Text

Response to Request for Additional Information Regarding Proposed Technical Specifications (TS) Change TS-505 License Amendments - Extended Power Uprate (EPU) - Supplement 29
	ML16217A161
Person / Time
Site:	Browns Ferry
Issue date:	08/03/2016
From:	James Shea; Tennessee Valley Authority
To:	; Document Control Desk, Office of Nuclear Reactor Regulation
References
	CAC MF6741, CAC MF6742, CAC MF6743, CNL-16-129
	Download: ML16217A161 (152)
	v • d • e

Tennessee Valley Authority, 1101 Market Street, Chattanooga, Tennessee 37402 CNL-16-129 August 3, 2016 10 CFR 50.90 ATTN: Document Control Desk U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 Browns Ferry Nuclear Plant, Units 1, 2, and 3 Renewed Facility Operating License Nos. DPR-33, DPR-52, and DPR-68 NRC Docket Nos. 50-259, 50-260, and 50-296

Subject:

Proposed Technical Specifications (TS) Change TS-505 - Request for License Amendments - Extended Power Uprate (EPU) - Supplement 29, Response to Request for Additional Information

References:

1. Letter from TVA to NRC, CNL-15-169, "Proposed Technical Specifications (TS) Change TS-505 - Request for License Amendments - Extended Power Uprate (EPU)," dated September 21, 2015 (ML15282A152)

2. Letter from NRC to TVA, Browns Ferry Nuclear Plant, Units 1, 2, and 3 -

Request for Additional Information Related to License Amendment Request Regarding Extended Power Uprate (CAC Nos. MF6741, MF6742, and MF6743), dated July 21, 2016 (ML16194A229)

By the Reference 1 letter, Tennessee Valley Authority (TVA) submitted a license amendment request (LAR) for the Extended Power Uprate (EPU) of Browns Ferry Nuclear Plant (BFN) Units 1, 2 and 3. The proposed LAR modifies the renewed operating licenses to increase the maximum authorized core thermal power level from the current licensed thermal power of 3458 megawatts to 3952 megawatts. The Reference 2 letter provided an Nuclear Regulatory Commission (NRC) Request for Additional Information (RAI) related to probabilistic risk assessment. The due date for the response to NRC RAI APLA-RAI 09.01, provided by the Reference 2 letter, is August 12, 2016. The enclosure to this letter provides the response to NRC RAI APLA-RAI 09.01.

U.S. Nuclear Regulatory Commission CNL-16-129 Page 2 August 3, 2016 TVA has reviewed the information supporting a finding of no significant hazards consideration and the environmental consideration provided to the NRC in the Reference 1 letter. The supplemental information provided in this submittal does not affect the bases for concluding that the proposed license amendment does not involve a significant hazards consideration. In addition, the supplemental information in this submittal does not affect the bases for concluding that neither an environmental impact statement nor an environmental assessment needs to be prepared in connection with the proposed license amendment.

Additionally, in accordance with 10 CFR 50.91 (b)(1 ), TVA is sending a copy of this letter to the Alabama State Department of Public Health.

There are no new regulatory commitments associated with this submittal. If there are any questions or if additional information is needed, please contact Edward D. Schrull at (423) 751-3850.

I declare under penalty of perjury that the foregoing is true and correct. Executed on the 3rd day of August 2016.

Respectfully,

~~~

J. W. Shea Vice President, Nuclear Licensing

Enclosure:

Response to NRC Request for Additional Information APLA-RAI 09.01 cc:

NRC Regional Administrator - Region II NRC Senior Resident Inspector - Browns Ferry Nuclear Plant State Health Officer, Alabama Department of Public Health

ENCLOSURE Response to NRC Request for Additional Information APLA-RAI 09.01

ENCLOSURE E-1 APLA-RAI 09.01 In RAI No. APLA-RAI 09, dated April 14, 2016, the NRG staff requested the licensee to explain how several human error probabilities (HEPs) were quantified for both the extended power uprate (EPU) and CL TP in order to confirm the reasonableness of these calculations and to understand why the increase in some HEPs are much larger than others. In response to APLA-RAI 09 dated April 27, 2016, the licensee discussed the quantification of these HEPs for both the EPU and CLTP. The response stated for human failure event (HFE)

HFFAOASD_RCIC that "the dependency levels are higher [a medium dependency was assigned for the CL TP case and a high dependency was assigned for the EPU case] for the abandonment action due to the longer execution time and stress levels." It is not clear why the HEP for HFFAOASD_RCIC is increased by only a factor of 1.15 at the CL TP level to the EPU level, while HEP for HFEs HFA_0002RPV_LVL and HFA_OHCllNIT30 are increased by much larger factors (i.e., a factor of 2.76 and 2.04, respectively).

