ML23010A191: Difference between revisions

From kanterella
Jump to navigation Jump to search
(StriderTol Bot insert)
 
(StriderTol Bot change)
Line 15: Line 15:


=Text=
=Text=
{{#Wiki_filter:Dependency Analysis Using the Integrated Human Event Analysis System (IDHEAS) HRA Methodology Michelle Kichline Senior Reliability and Risk Analyst U.S. Nuclear Regulatory Commission
{{#Wiki_filter:}}
 
===Background===
* The NRC developed the IDHEAS General Methodology (IDHEAS-G) [1] as a general framework that can be used to create application-specific HRA methods.
* IDHEAS-G includes a new dependency framework that analyzes the dependency between two HFEs by identifying and evaluating how failure of the first human action effects the context of subsequent human actions.
* IDHEAS for Event and Condition Assessment (IDHEAS-ECA) [2] was created for quantifying the HEPs for nuclear power plant PRAs.
* IDHEAS Dependency Model (IDHEAS-DEP) [3] was developed using the new dependency framework in IDHEAS-G and the quantification methodology in IDHEAS-ECA.
2
 
Development of IDHEAS-DEP
* NRC created a working group (WG) with members from RES, NRR, Region 2, and EPRI.
* RES developed a set of dependency relationships and factors and presented them to the WG.
* 6 WG members analyzed an HFE pair for dependency using IDHEAS-ECA and presented their results to the WG.
* The WG discussed each example and how the dependency impacts could be evaluated using IDHEAS-ECA.
* The WG reviewed the dependency factors created by the project team and the draft guidance document.
3
 
Process Overview 4
 
IDHEAS Dependency Types
* Consequential - Occurrence of the preceding HFE (HFE1) changes the context for performing the subsequent HFE (HFE2) from the context that was assumed when the HFE was analyzed without dependency.
* Cognitive - Dependency in the cognitive information for two consecutive HFEs.
* Resource-sharing - The two HFEs share limited resources such as critical tools, staffing, water, or electricity.
5
 
Dependency Relationships IDHEAS-DEP defines five different dependency relationships:
R1 - Functions or Systems R2 - Time Proximity R3 - Personnel R4 - Location R5 - Procedure 6
 
Entry Conditions There are no relevant HFE1 and HFE2 are in the human action success same PRA event sequence or events between HFE1 and minimal cutset HFE2.
OR The initiating event is caused by human actions and is analyzed as the first HFE, such that the subsequent HFEs need to be assessed for dependency.
Proceed to Predetermination (Step 1) 7
 
Predetermination Analysis Dependency Assessment Guidelines Relationship
: 1) HFE1 and HFE2 use the same procedure, AND                                  YES Complete      2) HFE1 is likely to occur because of issues associated with the common      NO Dependency        procedure (such as having an ambiguous or incorrect procedure), AND
: 3) There is no opportunity to recover from the issue with the procedure between HFE1 and HFE2.
R1 -          1) HFE1 and HFE2 have the same functions or systems, OR                      YES Functions or  2) HFE1 and HFE2 have coupled systems or processes that are connected due    NO Systems          to automatic responses or resources needed.
R2 -          1) HFE1 and HFE2 are performed close in time, OR                              YES Time Proximity 2) The cues for HFE1 and HFE2 are presented close in time.                    NO R3 -          1) HFE1 and HFE2 are performed by the same personnel.                        YES Personnel                                                                                    NO R4 -          1) HFE1 and HFE2 are performed at the same location, OR                      YES Location      2) The workplaces for HFE1 and HFE2 are affected by the same condition (such  NO as low visibility, high temperature, low temperature, or high radiation).
R5 -          1) HFE1 and HFE2 use the same procedure.                                      YES 8 Procedure                                                                                    NO
 
Basis for Screening Process
* Dependency relationships between HFEs can result in one or more dependency factors.
* Each dependency factor potentially impacts some PIFs associated with HFE2.
  - The impact of each dependency factor on HFE2 is based on how occurrence of HFE1 changes the context for HFE2.
  - If occurrence of HFE1 would not result in any changes to the context associated with HFE2, the dependency factor may be discounted.
  - Each undiscounted dependency factor potentially results in new PIFs, new PIF attributes, or worsening of the PIF attributes that were originally assessed in the individual HEP of HFE2.
* Some PIF attributes impact HEPs more significantly than others. The screening process focuses on evaluation of the more significant PIF attributes.
* The screening process groups the impact of the most likely affected PIF attributes for each dependency factor into Low, Medium, and High impact categories.
* The dependency impact values (Pd) are based on IDHEAS-ECA.
9
 
