ML24302A087

From kanterella
Jump to navigation Jump to search
HRA Data Workshop 2024-10-04 NRC
ML24302A087
Person / Time
Issue date: 10/06/2024
From: Chang Y
NRC/RES/DRA/HFRB
To:
References
Download: ML24302A087 (1)


Text

HRA Data R&D in the NRC Yung Hsien James Chang, Ph.D.

Senior Reliability and Risk Analyst U.S. Nuclear Regulatory Commission Presented at HRA Data Workshop Sendai, Japan October 6, 2024

2 Presentation Outline

  • Overview of recent NRC HRA R&D activities
  • Near-term focuses on HRA data
  • An example of data generation and integration
  • Opportunities for collaborations

3 Overview of Recent NRC HRA R&D Activities

- Complete IDHEAS Suite Development Function Products High-Level HRA guidance IDHEAS-General Methodology (NUREG-2198)

HEP quantification IDHEAS-ECA (NUREG-2256); IDHEAS-AT Power (NUREG-2199);

IDHEAS-Dependency (RIL 2021-14); Error Recovery (TBD);

Time Uncertainty (TBD); and HEP Uncertainty (TBD);

Expert elicitation (NUREG-2255, draft)

Data basis IDHEAS-DATA (in process) and SACADA, Guidance on data generalization and integration.

Software tool IDHEAS-ECA tool (present status: rev. 1.2)

Quality control IDHEAS-ECA desktop guide (in development); independent reviews IDHEAS-ECA Evaluation of SPAR HFEs (in process) - also demonstrate documentation.

NUREGs are available in https://www.nrc.gov/reading-rm/doc-collections/nuregs/staff/index.html.

Research Information letters (RIL) are available in https://www.nrc.gov/reading-rm/doc-collections/research-info-letters/index.html.

IDHEAS Suite

4 Recent HFRBs HRA R&D Products

  • HRA R&D five-year plan

Time and Time Uncertainty

  • RIL 2024-04 Human Reliability Analysis for Calculating Effects of Performance Influencing Factors: A Review of Historical Approaches
  • RIL 2024-05 Simulator Data Analysis to Inform Digitalized Environment Impacts on Human Reliability
  • RIL 2024-11 Use of Expert Elicitation Guidance to Inform a Small-Scale Knowledge Judgment Application
  • NUREG-2255 Expert Elicitation Guidance (Coming soon)

5 Near-Term Focuses on HRA Data

  • Publish IDHEAS-DATA reports
  • PNNL review of the draft IDHEAS-DATA (A RIL)
  • Develop IDHEAS-ECA desktop guide
  • To improve the reliability and consistency of applying IDHEAS-ECA method
  • Enhance guidance on the selections of Cognitive Failure Modes (CFMs), performance influencing factor attributes (PIFAs), and error recoveries
  • Discuss how the original data of IDHEAS-DATA were generated to support analysts judgments in applying IDHEAS-ECA when needed
  • Add and update data
  • Continue to explore opportunities to collect HRA data

Data Generalization and Integration in IDHEAS-DATA 6

7 An Example of Data Generalization and Integration

  • Performance Influencing Factor Attribute (PIFA) SF3 Scenario is unfamiliar
  • Applicability
  • Scenario is unfamiliar
  • A bias or presence for wrong strategies exists
  • Personnel are unfamiliar with system failure modes
  • Personnel are unfamiliar with worksite for manual actions
  • Plans, policies, and procedures to address the situation are lacking
  • Degrees of unfamiliarity - current guidance
  • SF3.1: Scenario is trained on but infrequently performed
  • SF3.2: Scenario is unfamiliar and rarely performed, e.g.,

Notice adverse indicators that are not part of the task at hand Notice incorrect status that is not a part of the routine tasks

  • SF3.3: Scenario is extremely rarely performed, e.g.,

Lack of plans, policies and procedures to address the situation No existing mental model for the situation Rare and unexpected events

8 Selection Detection HEP Understanding HEP Decisionmaking HEP Action Execution 1

1.0E-03 1.0E-02 1.0E-02 1.0E-03 2

3.8E-03 3.3E-02 3.3E-02 9.0E-03 3

6.5E-03 5.5E-02 5.5E-02 1.7E-02 4

9.3E-03 7.8E-02 7.8E-02 2.5E-02 5

1.2E-02 1.0E-01 1.0E-01 3.3E-02 6

1.6E-02 1.4E-01 1.4E-01 9.6E-02 7

2.0E-02 1.8E-01 1.8E-01 1.6E-01 8

2.5E-02 2.2E-01 2.2E-01 2.2E-01 9

2.9E-02 2.6E-01 2.6E-01 2.9E-01 10 3.3E-02 3.0E-01 3.0E-01 3.5E-01

SF 3 Data Sources (1/2)

- In IDHEAS-DATA 9

SF 3 Data Sources (2/2)

- In IDHEAS-DATA 10

[26] SACADA - Scenario Structure and Context Characterization 11

[26] SACADA Situational Factors Specific to Diagnosis/Understanding 12 Diagnosis Basis:

o Skill o

Procedure o

Knowledge Based Familiarity:

o Standard: Crew has previously trained on this challenge.

o Novel: This involves a change to the way the challenge is addressed, such as a new procedure, scenario, or role.

o Anomaly: Standard training must be adapted to fit an anomalous situation (e.g., the procedures do not cover the circumstances.)

