ML19092A470
| ML19092A470 | |
| Person / Time | |
|---|---|
| Issue date: | 03/14/2019 |
| From: | Chang Y NRC/RES/DRA/HFRB |
| To: | |
| J. Chang 415-2378 | |
| Shared Package | |
| ML19092A461 | List: |
| References | |
| Download: ML19092A470 (17) | |
Text
SACADA Cicada (from Google images)
Status and the Data Available to Public Y. James Chang, Ph.D.
Human Factors and Reliability Branch Division of Risk Analysis Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Presented at the HRA Data Workshop Rockville, MD March 14-15, 2019 1
Current SACADA Work Emphases
- Outreach to more plants to use SACADA
- Develop SACADA-2
- Make SACADA data available to HRA community SACADA: Scenario Authoring, Characterization, and Debriefing Application 2
Outreach
- STP continues to use SACADA-1
- Will migrate to SACADA-2 when ready
- Votgle 3&4 is piloting SACADA-2 since 1/2019
- Evaluate SACADA-2 for three training cycles
Develop SACADA-2
- The development was motivated by the industrys feedback on SACADA-1
- To have all operator performance information in one place, i.e.,
simulator training, job performance measure, written tests, and actual operations
- Able to track the performance of an individual operator
- Key other improvements
- Significantly reduced IT demand - Do not need SQL
- Comprehensive incorporation of INPO taxonomy for debriefing
- Improving debriefing taxonomy. May shades light on modeling error recovery.
- STP and Vogtle 3&4 instructors support the development
- Aim to have all SACADA-2 functions implemented in summer 2019 4
SACADA-2 Authoring Screen Shot 17
SACADA-2 Characterization Screen Shot 22
Debriefing Screen Shots Evaluate the Performance Link to INPO Operator Fundamentals 30
Debriefing Screen Shot Understand the Deficiencies
- Seven classes
- Monitoring/Detection
- Diagnosis/Understanding
- Procedure Following/Decisionmaking
- Manipulation
- Supervision
- Teamwork
- Communication
- Stimulate crew to think
- Can choose more than one class 32
Reports Exportable formats: PDF, MS Word, and Excel 44
SACADA-1 vs. SACADA-2 STP and Vogtle Electric Generating Plant, Units 3 and 4, Provide Technical Support for SACADA-2 Development SACADA-1 SACADA-2 Collect Simulator Data Collect Job Performance Measure (JPM) Data Collect Written Exam Results Directly Support Crew Notebook Provide Simulator Guide Track Crew Performance (Simulator only) (All areas)
Track Individual Performance (All areas)
All areas: Simulator, JPM, and written exams.
10
- Basic human error probabilities (HEPs)
- Performance influencing factors (PIFs) effect
- Error recovery effect
- Task dependency effect
- Types of data and preliminary assessment
- Basic HEPs - Okay to Good
- Methods are demonstrated by Drs. Pamela Nelson and Ali Azarm in 2018
- A key uncertainty is the success criteria difference between UNSAT and human err in HRA
- PIFs effect - Good
- Demonstrated by Drs. Jing Xing and Song-Hua Shen in this workshop
- Error recovery effect - Maybe
- Havent analyzed the data for error recovery
- Task dependency effect - Maybe
- Training data may not help. Exam data and experiment data could help.
- Compare with HuREX results
- Still to be done 11
Make SACADA Data Available to HRA Community
- Abide the NRC agreements with the data providers on SACADA data publications
- The NRC is sensitive to the data providers concern about releasing data to the HRA community
- Only release system-neutral statistics for HRA
- Alarm detection example:
Context Performance Statistics 12
Planned Data/Information to be Available
- Statistics of context and performance in various macro-cognitive functions
- Detecting alarms, reading parameters, diagnosis, deciding, actions, and external communication
- Statistics of context and failure modes (and the system-neutral performance information of the failure modes)
- Do not plan to release the system-specific statistics, e.g., crew performance of training objective elements and of malfunctions 13
Data Distribution 14
High Percentage of Non-characterized Data Points 15
Identified Some Reasons but Not All Reasons Malfunctions/Conditions Initial Condition 15 50 80 100 120 Time Fire in No. 12 Condensate Pump Motor US Enter POP04-ZO-0008, Fire/Explosion Elements (Training Activate the fire alarm, make the announcement, CREW Objective and call out the Fire Brigade Elements)
CREW Other item to discuss Not-characterized 16
Conclusion
- SACADA has made progress in
- Reaching out to more potential users
- Improving usability for operator training
- Demonstrating the use of data for HEP estimates
- Expanding the scope of data collection
- Establishing a long-term data collection program
- Tasks ahead
- Continue the existing activities
- Ensure data quality and make data available to public
- Compare human performance indications with the other empirical data 17