ML20205L993
| ML20205L993 | |
| Person / Time | |
|---|---|
| Site: | Callaway  |
| Issue date: | 10/21/1988 |
| From: | Schnell D UNION ELECTRIC CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| ULNRC-1849, NUDOCS 8811020267 | |
| Download: ML20205L993 (13) | |
Text
[,
c 1901 Gratiot Street.
Pcat Ofice Box !49 St. Iows. Miuouri63166 314 554 2650 l
- Union o 'sm "
c' *cr '**
Etucraic g<',vcc ?'c""c"'
E U.
S. Nuclear Regulatory Commission ATTN:
Document Control Desk Mail Station'Pl-137 Washington,'D. C. 20555
[-
Gentlemen ULNRC-1849 DOCKST NUMBER 50-483 CALLAWAW PLANT STEAM GENERATOR TUBE RUPTURE - OPERATOR ACTION TIMES Reference NRC letter from T. W. Alexion to D.
F.
Schnell, dated 8/18/88 The reference letter transmitted a request for additional information concerning operator action times used in the Callaway Steam Generator Tube Rupture Analysis.
The attachment to this letter provides the basis for the assumed action times.
Very truly yours,
(
Donald F.
Schnell DS/do Attachment 80
\\
0811020267 8G102.
l PDR ADOCK 05000403 P-PDC
l STATE OF MISSOURI -).)
Alan C.
Passwater, of lawful age, being first duly sworn upon oath says that he.is Manager, Licensing and Fuels (Nuclear) for-Union Electric Company; that he has read the foregoing document and knows the. content thereof; that he has executed the same for and on behalf of said company with full power and authority to do so; and that the facts therein stated are true and correct to the best of his knowledge, information and belief.
By _ _
Alan C. Passwater Manager, Licensing and Fuels Nuclear SUBSCRIBED and sworn to before me this c2/
day of W,198 0 L
BARbARMJ. EFA((
NOTAAY FUBUC, STATE 05 MIS $00Ri MY COMM:SSION EXPIRES april 22, 1939 i
ST. LOUIS COUNTY.
i t
i t
I I
.....-m----
r cc:
Gerald Charnoff, Esq.
Shaw, Pittman, Potts & Trowbridge 2300 N.
Street, N.W.
Washington, D.C.
20037 Dr. J. O. Cermak CFA, Inc.
4 Professional Drive (Suite 110)
Gaithersburg, MD 20879 R. C. Knop Chief, Reactor Project Branch 1 U.S.
Nuclear Regulatory Commission Region III 799 Roosevelt Road Glen Ellyn, Illinois 60137 Bruce Little Callaway Resident Office U.S.
Nuclear Regulatory Commission RRil Steedman, Missouri 65077 7bm Alexion (2)
Office of Nuclear Reactor Regulation U.S.
Nuclear Regulatory Commission Mail Stop 316 7920 Norfolk Avenue Bethesda, MD 20014 Manager, Electric Department Missouri Public Service Commission P.O.
Box 360 Jefferson City, MO 65102 U.S. Nuclear Regulatory Commission ATTM Document Control Desk Washington, D.C.
20555
=.
t bcc:
D. Shafer/A160.761
- /QA Record (CA-75 8)
Nuclear Date E210.01 DFS/ Chrono D.
F. Schnell J.
E.
Birk A.
P. Neuhalfen M. A.
Stiller G.
L. Randolph R. J.
Irwin H. Wuertenbaecher W. R. Campbell A. C.
Passwater R.
P. Wendling D.
E. Shafer D.
J. Walker O. Maynard (WCNOC)
N. P. Goel (Bechtel)
T.
P. Sharkey NS RB (Sandra Auston) 6.A.H h T.A. y 6
i l
1 l
i l
l
RESPONSE TO THE NRC'S REQUEST FOR ADDITIONAL INFORMATION CONCERNIJJG_S_G_TR OPERATOR ACTION TIME _S QUESTION After consideration of the staff's teleconference with the Licensee on April 8, 1988, the staff has determined that the operator response times provided by the Licensee in its February 3,
1987 response to the staff's request for information of November 12, 1986 are inappropriate for purposes of the staff's review.