The staff observed that the execution analysis in the Electric Power Research Institute (EPRI) human reliability analysis (HRA) calculation sheet for HFFAOASD_RCIC includes 9 procedure steps, with recovery considered only in procedure step 1. Procedure steps 2 through 9, which are considered "high" stress, did not credit recovery because of an assumption made for the HEP of each step (i.e., "it is better to use item Ref. 8a [that is a HEP value of 2.?E-04] even though it is for normal stress and then not credit a recovery step"). Not crediting recovery in procedure steps 2 through 9 for HFFAOASD_RCIC causes the HEPs associated with these steps to be the same between the CL TP and EPU. This results in a smaller increase in total HEP for HFFAOASD_RCIC, between the CLTP and EPU cases, than that had recovery been credited, potentially leading to underestimating the change in risk. When recovery is credited in each procedure step, the EPU HEP for HFFAOASD_RCIC could be 3 times larger (not 1.15 times larger as currently indicated by Attachment 44 of the EPU LAR) than that for the CL TP case. This issue may also apply to other HFEs considered in the EPU risk evaluation, such as those HFEs identified in Part 1 below.

The staff requests the licensee to address the following:

a. Provide the detailed HRA calculation sheets (e.g., as generated by the EPRI HRA calculator) for the following HFEs for both the CL TP and EPU cases:

HFA_0003PMP START ("Operator fails to restart RFW [reactor feedwater] after Level 8 trip")

HFA_0071 L8RESTART ("Operator fails to restart RCIC [reactor core isolation cooling] after Level 8 trip")

ENCLOSURE E-2

HFA_0073L8RESTART ("Operator fails to restart HPCI [high pressure coolant injection] after Level 8 trip")

HFFA_ 1 SHV0760540_35 ("Local action - close 1-SHV-076-0540 (2-and 3-for Units 2 and 3) within 35 minutes")

HFFA0268480CRSTIE ("Failure to transfer deenergized 480V board to alternate supply (fire)")

HFFAOASD_RCIC ("Operator fails to start RCIC")

b. For each internal events and fire HFE considered in the EPU risk evaluation (i.e.,

HFEs in Tables 4-4 and 4-9 of Attachment 44 of the EPU LAR) that have a different dependency level between the CL TP and EPU cases, confirm that recovery is appropriately represented in the HFE's cognitive analysis and execution analysis such that the total HEP is realistically estimated for the CL TP and EPU cases. If it is not, then update the associated HEPs for the CL TP and EPU cases to appropriately address recovery, and provide the updated HEPs along with an explanation of how they were changed (provide sufficient detail and numerical values to understand the basis for the updated HEPs).

c. If changes were made to the HEPs in Part b of this RAI, then provide:

1. Updated risk results in Sections 5.1 ("Internal Event Results"), 5.2 ("Fire Risk Results"), and 5.6 ("Total Risk") in Attachment 44 of the EPU LAR, as applicable.

If Regulatory Guide (RG) 1.17 4 risk acceptance guidelines are exceeded, then please provide a detailed justification to support the conclusion that no "special circumst.ances" are created by the proposed EPU, include a discussion of which metrics are exceeded and the conservatisms in the analysis and the risk significance of these conservatisms.

2. Updated risk results in Table 8-1 ("Summary Risk Results for the Combined Sensitivity Study") of the response to APLA-RAI 08, dated April 27, 2016. If RG 1.17 4 risk acceptance guidelines are exceeded, then provide a detailed justification to support the conclusion that no "special circumstances" are created by the proposed EPU, include a discussion of which metrics are exceeded and the conservatisms in the analysis and the risk significance of these conservatisms.

TVA Response:

Response to RAI 09.01 a.

The requested detailed Human Reliability Analysis (HRA) calculation sheets are provided in of this RAI response. While the text of the HRA Calculator sheets was updated for clarity and to add more explanations, the selections made for the cognitive and execution evaluations and the associated HEPs were not altered. The human failure events (HFEs) included in this document are listed below.