Screening Analysis Potential Dependency            Basis for Discounting the Potential Dependency Impact Factors                      Dependency Factor R1.1 Same functions or systems      A - HFE2 was trained in the scenarios    This cognitive dependency affects the PIF for scenario leads to cognitive dependency        that HFE1 occurs (e.g., Feed & Bleed    familiarity, which addresses the mental model. Scenario is the last action after others fail) so familiarity is applicable when there is something wrong with A. Occurrence of HFE1 leads to        there is no unfamiliarity due to HFE1. the mental model and there is no diverse method available to the scenario or parts of the    B - HFE2 is well trained on in various  correct the wrong mental model.
scenario being different from    scenarios such that personnel are        Low: Pd = 5E-2 what was typically trained;      unlikely to develop a wrong mental        Parts of scenario become unfamiliar (e.g., different from thus, the scenario associated    model due to occurrence of HFE1.            what was trained on), OR with HFE2 becomes less          A/B - There is no cognitive link          HFE1 creates a biased mental model or preference for familiar. (Note: Occurrence      (similar thought process) between            wrong strategies.
of HFE1 alters the scenario      the two HFEs; thus, occurrence of Medium: Pd = 1E-1 for HFE2; thus, HFE1 causes      HFE1 has no impact on scenario Parts of scenario become unfamiliar (e.g., different from some level of unfamiliarity      familiarity or mental model what was trained on), AND with HFE2)                        associated with HFE2.
HFE1 creates a biased mental model or preference for B. Occurrence of HFE1 leads to      B - There are opportunities between wrong strategies.
an incorrect or biased mental    the HFEs to break the incorrect model of the situation            mental model, such as multiple          High: Pd = 3E-1 associated with HFE2.            crews or diverse cues.                    HFE1 creates a mismatched or wrong mental model for HFE2 due to close cognitive links between HFE1 and HFE2 10                                                                            (e.g., thought process).
 
Screening Dependency Impact Calculate the screening dependent HEP of HFE2 by taking the probabilistic sum of the individual HEP of HFE2 (P2) and each of the undiscounted dependency impact values (Pd), as follows:
Dependent HEP of HFE2 = 1  1  1    =1 1      1      1 NOTE: When the dependency impact values are small, the screening dependent HEP can be approximated by summing the dependency impact values and the individual HEP of HFE2. This approximation should not be used when any High dependency impact values are applicable.
11
 
Dependency Impact HFE          Potential Dependency No Relationship                Factors                  Impact        Low          Medium          High R1 Functions or R1.1 Same functions or systems leads to cognitive dependency                  0.0          5E-2            1E-1          3E-1        Potential R1.2 Same functions or systems leads systems      to consequential dependency R1.3 Same functions or systems leads 0.0          1E-2            5E-2          2E-1 Dependency to resource-sharing dependency            0.0          2E-3            1E-2          5E-2 R2 Time R2.1 Close time proximity in performing HFE1 and HFE2 leads to Varies depending on the ratio of time available to time required (Ta/Tr) for performing HFE2 Impacts consequential dependency                  >4        3 and  4  2 and < 3        1 and < 2 proximity                                              0.0        1E-3            1E-2          1E-1 R2.2 Close time proximity in receiving the cues for HFE1 and HFE2 leads to      0.0          5E-3            5E-2          1E-1 consequential dependency R3          R3.1 Same personnel leads to cognitive dependency 0.0          5E-2            1E-1          3E-1 Personnel    R3.2 Same personnel leads to consequential dependency                  0.0          2E-3            1E-2          3E-2 R3.3 Same personnel leads to resource-sharing dependency              0.0          2E-3            1E-2          5E-2 R4          R4.1 Same location leads to consequential dependency                  0.0          2E-3            5E-3          2E-2 Location    R4.2 Same location and time leads to consequential dependency                  0.0          2E-3            5E-3          7E-3 R5          R5.1 Same procedure leads to 12        cognitive dependency                      0.0          5E-3            5E-2        3.5E-1 Procedure
 