Specifically, the staff understands that the Licensee has estimated these times rather than demonstrated them.
In its letter to the Westinghouse Owner's Group dated March 30, 1987, the staff required "... demonstration runs... to show that the accident can be mitigated within a period of time compatible with overfill prevention, using design basis assumptions regarding available equipment, and to demonstrate that the operator action times assumed in the analysis are realistic."
The Licensee has provided an analysis that uses "best-guess averages" for operator response times for critical tasks, such as identifying the Steam Generator Tube Rupture through recognition of increasing, uncontrolled level or high radiation, including manual sampling of steam generator liquid.
The Licensee should provide further information that d.emonstrates that the operator response times assumed in the Callaway analyses are realistic, e.g.
are representative of the operator population at the plant, and that the maximum response times fall within the bounds of analysis.
This information should address both control roon actions and actions performed outside the control room such as manual steam generator liquid sampling.
R E S P_O M E 1.
Clarification of SNUPPS involvement with WOG in SGTR Analyses The SNUPPS utilities (Union Electric and Wolf Creek Nuclear Operating Company) did not participate in the Westinghouse owner's Group (WOG) SGTR Analysis Group Subcommittee.
- Instead, SNUPPS chose to analyze the SGTR scenarios separately from the WOG in order to pay proper attention to neveral unique aspects of the SNUPPS design.
Nevertheless, Callaway and Wolf Creek, have participated in other WOG activities.
One such activity, was the verification and validation of the generic Westinghouse Emergency Operating Guidelines (ERGS).
These activities were performed at Callaway and Seabrook in 1982 and 1984 respectively.
The operator action t
times observed during the ERG tube rupture scenarios were taken into consideration when developing the SNUPPS SGTR operator p
i action times as discussed below.
Page 1
II'.
Basis for Action Times _Used in_the Call _away SGTR_ Analyses.
Union Electric and Wolf Creek have jointly submitted justifications for the original cperator action times in Section 2 of SLNRC 86-1 (1-8-86) and Responses 3 and 4 of SLNRC 86-0S (4-1-86).
In theJe submittals, five key operator action were identified:
1.
Faulted SG Isolation, 2.
RCS Cooldown Ireitiation, 3.
RCS Depressurization, 4.
Termination of Safety Injection (SI), and 5.
Trar4sition to cold Shutdown.
The first four operator actions are specifically modelled in the thermal hydraulic analysis and directly affect accident consequences.
As described in the aforementioned SLNRC transmittals, these critical actions times were derived from the following:
1.
Plant Simulator Exercises, 2.
SGTR Events at Domestic Nuclear Facilities 3.
Draft ANS Standard 58.8, Rev.
2, and 4.
Actual Callaway Plant Experiences The assumed action times are presented in Table 1.
The assumed Overfill actions are most representative of the simulator exercises since the assumption of manual closure of the ARV block valvo delays timely isolation of the faulted SG.
The following sections discuss how these operator action times were derived.
Table 1.
Operator Action Times Used in Callaway SGTR Analyses Stuck-ARV Overfill Op_e ra t o r Ac t.i_on SLNRC 86-01 ULNRC-1518 1.
Faulted SG Isolation 28 min 16 min 2.
RCS Cooldown Initiation 40 min 24 min 3.
RCS Depressurization 55 min
>35 min 4.
Termination of SI 58 min
>38 min A.
P_lant Simu_lator Exper_iences Simulator data was heavily weighted in dotermining operator responne because it reflects how SNUPPS plant operators have performed and are expected to perform.
Page 2
a Si'mulator based operator action times are available from four sources:
1.
Operator Training on the callaway Simulator y
2.
Verification and Validations of the Westinghouse generic ERGS, Rev. O at the Callaway Simulator.in 1982, 3.
Composite Verification and Validation data including Rev. 1 of the generic ERGS at the Seabrook simulator in 1984, and 4.