HFA_0003PMP_START-CLTP HFA_0003PMP_START-EPU HFA_0071L8RESTART-CLTP HFA_0071L8RESTART-EPU

ENCLOSURE E-3 HFA_0073L8RESTART-CLTP HFA_0073L8RESTART-EPU HFFA_1SHV0760540_35-CLTP HFFA_1SHV0760540_35-EPU HFFA0268480CRSTIE-CLTP HFFA0268480CRSTIE-EPU HFFA0ASD_RCIC-CLTP HFFA0ASD_RCIC-EPU Response to RAI 09.01 b.

For the HFEs on the list in item a., there is no difference in Execution Recovery step assignment between the Current Licensed Thermal Power (CLTP) and Extended Power Uprate (EPU) cases; the Recovery is applied to particular steps of the procedure and does not change for the two cases. The assignment of recovery steps was reviewed and considered appropriate for the errors modeled.

The difference between the CLTP and EPU human error probabilities (HEPs) for these HFEs is therefore due to the timing differences from the Modular Accident Analysis Program (MAAP) runs for the two cases and the impact that has on the Cognitive portion of the HEP, if credited, and the dependency level for the recovery actions.

Application of recovery to the Execution steps is part of the analytical process of HRA. It is based upon the analysts identification and evaluation of the steps that will be reviewed by another operator. A combination of operational and HRA experience and an understanding of the specific procedures is used to make these determinations.

When there are multiple Execution steps in an HFE, it is common for the analyst to apply recovery to combinations of steps rather than to each individual one, based on the analysts knowledge of how the actions are taken and reviewed. In the case of the HFFA0ASD_RCIC action, it is considered likely that the operator will go through the steps to do the switch transfers and the check will happen afterwards.

An additional analytical variable is the selection of the type of Execution failure and associated HEP in HRA Calculator from Technique for Human Error Rate Prediction (THERP). In the case of HFFA0ASD_RCIC, the selection available did not match well the type of error being modeled and the variation in HEPs for these selections is significant.

The primary error for HFFA0ASD_RCIC is that the operator fails to make the proper switch transfer. The stress level is high and the operator is assumed to be using portable lighting.

The error associated with THERP Table 20-12, Item 8c (Reference 1) is for both high stress and violation of a stereotype (this was incorrectly called prototype in the HFE). In this case, there is no stereotype violation, but the choices available to the analyst are limited.

Selection of Item 8c would have resulted in an exceptionally high HEP of 1.3E-01.

Therefore, the analyst elected to use Item 8a with a more appropriate HEP of 2.7E-04, even though it is for normal stress, and then not credit a recovery step that would have lowered the HEP even further.

ENCLOSURE E-4 In summary, the analyst evaluated the Execution actions taken in HFFA0ASD_RCIC and determined that recovery would not have been applied for each individual switch action based on experience with such actions, and had to apply the most reasonable HEP from the THERP options provided in HRA Calculator, given that the error mode choices available did not completely match the situation.

In the set of HFEs whose different timing data led to a changed dependency level between the CLTP and EPU cases, HFE HFFA0ASD_RCIC was the only action with a combination of

1) recovery not applied to every step and 2) an execution failure mode applied that did not match well with the type of error.

For the HFEs cited for comparison in the RAI, HFA_0002RPV_LVL and HFA_0HCIINIT30, there is only one Execution step, so recovery was applied to that single action step. The analysis in these cases is much more straightforward.

The HFEs of Tables 4-4 and Table 4-9 of LAR Attachment 44 were reviewed. Those that have a difference between HEPs of the CLTP and EPU cases were categorized in three groups, as follows:

1. HFEs whose timing data differences between the CLTP and EPU cases did not lead to a change in the dependency level (as determined in the Time Window tab of the HRA Calculator).

For such HFEs, keeping the same HEP between the CLTP and EPU cases would be warranted because the dependency level stayed the same. However, a lower HEP was calculated in the CLTP case by either applying an additional cognitive recovery, or by lowering the cognitive dependency level of an already credited recovery. This was done to inject a reasonable conservatism in the delta risk evaluations, intended to take into account the added potential for recovery afforded by the additional time in the CLTP case.

2. HFEs whose timing data differences between the CLTP and EPU cases led to a change in the dependency level, which was carried over as-is in the cognitive and execution recoveries.

For these HFEs, the recoveries credited in the EPU case for the cognitive analysis and the execution analysis were carried over in the CLTP case, but with the lower dependence level granted by the additional time from the CLTP case. Recoveries were appropriately represented.

3. HFEs whose timing data differences between the CLTP and EPU cases led to a change in the dependency level, which was selectively carried over in the cognitive and execution recoveries.

A specific discussion is required for each of these HFEs, as follows.