Outcome:

o Procedure: The diagnosis results in performing a procedure step or transitioning within or external to the procedure currently being performed. (Specify) Procedure ID:_____ Step:______

o Skill: Diagnosis results in the performance of a skill-based activity.

o Knowledge Based: Diagnosis results in the performance of a knowledge-based activity.

Information Integration

Timing of Information: Includes slow information feed or delayed information.

Ambiguous Information: Information provided by system is vague, unclear, or does not point to the nature of the problem.

Integration Required: Integration of multiple pieces of information is required for diagnosis.

Information Specificity:

o Specific: Alarm/alarm pattern/indication(s) point to the specific system problem.

o Not Specific: The alarm(s) and/or indication(s) do not directly point to the specific system problem, which requires operator cognitive effort to integrate the information and identify the specific system problem.

o Not Applicable.

Information Quality:

Missing Information: include masked information

Misleading Information: Information points to an incorrect diagnosis

Conflicting Information: Information points to more than one possible diagnoses, or conflicts with other alarms or indications.

Others (specify):____________________

[26] SACADA Situational Factors Specific to Decisionmaking 13 Situational Factors and Optional Statuses Decision Basis:

o Procedure: The decision is driven by procedures or other guidance.

o Skill: Skill-driven decision: without procedure, operator can make decision from memory.

o Knowledge: No procedure applicable; crew relies on engineering or technical knowledge and operating experience.

Familiarity:

o Standard: Crew has previously trained on this challenge.

o Novel: This involves a change to the way the challenge is addressed, such as a new procedure, scenario, or role.

o Anomaly: Standard training must be adapted to fit an anomalous situation (e.g., the procedures do not cover the circumstances.)

Uncertainty:

o Clear: No uncertainty or competing goals - clear decision criteria.

o Uncertain: Lack of information or ambiguous decision criteria.

o Competing Priorities: Multiple competing goals, foreseeable severe consequences.

o Conflicting Guidance: Policies, practices and procedure have conflicting guidance.

Other (Specify):__________________

[26] Diagnosis-Specific Context Characterization 14

[26] Overarching Context Characterization 15

[26] Performance Statistics 16

[26] Causal Analysis 17

[26] Use SACADA Data to Inform Desktop Guide Development 18

[4, 5] are Two RESS Papers (Part I and II) 19

[4, 5] Example Data 20

[4, 5] Decisionmaking/Understanding Data Points

- in IDHEAS-DATA and IDHEAS-ECA 21 Varies, [4] Table 7 (Decisionmaking)

EOCs due to cognitive errors in identifying or defining the task.

- 1.1E-3 (1/888) Frequently performed task

- Task remembered incorrectly.

- No additional error forcing factor.

- 2.3E-2 (4/173) Rarely performed task

- Task remembered incorrectly.

- No additional error forcing factor.

- 1.4E-1 (3/22) Rarely performed task

- Task remembered incorrectly.

- Additional error forcing factors present.

SF3: Infrequently performed scenarios Basic HEP 1.00E-03 Selection Decisionmaking HEP 1

1.0E-02 2

3.3E-02 3

5.5E-02 4

7.8E-02 5

1.0E-01 6

1.4E-01 7

1.8E-01 8

2.2E-01 9

2.6E-01 10 3.0E-01 In IDHEAS-DATA Implemented in IDHEAS-ECA

[4, 5] Action Execution Data Points

- In IDHEAS-DATA and IDHEAS-ECA 22 Varies, Table 9 [5] (Action Execution) (This table replaces [4] Table 8).

- EOOs, Carrying out a sequence of tasks

- Memorized task step not remembered.

- 1.0E-3 (3/3067) Frequently performed, no error forcing factor.

- 2.1E-2 (1/48) Rarely performed, no error forcing factor.

- 1.6E-2 (3/185) Rarely performed, moderately high level of stress.

- 4.9E-2 (2/41) Rarely performed, moderately high level of stress, ergonomically deficient work environment.

- 7.8E-2 (1/13) Rarely performed test procedure consisting of many sub-steps, task is part of professional knowledge and has to be remembered, high task load.

- 1.4E-1 (1/7) Rarely performed, moderately high level of stress, error prone PSFs and dynamic work environment.

- 3.3E-1 (1/3) Extremely rarely performed, no error promoting factors 6.6E-5 (1/15200) #58

- Pulling and replacing a simulation pin on an electronic module front cover in a control cabinet

- Replunging omitted.

- Highly trained task, not part of a written procedure but part of professional knowledge, favorable ergonomic design.