Training of Operators from several utilities at tF.e Westinghouse Training Center in Zion, Illinois on the SNUPPS and Zion simulators.
A1.,_O_perator Exercises on the_Callaway_ Simulator Nominal Callaway simulator SGTR exercise data is summarined in Table 2.
These events depict typical SGTR scenarios and involve a varying degree of break flow.
The assumed break flow at SGTR initiation for the analyzed worst-case sconarios was 550 gpm.
Approximately 50 callaway operators have participated in these recorded exercises and all operators receive periodic training on the event.
In addition to these tabulated events other exercises have been held at callaway.
The results at these exercises were not included in this table since the times obtained were very short, reflecting the large amount of emphasis on SGTR events in training prior to the simulator exercises.
These results however were reported in SLNRC 86-01, Section 2.0 as results of the Callaway Validation & Verification, Rev. 1 exercises.
Table 2 - Callaway Simulator Nominal CGTR Exercises III III 1983 1985
_12]_13 12/19 12/29 8/08 g/08 9415
- _8115, 8/2_2 RL2_2 Break Flow 400 400 400 800 800 500 500 500 500 (gpm)
Icolate 12 11.6 16.2 11.1 10.9 9.25 11.3 16.5 8.3 Fculted S/G Initiate RCS Note 2 Note 2 27.7 13.8 15.3 20.25 16.3 20.7 23 Cooldown Complete RCS Note 2 Note 2 40.9 31.5 42.0 34.0 39.0 36.0 36 D: pressurization Torminate SI 33 33 40.9 32.5 37.5 42.0 28.7 37.2 34.8 NT - Not Taken Notes:
1.
All values in minuten unless indicated.
2.
Value not takon.
Page 3
Two SGTR scenarios' assuming a failed open AFW Control Valve (the single _ failure for the SNUPPS Overfill scenario) were recorded in 1986 and are presented in Table 3.
The quickened operator action
' times in relation to the earlier exercises are indicative of increased operator familiarity with the E-3 procedure and the unique aspects of this scenario.
Trble 3 - Callaway Simulator SGTR Events with Auxiliary Feedwater Controller Failure 11/10/86 11210/86 Break Flow 500 gpm 500 gpm Icolate Faulted S/G 11.1 min 11 min Initiate RCS Cooldown 12,7 min 14 min Complete RCS Depressurization 23.3 min 20 min Termination of SI 24.6 min 25 min A2.
Coordinated Simulator Exercises Data from coordinated simulator exercises were also used as sources of information regarding operator action times.
Data from three simulator sources were available:
1.
The verification and validation of the Westinghouse generic ERGS, Rev. O at the Callaway Simulator in 1982, 2.
The composite verification and validation data including Rev. 1 of the generic ERGS at the Seabrook simulator in 1984, and 3.
The results of 24 exercises run at the Westinghouse Training Center in Zion, Illinois using both the SNUPPS and Zion simulators with crews from various utilities.
The results of these exercises are presented in Table 4.
For comparison purposes the range of operator action times for the above Callaway exercises are also presented.
The consistency of these operator action times is additional evidence which supports the validity of the results of the Callaway exercises.
Page 4
n; j
.)
r Table 4 - Coordinated' Simulator Exorcises II)
II)
I1)
Callaway(1)
Seabrook Westinghouse Callaway 1983 & 1985 V&V, Rev. O V&V, Rev. 1 24 Exercises Range of_ Times
~1982 1984 Zion /SNUPPS Break Flow 400-800 100-800 Not Not
. (gpm)
Available Available Icolated 8-16 19 10 9 (Note 2)
Fculted S/G Initiate RCS 13-28 24 23 16 (Note 2)
Cooldown Complete RCS 20-42 30 33 32 (Note 2)
Dspressurization Terminate SI 25-42 40 41 37 (Note 2)
Notes:
- 1. All values in minutes unleos indicated.
2.' Operator action times were recorded from time of reactor trip.
For purposes of comparison 3 minutes was added to each operator action time to reflect the time from initiation of tube rupture to reactor trip.