Detailed Analysis Basis for Discounting the Potential Potential Dependency Factors                                                                      Dependency Impact Dependency Factor R1.1 Same functions or systems leads to    A - HFE2 was trained in the scenarios that  Potentially affected CFMs:
cognitive dependency                        HFE1 occurs (e.g., Feed & Bleed is the last All CFMs action after others fail) so there is no A. Occurrence of HFE1 leads to the          unfamiliarity due to HFE1.                  Potentially impacted PIFs:
scenario or parts of the scenario      B - HFE2 is well trained on in various      SF - Scenario Familiarity being different from what was          scenarios such that personnel are unlikely typically trained; thus, the scenario  to develop a wrong mental model due to      PIF attributes that are most likely impacted by associated with HFE2 becomes less      occurrence of HFE1.                        the dependency factor:
familiar. (Note: Occurrence of HFE1    A/B - There is no cognitive link (similar    SF1 - Unpredictable dynamics in known alters the scenario for HFE2; thus,    thought process) between the two HFEs;          scenarios HFE1 causes some level of              thus, occurrence of HFE1 has no impact on    SF2 - Unfamiliar elements in the scenario unfamiliarity with HFE2)                scenario familiarity or mental model        SF3 - Scenarios trained on but infrequently B. Occurrence of HFE1 leads to an          associated with HFE2.                            performed incorrect or biased mental model of    B - There are opportunities between the      SF4 - Bias or preference for wrong the situation associated with HFE2. HFEs to break the incorrect mental model,        strategies exists, mismatched mental such as multiple crews or diverse cues.          models 13
 
Simplified Example
* Inadvertent safety injection (SI) actuation due to a stuck open pressurizer spray valve at power.
* SI triggers a reactor trip. All automatic actions occur as expected.
14
 
Human Failure Events
* HFE 1: Terminate SI
  - Performed using procedure ES-1.1 and takes 15 min to complete.
  - E-0 is entered due to the reactor trip. E-0 directs entering ES-1.1 to terminate SI flow if RCS pressure is stable or rising and other parameters are within criteria.
* HFE 2: Start high pressure recirculation (HPR)
  - Performed using procedure ES-1.3 and takes 12 min to complete.
  - E-0 directs entering ES-1.3 when RWST level reaches 30%. ES-1.3 directs operators to start HPR once RWST level drops to 20%.
15
 
Timeline T=0 SI Actuation              T = 5.5 hrs                T = 8 hrs Reactor Trip            RWST at 30%              Core Damage T = 30 min                T = 6.5 hrs Pressurizer              RWST at 20%
filled with water
* 5.5 hours available to terminate SI.
* 1.5 hours to start high pressure recirculation.
16
 
Relationship                      Assessment Guidelines Complete        HFE1 and HFE2 use the same procedure, AND                    YES Dependency      HFE1 is likely to occur because of issues associated with the Step 1:
NO common procedure (such as having an ambiguous or incorrect procedure), AND There is no opportunity to recover from the issue with the procedure between HFE1 and HFE2.
Justification: (1) HFE1 uses procedures E-0 and ES-1.1.
HFE2 uses procedures E-0 and ES-1.3, or E-1 and ES-1.3. E-0 Predetermination is the shared procedure. (2) The cues of HFE1 and HEF2 are R1 Functions or different. The common procedure is not a likely cause of dependency.
HFE1 and HFE2 have the same functions or systems, OR          YES Analysis HFE1 and HFE2 have coupled systems or processes that          NO Systems        are connected due to automatic responses or resources needed.
Justification: HFE1 and HFE2 use the same system (ECCS) for different functions. HFE1s function is to control RCS inventory. HFE2 provides core cooling.
R2            HFE1 and HFE2 are performed close in time, OR                YES Time Proximity  The cues for HFE1 and HFE2 are presented close in time.      NO R3            HFE1 and HFE2 are performed by the same personnel.            YES Personnel                                                                      NO Justification: HFE1 and HFE2 are assumed to be performed by the same crew.
R4            HFE1 and HFE2 are performed at the same location, OR          YES Location        The workplaces for HFE1 and HFE2 are affected by the          NO same condition (such as low visibility, high temperature, low temperature, or high radiation).
Justification: HFE1 and HFE2 are performed in the main control room. No environmental factors are expected to affect the reliability of HFE1 and HFE2.
R5            HFE1 and HFE2 use the same procedure.                        YES Procedure                                                                      NO Justification: HFE1 and HFE2 could share the same procedure 17                (E-0).
 