In the EPU case, it was noted that for HFE HFA_0071L8RESTART, representing the failure of the operators to restart Reactor Core Isolation Cooling (RCIC) after its trip on Reactor Level L-8, the recovery step of the skip a step in procedure entry was assigned a complete level of dependency, while the recovery of another cognitive failure mechanism (failure of attention) was assigned a medium level of dependency. In contrast, the same HFE in the

ENCLOSURE E-5 CLTP case had these recovery steps both assigned to the low dependency level. This cognitive recovery dependency level for the EPU case is considered to be appropriate, because, in the relatively short time window of the human action, the operators may collectively make the mistake of skipping a step in the procedure. This can be explained as follows: the recovery step is stated as when RPV level drops below minus 120. RPV level could be recovered quickly with HPCI, but it may be difficult to turn level around with RCIC if level was not maintained and a step was skipped at a higher power level, leading to less available time to recover. At a lower power level, however, it would not be as significant since the operators would have more time to recover. The resulting HEP in the EPU case, 7.97E-03, is approximately 11 percent higher than the HEP obtained if the dependency level had been assigned to the medium level (7.13E-03), a moderate increase. Similarly, the EPU plant risk Core Damage Frequency (CDF), Large Early Release Frequency (LERF), and delta risk values (delta CDF, delta LERF), are slightly more conservative with this complete dependency level than if the medium level had been used. In particular, when, in the cutset files, an HEP of 7.13E-03 for HFA_0071L8RESTART is used in lieu of 7.93E-03, the CDF and LERF decreases slightly, by less than 1 percent. In summary, the EPU plant risk metrics presented in Attachment 44 of the EPU LAR are marginally more conservative with the 7.97E-03 value than with the lower value of 7.13E-03.

In the CLTP case, it was noted that for HFE HFA_0003PMP_START, representing the failure of the operators to restart a feedwater pump after its trip on Reactor Level L-8, the execution recoveries were left at the low dependency level used for the EPU case. A low dependency level is considered to be appropriate because feedwater is the normally operated system to provide both inventory and decay heat removal. As such, it could be expected that an error of execution to restart a feedwater pump would be promptly noticed and likely to be recovered. This is because if feedwater is available, RPV level could be recovered even faster than with HPCI or RCIC due to its larger makeup capability. With lower flow rate systems, it could be challenging to turn the level around within the recovery window if a step is missed, while a higher flow system (such as feedwater) can recover the RPV level more quickly. In that light, a low level of dependency is justified. If, however, a medium dependency level was used for the recovery of execution steps in the EPU case, the HEP of HFA_003PMP_START would increase from 1.14E-02 to 1.18E-02, a 3.5 percent increase. When this updated value is used in the cutset files of the EPU case, the CDF and LERF values are unchanged, thereby showing that increasing the execution recovery dependency level from low to medium has an insignificant effect on the risk results.

It was noted that for the failure to transfer a de-energized 480V board to alternate supply, represented with HFE HFA_0268480CRSTIE in the Internal Events PRA and HFFA_0268480CRSTIE in the Fire PRA, the execution recoveries were kept at a high level of dependencies in both the EPU and CLTP cases. This was done because the relatively complex actions modeled under these HFEs are performed locally and within a short time window. A lower recovery dependency level in the CLTP case would lead to an unrealistic value of the associated HEP.

Finally, it was noted that for HFEs HFA_0HCIINIT30 (operator failing to initiate high pressure injection within 30 min) and HFE HFA_0LPINIT10 (operator failing to initiate low pressure injection within 10 min), the execution recoveries were kept at a low level of dependencies in both the EPU and CLTP cases. This was done because these actions are intended to recover adequate reactor water level, a highly visible parameter that is constantly monitored by the operators. A low level of recovery dependency appropriately captures the likelihood that the operators would recover a failed injection. In contrast, selecting a higher level of

ENCLOSURE E-6 recovery dependency for these actions would provide an unrealistically high HEP in both the CLTP and EPU cases.

It is concluded that the recovery analysis is appropriately represented in the HFEs cognitive analysis and execution analysis such that the total HEP is realistically estimated for both the CLTP and EPU cases.

Response to RAI 09.01 c.

No action is required for this request as no changes were made to HEPs in part b. of this RAI response.

Reference

1.

NUREG/CR-1278, Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications, dated August 1983.