One SGTR with Loss of Offsite Power (LOOP) was recorded in 1982 at the Callaway Simulator and documented in WCAP-10204.
Such an accident would be very similar to those events observed during the analyzed overfill scenario (See ULNRC-1518) and was therefore considered in developing operator action times.
Tab _le 5 - Callaway Simulator SGTR Event with LOOP Power Assumed Break Flow 500 gpm Isolated Faulted S/G 15.0 min Initiate RCS Cooldown 17.1 min Complete RCS Depressurization 38.4 min Terminate SI 43.0 min In reviewing Tables 2 thru 5, it can be noen that in all but a few exceptions, the Operator action times chosen for the two analysis scenarios are conservative given the actual scencrio and break flow rates assumed.
Specifically it can be seen that the operator action time were chenen which envelop credible SGTP scenarios under present EOP guidance.
Page 5
B.
Actual SGTR Events A table of Operator response times for actual domestic SGTR events is presented below:
Table 6. Domestic Tube rupture Events Time After Occurrence of SGTR (Min.)
Action Prairie Ginna North Island Anna Isolate Faulted SG 30 13
~ 18 Initiate RCS Cooldown 43 15
~ 21 Complete RCS Depress.
51 45
~ 36 Terminate Safety Injection 55 72
~ 36 The Prairie Island event was less than a full double-ended rupture and consequently the accident progressed more slowly than those analyzed by SNUPPS.
As a result, operator action times are longer than an SGTR with SNUPPS-assumed break flow rates.
This observation is supported by the fact that operator action times in the early stages of the Ginna and North Anna events were shorter than at Prairie Island.
The Ginna event was complicated by several factors which are not expected in an SGTR at SNUPPS facilities.
These factors include a Stuck-Open Pressurizer PORV that resulted in lower than intended RCS pressures, and the occurrence and persistence of a steam bubble in the reactor vessel head.
The steam bubble occurred, in part, because the Ginna Plant upper head is at the hot leg temperature.
At the SNUPPS facilities the upper head is maintained at the cold leg temperature.
l Additionally, the proceduren in effect at Ginna were not explicit about the criteria for terminating SI with a steam bubble present in the reactor vessel.
The operators, taking a conservative approach to maintaining core cooling, allowed SI to continue and therefore repressurized the RCS.
This in the dominant reason for the 45 minuto period to termination of break flow.
Given the experience at Ginna, improvements were made in procedures and operator training which have reduce operator action times.
These improvements were enhibited in the North Anna SGTR.
However, the operator actions were delayed at North Anna as a result of what was believed to be radiation monitor malfunctions.
Given the circumstances surrounding the North Anna event, the SNUPPS operator action times are supported by this latest SGTR.
C.
Dr.af t_N{S_ Standard _5L 8,_Rev. 2 Section 2 of SLNRC 86-01 and SLNRC 86-05 discusses the use of Draft ANS Standard 58.8, Rev. 2 in developing operator action times.
Page 6
Di Actual Callaway Plant Experience Actual Callaway Plant experience in isolating stuck ARVs was factored into the development of the SNUPPS operator action times as discussed in response to Question 3 in SLNRC 86-05.
III.
Operator Action Discussed _in the_ULNRC-1442 (2-3-87)
The ULNRC-1442 responses were provided at the staff's request for additional insight into actions that a SNUPPS operator would be expected take in order to isolate and identify the faulted SG.
Recognizing this request and the variety of avenues that an operator may take, "estimated" action times were given for the analyses scenarios.
The provided times did not represent actual m;asured times but rather illustrated what considerations were available to the operators given the scenario..
The discussion also was formulated to assist the NRC reviewers in understanding the systematic approach to the SNUPPS Emergency Procedures in identifying and mitigating an SGTR.
As previously discussed, the analysis is not sensitive to these subordinate steps as long as the operator reaches the milestones within the assumed time frame (See Table 1).
Therefore, demonstrating the subordinate times would not benefit the analysis.
In support of the steam generator isolation action times, two actions taken outside the control room were discussed in ULNRC-1442:
manual closure of the faulted SG ARV and manual sampling of the faulted SG.
