ML20052G062
| ML20052G062 | |
| Person / Time | |
|---|---|
| Site: | Maine Yankee |
| Issue date: | 05/11/1982 |
| From: | Garrity J Maine Yankee |
| To: | Clark R Office of Nuclear Reactor Regulation |
| References | |
| REF-GTECI-A-49, REF-GTECI-RV, TASK-A-49, TASK-OR JHG-82-83, MN-82-92, NUDOCS 8205140275 | |
| Download: ML20052G062 (7) | |
Text
..
n g
f MAIRE ~%n 9ARHEE Alom/CPol'HRCOMPARU*
,uaugraja$g$s (p Mil (207) 623-3521
-O May 11, 1982 MN-82-92 JHG-82-83 D
United States tbclear Regulatory Commission e
Washington, D. C. 20555 g
9 Attention: Office of Nuclear Reactor Regulation RECEIVED Division of Licensing 2
MAY 131982* I sperating Reactors Branch #3 S;
m, Mr. Robert A. Clark, Chief n:.cm uau tt p
Subject : Reactor Vessel Pressurized Thermal Shock f
Go
References:
(a) License No. DPR-36 (Docket No. 50-309)
(b) USNRC letter to MYAPCo dated March 18, 1982 (c) MYAPCo letter to USNRC dated January 21, 1982, MN-82-08 (d) MYAPCo letter to USNRC dated December 31, 1981, FMY-81 -189 (e) Letter Southern California Edison to USNRC dated December 31, 1981, " Transmittal of CEN-189, ' Evaluation of Pressurized Thermal Shock Effects Due to Small Break LOCA's with Loss of Feedwater for Combustion Engineering NSSS's,' December,1981" (f) MYAPCo letter to USNRC dated November 2,1981, FMY-81-163 (g) MYAPCo letter to USNRC dated September 29, 1981, FMY-81-148 (h) MYAPCo letter to USNRC dated January 30,1981, FMY-81-11 (i) MYAPCo letter to USNRC dated April 28,1981, FMY 81-65, YAEC-1259 Maine Yankee Cycle 6 Core Performance Analysis
Dear Sir:
Reference (b) requested additional information to supplement our submittals concerning pressurized thermal shock (PTS), References (c), (d), (f) and (g).
This letter provides a response to this request.
Maine Yankee provided a description of a joint program with the Electric Power Research Institute (EPRI) which addressed the issue of PTS in Reference (c).
In the time since the program description was provided, we have focused our attention and concentrated our available resources on moving the program forward.
The program has as its objective development of the ability to perform detailed plant specific analysis of the pressurized thermal shock i
problem. We believe this approach is necessary in view of the potential for the pressurized thermal shock issue to impact upon us through backfits or other operational recuirements which may be imposed and in flew of the level i
of detail reflected in NRC staff questions.
The methods being developed in this program, however, have not been used in developing responses to the staff questions transmitted via Reference (b).
Oh am88%
o P
MAINE YANMEE ATOMIC POWER COMPANY United States Nuclear Regulatory Commission Page Two Attn: Robert A. Clark, Chief May 11, 1982 Our response to the staff's questions are provided below:
Concerninq-Operator Actions Question 1.
In CEN-189, only two cases are considered for a SBLOCA with concurrent loss of feedwater.
In one case, PORVs are opened by the operator at 10 minutes to prevent core uncovery.
In the other case, feedwater is restored to the steam generator in 30 minutes to prevent core uncovery. For both cases, the report stated that 15-30 minutes would provide ample time to initiate feedwater prior to dryout.
Provide the analysis or basis to justify that 15 to 30 minutes is ample time for correct operator action.
Response
The basis for the 15 to 30 minutes as ample time for initiation of feedwater prior to dryout is derived from the complete loss of feedwater transient analysis. Dryout times for Maine Yankee have been determined for the complete loss of feedwater at full power (2630 MWt).
The latest analysis for Maine Yankee is documented in Reference (1).
Assuming a complete and instantaneous loss of feedwater from full power with reactor coolant pumps running, the Reference (1) analysis predicts a steam generator dryout time of about 13 minutes.
Consideration of reactor coolant pumps running continuously is important because it adds an additional source of heat to be removed through the steam generators and delays reactor trip.
i In the Reference (1) analysis, the reactor trip signal is genera'.ed on low steam generator _ level at approximately 26 seconds into the transient.