Step 2:
Screening Analysis -
Same Personnel
* Individual HEP = 3.5E-3
* Pd(R3.2) = 2E-3
* Screening dependent HEP = 5.5E-3 18
 
Potential Dependency Factors R3.2 Use of the same Dependency Impact Potentially affected CFMs:
Step 3:
personnel leads to                All CFMs consequential dependency A. Mental fatigue, time Potentially impacted PIFs:
MFMental fatigue, stress, and time pressure Detailed Analysis -
pressure, or stress level increase due to the same personnel performing MTMultitasking, interruptions, and distractions PIF attributes that are most likely impacted by the Same Personnel HFE1 and HFE2 (e.g.,          dependency factor:
HFE1 could cause high          MF1Sustained (> 30 minutes), highly demanding stress or mental fatigue          cognitive activities requiring continuous attention because several layers of defense in depth have (e.g., procedure-situation mismatches demand
* Individual HEP = 3.5E-3 constant problem-solving and decisionmaking; failed, such as in situations beyond the EOPs).
information changes over time and requires sustained attention to monitor or frequent checking)
* The dependency impact was B. Personnel need to perform MF2Time pressure due to perceived time urgency          accounted for in the original MF3Lack of self-verification due to rushing task HFE1 and HFE2 at the same time (i.e., personnel completion (speed-accuracy tradeoff)                  analysis of the individual HEP.
MF5Long working hours (> 4 hours) with highly must switch between tasks).                          cognitively demanding tasks
* HEP with dependency = 3.5E-3 Note: The analyst added MF5 to the list.
MT3Concurrent visual detection and other tasks
* HEP without dependency MT4Concurrent auditory detection and other tasks MT5Concurrent diagnosis and other tasks                applied = 2.6E-3 MT8Concurrently executing action sequence and performing another attention/working memory task
* Pd(R3.2) = = 9E-4 MT9Concurrently executing intermingled or interdependent action plans 19
 
Example Results
* Original individual HEP = 3.5E-3
* Screening dependent HEP = 5.5E-3
  - Screening consequential dependency impact due to use of same personnel: Pd(R3.2) = 2E-3
* Detailed dependent HEP = 3.5E-3
  - Detailed consequential dependency impact due to use of same personnel: Pd(R3.2) = = 9E-4
  - Individual HEP without dependency impact = 2.6E-3 20
 
Conclusion
* IDHEAS-DEP leverages the IDHEAS-ECA quantification method to develop screening dependency values that can be applied to HEPs for HFEs modeled using any HRA method.
* Both the screening analysis and detailed analysis provide a dependent HEP for HFE2 and identify the dependency relationships, dependency factors, and PIFs impacted by the occurrence of HFE1
* In general, the dependent HEPs calculated using IDHEAS-DEP are lower than those calculated using THERP.
* In some cases, the independent HEP may already account for the dependency impacts (such as for feed and bleed or initiating high pressure recirculation).
21
 
Questions?
22
 
Acronyms
* CFM - Cognitive Failure Mode
* IDHEAS-G - IDHEAS General
* EPRI - Electric Power Research        Methodology Institute
* NRC - Nuclear Regulatory Commission
* HEP - Human Error Probability
* NRR - Office of Nuclear Reactor
* HFE - Human Failure Event            Regulation
* HRA - Human Reliability Analysis
* PIF - Performance Influencing Factor
* HPR - High Pressure Recirculation
* PRA - Probabilistic Risk Analysis
* IDHEAS - Integrated Human Event
* RCS - Reactor Coolant System Analysis System
* RES - Office of Nuclear Regulatory
* IDHEAS-DEP - IDHEAS Dependency        Research Analysis
* SI - Safety Injection
* IDHEAS-ECA - IDHEAS for Event and
* WG - Working Group Condition Assessment 23
 
References
: 1. U.S. Nuclear Regulatory Commission, NUREG-2198, The General Methodology of an Integrated Human Event Analysis System (IDHEAS-G),
May 2021, ADAMS Accession No. ML21127A272.
: 2. U.S. Nuclear Regulatory Commission, NUREG-2256, Integrated Human Event Analysis System for Event and Condition Assessment (IDHEAS-ECA), October 2022, ADAMS Accession No. ML22300A117.
: 3. U.S. Nuclear Regulatory Commission, Research Information Letter 2021-14, Integrated Human Event Analysis System Dependency Analysis Guidance (IDHEAS-DEP), November 2021, ADAMS Accession No. ML21316A107. (Will be NUREG-2258) 24}}

Revision as of 12:49, 15 November 2024

IDHEAS-DEP for EPRI HRA Ug Jan 2023
ML23010A191
Person / Time
Issue date: 01/10/2023
From: Michelle Kichline
NRC/NRR/DRA/APOB
To:
References
Download: ML23010A191 (24)
v • d • e


Text

Retrieved from "https://kanterella.com/w/index.php?title=ML23010A191&oldid=3715287"