SF3: Infrequently performed scenarios Basic HEP 1.00E-04 Selection Action Execution HEP 1

1.0E-03 2

9.0E-03 3

1.7E-02 4

2.5E-02 5

3.3E-02 6

9.6E-02 7

1.6E-01 8

2.2E-01 9

2.9E-01 10 3.5E-01 In IDHEAS-DATA Implemented in IDHEAS-ECA

[4, 5] Data Generalization and Integration (1/2)

- IDHEAS-DATA vs. Original Data 23 Varies, Table 9 [5] (Action Execution) (This table replaces [4] Table 8).

- EOOs, Carrying out a sequence of tasks

- Memorized task step not remembered.

- 1.0E-3 (3/3067) Frequently performed, no error forcing factor. (#30, #59, and #64)

- 2.1E-2 (1/48) Rarely performed, no error forcing factor. (#35)

- 1.6E-2 (3/185) Rarely performed, moderately high level of stress. (#27, #34, and #65)

- 4.9E-2 (2/41) Rarely performed, moderately high level of stress, ergonomically deficient work environment. (#31 and #60)

- - 1.4E-1 (1/7) Rarely performed, moderately high level of stress, error prone PSFs and dynamic work environment. (#28)

- 3.3E-1 (1/3) Extremely rarely performed, no error promoting factors. (#66) 6.8E-4 (1/1470) #30

- Manually operating a local valve

- Valve not operated, step in a sequence of different steps not remembered.

- Frequently performed task.

7.42E-4 (1/1347) #59

- Operating a manual control at a MCR control panel.

- Task not remembered.

- Frequently performed task, part of professional knowledge, position of indicator lamp ergonomically unfavorable designed.

4.0E-3 (1/250) #64 (Action Execution)

- Disassembly of a circuit breaker at a local switch gear cabinet.

- Cable not disconnected, sub-step in a task consisting of four sub-steps not remembered.

- Memory based activity, very slow variation of process parameters, high task load.

2.1E-2 (1/48) #35 (Action Execution)

- Removing a ground connection from a switch cabinet.

- Ground connection not removed, memorized step not remembered.

- Rarely performed task, step not described in written procedure.

In IDHEAS-DATA Original Data

[4, 5] Data Generalization and Integration (2/2)

- Original Data 24 2.0E-2 (1/50) #27 (Action Execution)

- Repeated discontinuous manual control od pump discharge pressure.

- Task not remembered.

- Moderately high level of stress.

5.0E-2 (1/20) #34 (Action Execution)

- Open a valve by MCR panel control.

- Failed to open, memorized task step not remembered.

- Rarely performed task sequence, moderately high level of stress.

8.7E-3 (1/115) #65 (Action Execution)

- Discontinuous check of a process parameter on a MCR control panel.

- Check not performed, task not remembered.

- Memory based activity, very slow variation of process parameter, high task load.

3.6E-2 (1/28) #31 (Action Execution)

- Open a key control.

- Control not operated, step not remembered.

- Rarely performed task, enhanced level of stress due to time constraints, unfavorable ergonomic design.

7.8E-2 (1/13) #60 (Action Execution)

- Operating a circuit breaker at a local switchgear cabinet.

- Task not remembered.

- Rarely performed test procedure consisting of many sub-steps, task is part of professional knowledge and has to be remembered, high task load.

1.4E-1 (1/7) #28 (Action Execution)

- Carrying out a sequence of tasks.

- Memorized task step not remembered.

- Rarely performed, moderately high level of stress, error prone PSFs and dynamic work environment.

3.3E-1 (1/3) #66 (Action Execution)

- Disabling an automation in order to perform a test.

- Automatic operation not disabled, step not remembered.

- High demand on memory performance, rarely performed task.

25 Considerations in data generalization and Integration on [26] and [4, 5]

  • The HEPs in SACADA [26] and nuclear maintenance [4, 5]

are the tasks performed by a team. The data includes individual and team error recoveries.

  • The error mode in [4, 5] is memorized task step is not memorized. EOPs provide step-by-step instructions that makes the error mode less likely.
  • Inform the analysts about the considerations

26 IDHEAS-ECA Desktop Guide

  • Provide more details of the original data than documented in IDHEAS-DATA to support analysts applying IDHEAS-ECA method
  • Plan to incorporate the desktop guide into IDHEAS-ECA software tool

27 Opportunities for Collaborations

  • The NRC welcomes collaborations in HRA method, data, and tool
  • The NRC can provide free applications, source codes, and technical support of IDHEAS-ECA and SACADA
  • Collaborations required to go through the NRC approval process (e.g.,

establish a memorandum of understanding (MOU))

  • NRCs HRA data needs
  • Multiple sites: Existing data is from a nuclear site
  • BWR data: Existing data is from a PWR site
  • Field performance data: Existing data only contains control room actions. Field data, e.g., job performance measures, would expand the coverage
  • Timing data: to assess time uncertainty and the drivers