The delay in reactor trip without feedwater allows 26 full power ' seconds of heat to be rejected through the steam generators maximizing the secondary inventory loss. For the case where the reactor trip signal is generated earlier in the transient, the dryout times are significantly longer than the 13 minutes calculated for the worst case loss of feedwater event. For example, if the loss of feedwater transient is initiated by loss of AC and the reactor coolant pumps trip, the reactor trip signal would be initiated by loss of flow in 2-3 seconds. Under these conditions, dryout time is in the order of 30 minutes because boil-off of secondary inventory is by decay beat for the entire transient except for the first 2-3 seconds.
en
--m,-n
MQlNE VANKEE ATOMIC POWER COMPANV United States Nuclear Regulatory Commission Page Three Attn: Robert A. Clark, Chief May 11, 1982 The secondary inventory response during a SB LOCA without feedwater would behave similar to the loss of feedwater transient with loss of AC.
The reactor trip signal for a S8 LOCA with any appreciable depressurization would be generated on low reactor coolant system pressure within the first few seconds.
A portion of the energy generated in the core would be removed through the break while the remaining energy would be transferred through the steam generators.
Without feedwater, the portion of energy transmitted through the steam generators would bo13 off the secondary inventory. Since the heat transferred to the steam generators during a SB LIOCA would be bounded by a complete loss of feedwater transient with loss of AC, approximately 30 minutes is available for the operator to initiate or restore feedwater prior to steam generator dryout. Dryout times for SB LOCA sufficiently small in size to result in little or no depressurization are bounded by the complete loss of feedwater transient with AC available.
In this case, most of the heat would be rejected through the steam generators.
If the depressurization is slow enough not to reach the low pressure setpoint in 26 seconds then the reactor would trip on low steam generator level.
The secondary inventory response would behave similar to the complete loss of feedwater transient with AC available.
Therefore, steam generator dryout -
times for the spectrum of SB LOCA transients without feedwater are bounded by complete loss of feedwater events. Depending on the rate of depressurization and the fraction of energy removed through the break, dryout times would be in excess of 15 - 30 minutes.
This is a conservative estimate of the time required for operator action to restore feedwater.
Since Maine Yankee's auxiliary feedwater system is automatically i
initiated on low steam generator level operator action to restore feedwater would only be required in the event of loss of both main and auxiliary feedwater systems.
The auxiliary feedwater system is powered from the emergency diesels in addition to the normal AC power source.
l With the multiplicity of information available to the operator, j
the conservatisms in the analysis, compounded with the necessity i
that auto-initiation of auxiliary feed fails, the stated 15 to 30 minutes action time in the CEN 189 analysis is in our judgement adequate.
Question 2.
In CEN-189, provide an evaluation of the sensitivity of the transient to the time assumed for operator action (i.e., if the operator ooens the PORVs at 15 minutes, or 30 minutes, or restores feedwater alone at 15 minutes, or 20 minutes, or 45 minutes, what are the resulting pressure / temperature transients?).
i t
l MalNE VANKEE ATOMIC POWER COMPANY United States Nuclear Regulatory Commission Page Four Attn: Robert A. Clark, Chief May 11, 1982
Response
Figure 4-24 in CEN-189 shows the primary system response to a small break LOCA in which the FW was not restored until 30 minutes after the initiation of the accident.
The system pressure, in this case, decreases initially but the availability of High Head HPSI pumps at MY forces the pressure to go up to approximately 2,400 psia (PORV setpoint). At 30 minutes, the FW is restored. Better secondary heat sink causes a rapid system depressurization followed by a repressurization again to approximately 2,400 psia.
The repressurization is caused by the-Increase in HPSI injection at lower system pressures. Similar primary system pressure response is expected for cases where FW restoration time is varied.
In these variations, the exact time of depressurization and subsequent repressurization will depend on the time at which the FW is restored.
Fig. 4-25 in CEN-189 gives the corresponding temperature response in the cold leg.
The same response is expected for variations in FW restoration times.
In these cases, the time at which the temperature rapidly decreases to a lower limit will be controlled by the time at which the FW is restored.
No PORV opening time study was performed for the High Head HPSI system. However, initial pressure response for the PORV opening case is exoected to be similar to the FW restoration case described above.
The pressure will drop at first, but the High Head HPSIs will inject enough water to' increase the system pressure to the PORV setpoints.
The pressure will stay at 2,400 psia until the second PORV is opened by operator action.
At this time, the pressure will slowly start to drop in a manner similar to the PORV Cases reported in CEN-189.
l The initial temperature response will be similar to that presented in Fig. 4-25.
The initial temperature will be i
j maintained until an additional PORV is opened.
The fluid temperature then will start to decrease slowly as is shown in i
I other PORV cases reported in CEN-189.
Question 3.
In the Maine Yankee-specific analysis in CEN-189, Auxiliary Feedwater ( AFW) was assumed to be controlled by the operator to maintain the S/G at normal operating level. What is the resulting pressure / temperature transient if AFW flow remains uncont rolled?
i
Response
Failure by the operator to control the steam generator at the normal operating level superimposes a runaway feed water transient upon the small break LOCA.