Previous responses to NRC's requests for information have adequately documented Callaway's ability to manually close a stuck-open ARV within 20 minutes (See SLNRC 86-05, Response 3).
Therefore this critical action has been demonstrated and will not be discussed further.
i l
Manual Sampling is only o concern given a tube rupture coincident with loss of offsite power (LOOP).
LOOP soon after a SGTR initiation does complicate identification of an SGTR and the faulted SG.
Since the SG blowdown and condenser air discharge l
monitors are not powered from emergency power supplies, this l
information is not available in the control room.
- However, sufficient information is available tc identify a SGTR (See ULNRC-1442 for details).
Identification of the faulted steam generator lo also complicated by LOOP.
For the overfill scenario, uncontrolled narrow range rise would be recognized at approximately 6 to 10 minutes into the accident.
This leaves ample time for operators to complete isolation activities within 16 minutes.
For the stuck-open ARV scenario, steam generator level does not recover until 40 minutes after SGTR.
In this case, the affected SG would be identified by the following:
These 1.
Secondary Side Release Point Monitors (SSRM) monitors continuously check for gamma radiation in the steam release from the ARVs.
They are supplied with a reliable Page 7
~
power supply and are available after LOOP.
The readings of these monitors are available in the control room and after LOOP locally near the auxilliary shutdown panel.
These monitors are therefore ideal for detecting a situation similar to the SNUPPS stuck-open ARV scenario.
(For additional details about the SSRMs please refer to ULNRC-?.825, 9/2/88).
It should be noted that the SSRMs were not discussed in our response to questions answered in ULNRC-1442, 2.
Manual Sanpling - This method requires dispatching chemistry personnel to manually sample en SG for gross radioactivity.
Normal routine chemistry samples of a SG require 15 minutes to draw the sample, 5 minute to transport the sample and prepare it for counting, and 20-25 minutes to perform a detailed isotopic analysis of the sample.
However in the case of a SGTR, the sampling procedure would be modified and a hand held detector would be used to determine a gross activity indication after the sample was drawn in lieu of the more time consuming isotopic analysis.
Given an actual 15 minute sampling time based upon actual plant experience, this gross activity assessment with a hand held instrument would take approximately 20 minutes.
Note that emphasis above was placed on "procedural identifi-cation" Operators also have other available indications outside the emergency procedures guidance to identify the faulted SG including:
1.
Initial feedwater flow / steam flow mismatch, 2.
SG liquid conductively, 3.
SG pressure, and 4.
SG wide range level.
IV.
Critical Operator Action Times As previously discussed, select operator actions are used to influence the thermal hydraulic nnalyses of the SGTR scenarios.
As might be expected, some action have greater importance to the consequences of the accident than others.
Therefore, it is important that these "critical" actions are highlighted.
For the stuck-open ARV scenario, 99% of the thyroid dose is attributed to releases during time interval the ARV is stuck open (prior to faulted SG isolation).
The key operator action time is therefore the time to close the ARV, which the analysis assumes to be 20 minutes.
This value is well supported by the simulator exercises and the actual isolations which have been performed at Callaway (See SLNRC 06-05).
For the Overfill scenario, overfill occurs as a result of cumulative AFW flow and prolonged SI flow.
The key operator actions in this scenario are the operator's ability to identify a Page 8 i
e' jt
- tube rupture and move through the E-3 proceduro.
Simulator exercises demonstrated that these key actions can be accompliehed by Callaway operators in a tiraly manner.
These actions are jy further supported by the diverse instruments and methods exist by which to identify the SGTR as discussed above and in ULNRC-1442.
p V.
Conclusion The preceding discussions have outlined how-operator action times
.have been derived for the Callaway SGTR analyses.
These discussions provide evidence of critical action "demonstration" and that the values assumed represent maximum observed response times.
Given that the observations were made from a fairly large number of the present operating staff, the actions assumed in the SGTR scenarios remair conservative for the Callaway Plant.
s Page 9