sponsored Probabilistic Safety
Assessment Conference Columbia SC September 2013
[10] A Preliminary Approach to Human Reliability Analysis for External Events with a Focus on
Seismic HRA Mary Presley et al presented at PSA 2013 American Nuclear Society sponsored
Probabilistic Safety Assessment Conference Columbia SC September 2013
[11] Gap Analysis Insights Between External Events HRA Requirements and Current HRA Methods
Mary Presley et al presented at PSA 2013 American Nuclear Society sponsored Probabilistic
Safety Assessment Conference Columbia SC September 2013
[12] Standard for Level 1Large Early Release Frequency Probabilistic Risk Assessment for Nuclear
Power Plant Applications American Society of Mechanical Engineers (ASME) 2009
ASMEANS RASa2009
[13] Guidance for Performing the Integrated Assessment for Flooding United States Nuclear
Regulatory Commission (USNRC) 2012c JLDISG201205 Rev 0 draft September 2012


《香当网》用户分享的内容，不代表《香当网》观点或立场，请自行判断内容的真实性和可靠性！
该内容是文档的文本内容，更好的格式请下载文档