Human Reliability Analysis Calculation Reports

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 1 of 142 HFA_0003PMP_START-CLTP, OPERATOR FAILS TO RESTART RFW AFTER LEVEL 8 TRIP Plant Data File File Size File Date Record Date BFN_APLA-RAI 09.01 Response_20160708.hra 1331200 7/8/2016 7/8/2016 Name Date Analyst J. Branch (Update for CLTP) 7/6/2016 Reviewer E. Collins, JENSEN HUGHES 7/8/2016 HEP Summary P1 P2 Pcog Pexe Total HEP Error Factor Method CBDTM HCR/ORE Maximum THERP HEP 1.07E-03 2.44E-10 1.07E-03 8.82E-03 9.88E-03 5

RAW FV Risk Significant 1.03E+00 2.66E-06 NO Identification and Definition In the event of a transient, operators are instructed to maintain water level between +2 in. and

+51 in. (level 8) (EOI-1, RC/L-4). If water level reaches +55 in. and RFW is operating to provide injection, the pumps would be tripped and main turbine automatically trip (HPCI and RCIC trip at

+51 in.). Without any other source of injection, the water level would then decrease. Operators are instructed to restart injection before water level decreases to +2 in. Failing this, the water level would continue to decrease eventually leading to core uncovery and heatup.

Several other steps in the EOI also instruct operators to restore Feedwater, CRD, HPCI, RCIC, or other pump to restore level (e.g., EOI-1, step RC/L-4; C-1, step C1-4).

The operators monitor RPV level very closely during a transient. They would be anticipating tripping of the RFW, HPCI or RCIC turbine at Level 8 and restarting the pumps again if the water level decreased.

This HFE has been updated for the current licensing thermal power (CLTP) case. 4 to the LAR on Probabilistic Risk Assessment, pg. Att 44-72, discusses the following MAAP runs that were reviewed to identify the appropriate estimate for the time to core damage for Tsw.

(a) Browns Ferry Nuclear Plant Unit 1 PSA MAAP Thermal Hydraulic Calculations, Rev. 2, July 2008.

(b) NDN-000-999-2008-0006, BFN Probabilistic Risk Assessment - Thermal Hydraulics Analysis, Rev. 4, August 2015.

MAAP 4.0.7 run (407_CLTP)

Case 1 is a general transient scram with no injection.

Time to core damage (Tsw):

407_CLTP: Tsw = 0.70 hrs or 42 min The CLTP modeling uses 42 minutes for HFEs in all general transient (GTRAN) sequences for actions to initiate injection if all injection is lost at the scram and there are no stuck open relief valves.

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 2 of 142 Cues and Indications Initial Cue Low level in reactor pressure vessel (+2 in normal range)

Recovery Cue Cue Comments EOI-1 Step RC/L-4 instructs operators to maintain water level between +2 in. and

+51 in. Although operators would likely attempt to restart injection prior to water level dropping to +2 in., it is assumed that the operators allow the level to decrease to the procedural minimum before taking action. Thus the cue is taken as water level at +2 in.

Degree of Clarity Clarity of Cues and Indications are modeled explicitly in CBDTM Procedures Cognitive Procedure 1-EOI-1 (RPV Control, Unit 1, Browns Ferry Nuclear Power Plant) Revision:

Cognitive Step Number RC/L-4 Cognitive Instruction Restore and maintain RPV level... with one of the following sources Execution Procedure 1-EOI App 5D (Injection System Lineup HPCI) Revision:

Execution Instruction Job Performance Measure JPM: Not Selected Other Procedures 1-C-1 (Alternate Level Control, Unit 1, Browns Ferry Nuclear Power Plant) Revision: 0 Notes Response begins with EOI-1, and transitions to C-1 when level cannot be maintained. Although operators would likely attempt to restart injection prior to water level dropping to +2 in., it is assumed that the operators allow the level to decrease to the procedural minimum before taking action.

Training Classroom Training Simulator Training Crew Member Included Total Available Required for Execution Notes Reactor Operator No 0

2 Assume that HPCI or RCIC restart would take precendence, but RFW would be restarted.

Notes Assumptions The first cue the operators receive to restart RFW, HPCI or RCIC is when the water level reaches

+2 in. In reality, The operators monitor RPV level very closely during a transient. They would be anticipating tripping of the RFW, HPCI or RCIC turbine at Level 8 and restarting the pumps again if the water level decreased, thus pump restart would likely be initiated prior to water level dropping to +2 in.

Operator Interview Insights During observation of a demonstration of a representative scenario using the Unit 3 simulator, it was clear that the operators would track RPV level carefully.