This event sequence is j
considered beyond credibility and inconsistent with best estimate analysis.
j
)
j
MAINE YANKEE ATOMIC POWER COMPANY United States Nuclear Regulatory Commission Page Five Attn: Robert A. Clark, Chief May 11,-1982 The effect of overfilling the steam generator through multiple failures is to increase the primary system cooldown rate and depressurization, however, the temperature of the vessel at'the time of repressurization may not be significantly altered.
Question 4.
In your evaluation, the actions described do not provide the operator with clear direction for dealing with conflictina concerns that need to be evaluated when considering the operation of HPI and the charging flow as it relates to vessel integrity _
and maintaining core cooling.
Provide an evaluation of the need and effectiveness of procedure modifications to clearly identify the concerns in the emergency operating procedures themselves.
This should be done in contrast of depending upon upgrading operator training alone.
Response
Reference (c) provided a list of operator actions which are required to prevent pressurized thermal shock and to ensure vessel integrity.
In particular, Maine Yankee provided a summary of instructions for high pressure safety injection (HPSI) system operation and termination. Maine Yankee believes that operation of the HPSI system according to these instruction will ensure -
both vessel integrity and core cooling.
Maine Yankee is in the process of upgrading its' emergency operatirg procedures (80Ps) to fulfill the Task I comments for Action Plan Item I.C.1, Reference (h).
This process has explicitly addressed the issue of PTS and has resulted in the current instructions summarized above. ' Although other minor differences primarily relating to multiple-failures and operator errors have been discovered, it was concluded that the current EOPs provide adeouate guidance pertaining to prevention and mitigation of PTS in the interim period before procedures for I.C.1 are implemented (fall 1982). This conclusion assumed that i
the E0Ps would be supplemented by the PTS training program
~
described in Reference (c).
Procedure revisions to fulfill I.C.1 will reflect Maine Yankee's philosophy regarding the level of detail provided in E0Ps and that relegated to the bases documents and the operator training program.
l Concernino Probabilistic R{$ j g "aen: of Over Cooling Transients Question 1.
Provide existiN cowetation or references of such documentation related to Probabilistic Risk Assessments which would provide insight into the probabilities of overcooling events at your plant.
y v.
MAINE YANKEE ATOMIC POWER COMPANY United States Nuclear Regulatory Commission Page Six Attn: Robert A. Clark, Chief May 11, 1982
Response
Maine Yankee has no new information in this area.
In the event Maine Yankee performs a plant specific FRA in the future serious consideration will be given h, the inclusion of accident initiators and sequences leading t-overcooling /repressurization and potential foi c..'llenging reactor vessel integrity.
Concerning Overcooling Transients at -Your Plant QJestion 1.
Review the operating history at your plant and identify all overcooling transients as well as those events which could have become overcooling transients if not mitigated by plant controls on operator actions.
Provide summaries of the identified events.
Response
In a review of Nuclear Safety Audit and Review Committee Records, no events were discovered in which reactor coolant system average temperature was reduced by 1000F in one hour or less.
Thus Maine Yankee concludes that no overcooling event has occurred at the Maine Yankee Plant.
The records reviewed include Plant Information Reports, Licensee Event Reports, Abnormal Occurrence Reports, and specific Ieports to the NSAR Committee.
They date back to initial commercial operation in December of 1972. These records were supplemented, when necessary, by interviews with operations personnel.
Two incidents were discovered which had the potential to lead to overecoling transients because they involved large steam loads being imposed on the plant.
On January 14, 1973, a transistor failed in the steam dump valve temperature controller causing the twelve steam dump valves to i
cpen.
The operator terminated this transient by closing the main steam excess flow check valves and non-return valves.
The average reactor coolant system temperature decreased 430F.
Subsequently, plant design change number 10-73 eliminated the transistor in question and the potential that this single failure would cause another transient.
l On February 4,1973, during testing and adjustment of the turbine governor valves, these valves opened, resulting a rapid increase l
in steam flow.
Automatic operation of the excess flow check valves and prompt operator response minimized the effects of this j
event. Average reactor coolant system temperature decreased 200F.
l i
MAIP0E YANKEE ATOMIC POWER COMPANY s
United States Nuclear Regulatory Commission Page Seven Attn: Robert A. Clark, Chief.
May 11, 1982 If there are any questions, please do not hesitate to call.
Sincerely, MAINE YANKEE ATOMIC POWER COMPANY 6
John H. Garrity, Senior irector Nuclear Engineering &. Licensing ~
i i
1 i
9 f
,p y.
,-,,y-.-.
y, y-
,-v,w--
m.,
- 6_m,
-4
%-*-c
,v-
-. = -