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 3 of 142 Timing Analysis Time available for cognition and recovery 39 Minutes Time available for recovery 38.5 Minutes SPAR-H Available time (cognitive) 39 Minutes SPAR-H Available time (execution) ratio 39.50 Minutes EPRI Minimum level of dependence for recovery LD Notes This HFE has been updated for the current licensing thermal power (CLTP) case. 4 to the LAR on Probabilistic Risk Assessment, pg. Att 44-72, discusses the following MAAP runs that were reviewed to identify the appropriate estimate for the time to core damage for Tsw.

(a) Browns Ferry Nuclear Plant Unit 1 PSA MAAP Thermal Hydraulic Calculations, Rev. 2, July 2008.

(b) NDN-000-999-2008-0006, BFN Probabilistic Risk Assessment - Thermal Hydraulics Analysis, Rev. 4, August 2015.

MAAP 4.0.7 run (407_CLTP)

Case 1 is a general transient scram with no injection.

Time to core damage (Tsw):

407_CLTP: Tsw = 0.70 hrs or 42 min The CLTP modeling uses 42 minutes for HFEs in all general transient (GTRAN) sequences for actions to initiate injection if all injection is lost at the scram and there are no stuck open relief valves.

The initiation time is taken as the time when Level 8 is reached since this HEP applies to a failure to restart RFW following a Level 8 trip. Thus, the time delay is time from L8 to +2 in. = 0.06 - 0.02

= 0.04 hours4.62963e-5 days 0.00111 hours 6.613757e-6 weeks 1.522e-6 months , or 2.4 minutes. The time delay would be shorter for scenarios that do not involve L8. For the CLTP case, however, the MAAP timing discussed above for Tsw was also used to re-evaluate Tdelay, due to impacts from changes in RPV power since these operator actions may be derived based on other parameters such as time for battery depletion, time for irreversible equipment failure, or depletion of inventory. This re-evaluation resulted in a value of 2 minutes.

The median response time would be very short, because operators are monitoring water level closely during any transient. Based on response for similar actions median response time is taken as 30 seconds to restore RFW (per 1-AOI-3-1 4.2.12, 1-AOI-100-1 and 1-OI-3) once low level is detected.

Cognitive Analysis Pc Failure Mechanism Branch HEP

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 4 of 142 Pca: Availability of Information a

0.00E+00 Pcb: Failure of Attention m

1.50E-02 Pcc: Misread/miscommunicate data a

0.00E+00 Pcd: Information misleading a

0.00E+00 Pce: Skip a step in procedure e

1.98E-03 Pcf: Misinterpret Instructions a

0.00E+00 Pcg: Misinterpret decision logic k

0.00E+00 Pch: Deliberate violation a

0.00E+00 Initial Pc(without recovery credited) 1.70E-02 Notes Pca: Availability of Information Notes/Assumptions:

Pcb: Failure of Attention Notes/Assumptions: Workload would be expected to be relatively high due to attempts to restore some source of HPI. Action primarily triggered based on monitoring RPV level, which is indicated by a variety of displays and meters.

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 5 of 142 Pcc: Misread/miscommunicate data Notes/Assumptions:

Pcd: Information misleading Notes/Assumptions:

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 6 of 142 Pce: Skip a step in procedure Notes/Assumptions: Since a transition among three procedures might be required, and the unit supervisor would be following multiple paths through the EOIs, this is treated as a case of multiple procedures. Placekeeping aids are used in the form of a bracket in which the unit supervisor initials the steps as they are reached.

Pcf: Misinterpret Instructions Notes/Assumptions:

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 7 of 142 Pcg: Misinterpret decision logic Notes/Assumptions: Logic is straightforward, and this is a well-practiced scenario.

Pch: Deliberate violation Notes/Assumptions: There is no reason to expect other than verbatim compliance with this instruction.

Cognitive Recovery Initial HEP Self Review Extra Crew STA Review Shift Change ERF Review Recovery Matrix Dependency Level Multiply HEP by Override Value Final Value Pca n/a N/A 1.00E+00 0.0 Pcb 1.50E-02 X

LD 6.43E-02 9.65E-04 Pcc n/a N/A 1.00E+00 0.0 Pcd n/a N/A 1.00E+00 0.0

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 8 of 142 Pce 1.98E-03 X

LD 5.19E-02 1.03E-04 Pcf n/a N/A 1.00E+00 0.0 Pcg n/a N/A 1.00E+00 0.0 Pch n/a N/A 1.00E+00 0.0 Final Pc (with recovery credited) 1.07E-03 Notes Self-review is credited because RPV level is continuously monitored Execution Performance Shaping Factors Environment Lighting Normal Heat/Humidity Normal Radiation Non Radiation Area Atmosphere Normal Complexity of Response Execution Simple Equipment Accessibility (Cognitive)

Main control room Accessible Equipment Accessibility (Execution)

Main control room Accessible Stress High Plant Response As Expected:

Yes Workload:

High Performance Shaping Factors:

Negative Notes The workload is high and the PSFs are negative due to the fire Sigma Table Plant Type Response Type LB Sigma UB BWR CP1 0.4 0.7 1

CP2 0.2 0.58 0.96 CP3 0.59 0.75 0.91 PWR CP1 0.26 0.57 0.88 CP2 0.07 0.38 0.69 CP3 0.77 Sigma:

7.00E-01 HEP:

2.44E-10 Notes/Assumptions The cue is to start HPCI when level reaches +2 in. Although the operators are prepared for this condition, it is still a response of type CP1.

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 9 of 142 Execution Unrecovered Procedure Comment THERP Step No.

Instruction /

Comment Error Type Table Item HEP Stress Factor Override Restore and maintain water lvl between +2 in and +51 in

... with...

RFW, Appx 5A EOM 20-7b 1

4.3E-4 Comment Operator must recognize low water level and the need to restore feedwater High EOI-1 RC/L-04 Location:

Total Step HEP 2.15E-03 DEPRESS 1-HS 8A(9A)(10A),

RFPT 1A(1B)(1C)

SPEED CONT RAISE/LOWE R, and VERIFY amber light is illuminated.

EOM 20-7b 1

4.3e-04 High 1-EOI Appx 5A Step 4 Location:

Total Step HEP 2.15E-03 DEPRESS 1-HS 124A(150A)(1 75A), RFPT 1A(1B)(1C)

TRIP RESET, and VERIFY that the turbine trip is RESET EOM 20-7b 1

4.3e-04 High 1-EOI Appx 5A Step 8 Location:

Total Step HEP 2.15E-03 1-EOI Appx 5A Step 10 PLACE 1-HS-46-112A(138A)(1 63A), RFPT EOM 20-7b 1

4.3e-04 High

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 10 of 142 1A(1B)(1C)

START/LOCA L ENABLE, in START, AND VERIFY RFPT speed increases to approximately 600 rpm.

Location:

Total Step HEP 2.15E-03 SLOWLY ADJUST RFPT speed UNTIL fw flow to the RPV is indicated, using ANY of the following methods on Panel 1-9-5: 1-HS 8A(9A)(10A) switch in MNL GVRNOR; 1-SIC 8(9)(10)

SPEED CNTRL in MNL;1-LIC 5, RX LVL CNTRL in MNL with 1-SIC 8(9)(10)

AUTO EOM 20-7b 1

4.3e-04 High 1-EOI App 5A Step 12 Location:

Total Step HEP 2.15E-03 EOI-1 RC/L-7,8 If RPV water lvl drops below

-120 in....

augment RPV EOM 20-7b 1

4.3e-04 High

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 11 of 142 water lvl...

Location:

Total Step HEP 2.15E-03 ADJUST RFPT speed as necessary to control injection using the methods of step 12.

EOM 20-7b 1

4.3e-04 High 1-EOI App 5A Step 13 Location:

Total Step HEP 2.15E-03 WHEN....

RPV level is approximately equal to desired level AND automatic level control is desired, THEN

.... PLACE 1-LIC-46-5, REACTOR WATER LEVEL CONTROL, in AUTO with individual 1-SIC 8(9)(10),

SPEED CONTROL in AUTO EOM 20-7b 1

4.3e-04 High 1-EO1 Appx 5A Step 14 Location:

Main control room Total Step HEP 2.15E-03 Execution Recovered Critical Step No.

Recovery Step No.

Action HEP (Crit)

HEP (Rec)

Dep.

Cond. HEP (Rec)

Total for Step EOI-1 RC/L-04 Restore and maintain water lvl between +2 in 2.15E-03 1.12E-04

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 12 of 142 and +51 in... with

... RFW, Appx 5A EOI-1 RC/L-7,8 If RPV water lvl drops below -120 in.... augment RPV water lvl...

2.15E-03 LD 5.20E-02 1-EOI Appx 5A Step 4 DEPRESS 1-HS-46-8A(9A)(10A),

RFPT 1A(1B)(1C)

SPEED CONT RAISE/LOWER, and VERIFY amber light is illuminated.

2.15E-03 2.15E-03 1-EOI Appx 5A Step 8 DEPRESS 1-HS-3-

124A(150A)(175A

), RFPT 1A(1B)(1C) TRIP RESET, and VERIFY that the turbine trip is RESET 2.15E-03 2.15E-03 1-EOI Appx 5A Step 10 PLACE 1-HS 112A(138A)(163A

), RFPT 1A(1B)(1C)

START/LOCAL ENABLE, in START, AND VERIFY RFPT speed increases to approximately 600 rpm.

2.15E-03 2.15E-03 1-EOI App 5A Step 12 SLOWLY ADJUST RFPT speed UNTIL fw flow to the RPV is indicated, using ANY of the 2.15E-03 1.12E-04

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 13 of 142 following methods on Panel 1-9-5: 1-HS 8A(9A)(10A) switch in MNL GVRNOR; 1-SIC-46-8(9)(10)

SPEED CNTRL in MNL;1-LIC 5, RX LVL CNTRL in MNL with 1-SIC 8(9)(10) AUTO 1-EOI App 5A Step 13 ADJUST RFPT speed as necessary to control injection using the methods of step 12.

2.15E-03 LD 5.20E-02 1-EO1 Appx 5A Step 14 WHEN.... RPV level is approximately equal to desired level AND automatic level control is desired, THEN.... PLACE 1-LIC-46-5, REACTOR WATER LEVEL CONTROL, in AUTO with individual 1-SIC-46-8(9)(10),

SPEED CONTROL in AUTO 2.15E-03 2.15E-03 Total Unrecovered:

1.29E-02 Total Recovered:

8.82E-03

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 14 of 142 HFA_0003PMP_START-EPU, OPERATOR FAILS TO RESTART RFW AFTER LEVEL 8 TRIP Plant Data File File Size File Date Record Date BFN_APLA-RAI 09.01 Response_2016 0708.hra 1331200 7/8/2016 7/8/2016 Name Date Analyst J. Branch (Update for EPU) 7/6/2016 Reviewer E. Collins, JENSEN HUGHES 7/8/2016 HEP Summary P1 P2 Pcog Pexe Total HEP Error Factor Method CBDTM HCR/ORE Maximum THERP HEP 2.63E-03 2.21E-09 2.63E-03 8.82E-03 1.14E-02 5

RAW FV Risk Significant 1.03E+00 2.66E-06 NO Identification and Definition In the event of a transient, operators are instructed to maintain water level between +2 in. and

+51 in. (level 8) (EOI-1, RC/L-4). If water level reaches +55 in. and RFW is operating to provide injection, the pumps would be tripped and main turbine automatically trip (HPCI and RCIC trip at

+51 in.). Without any other source of injection, the water level would then decrease. Operators are instructed to restart injection before water level decreases to +2 in. Failing this, the water level would continue to decrease eventually leading to core uncovery and heatup.

Several other steps in the EOI also instruct operators to restore Feedwater, CRD, HPCI, RCIC, or other pump to restore level (e.g., EOI-1, step RC/L-4; C-1, step C1-4).

The operators monitor RPV level very closely during a transient. They would be anticipating tripping of the RFW, HPCI or RCIC turbine at Level 8 and restarting the pumps again if the water level decreased.

This HFE has been updated for the EPU case. Attachment 44 to the LAR on Probabilistic Risk Assessment, pg. Att 44-72, discusses the following MAAP runs that were reviewed to identify the appropriate estimate for the time to core damage for Tsw.

(a) Browns Ferry Nuclear Plant Unit 1 PSA MAAP Thermal Hydraulic Calculations, Rev. 2, July 2008.

(b) NDN-000-999-2008-0006, BFN Probabilistic Risk Assessment - Thermal Hydraulics Analysis, Rev. 4, August 2015.

MAAP 4.0.7 run of equivalent cases (407_EPU)

Case 1 is a general transient scram with no injection.

Time to core damage (Tsw):

407_EPU: Tsw = 0.59 hrs or 35.4 min A Tsw of 35 minutes typically used in the EPU Fire HRA for actions to initiate injection if all

ALPA-RAI-09.01 (1) Response - HRA Calculator Reports Page 15 of 142 injection is lost at the scram and there are no stuck open relief valves.

Cues and Indications Initial Cue Low level in reactor pressure vessel (+2 in normal range)