ML20117J021
| ML20117J021 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 05/20/1996 |
| From: | Mccollum W DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9605300074 | |
| Download: ML20117J021 (16) | |
Text
..
- e ll l
Duke h>uer Comparty WtuletR Afc0>unt,JR Catan ba Nuclear Genemtwn Department lice Pnulent 4SM Concord FHad (b10)S31J200 Olliie krk, SC29715 (813)831J426 Far i
DUKEPOWER May 20, 1996 U.
S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, DC 20555 1
Subject:
Catawba Nuclear Station, Unit 1, Docket No. 50-413 Response to Request For Additional Information By letter dated January 26, 1996, Duke Power Company submitted a proposed one time Technical Specification change that would allow operation of the Containment Purge Ventilation System during Modes 3 and 4 following startup from the steam generator replacement outage.
On April 26, 1996, the NRC Staff provided a Request for Additional Information concerning the proposed Technical Specification change.
Attached are responses to each of the Staf f's seven questions.
The Staf f's RIA focused on the ability of the containment purge isolation valves to close under design basis accident conditions and on the need to purge containment following steam generator replacement.
It is our conclusion that purging of lower containment during heatup will be necessary for the following reasons and as discussed more fully in our January 26, 1996 submittal:
1.
Thermal decomposition products (TDP) will be produced which could cause a respiratory hazard to personnel.
Expected sources of TDP include lubricants, cutting fluids and nondestructive examination gels on system piping and the new steam generators.
In an attempt to eliminate a significant portion of the TDPs, the new insulation is being pre-heated.
2.
Concentrations of TDPs in the Catawba Unit 1 I
containment are expected to exceed allowable 9605300074 960520 PDR ADOCK 05000413 P
PDR t
i; t
k
- aweumwws,
U.
S. Nuclear Regulatory Commission May 20, 1996 Page 2 concentrations for personnel expocure.
This is due in l
large part to the small net free volume of an ice 3
3 condenser containment (237,000 f t vs. >2,000,000 ft for a typical dry containment).
l 3.
Activities such as support gap measurements, system leak checks and routine surveillances make containment entry necessary during heat-up following steam generator replacement.
4.
In order for personnel to perform these necessary tasks I
in lower containment, measures must be taken to reduce the expected concentrations of TDPs.
OSHA Standard 1910.134 recommends engineering control measures (i.e.,
ventilation), with use of respiratory personal protective equipment as a last resort.
The proposed purging of the Catawba Unit 1 containment does not present a significant risk for the following reasons:
1.
The reactor will be subcritical during the heatup process in Modes 4 and 3.
Approximately one third of the core will contain new fuel with no fission product inventory.
The remainder of the core will have decayed approximately 90 days prior to entering Mode 4.
Therefore, the source term fission product inventory is significantly less than during power operation.
2.
The proposed purging of lower containment is a one time occurrence that would take place over approximately 5 days.
3.
The probability of a LOCA is considered extremely small and no greater than before the outage.
The reactor coolant piping welds to the new steam generators will be performed and inspected to the ASME Code.
Leak-before break analyses that were submitted and approved by the NRC (K. N. Jabbour's letter of April 7, 1987) demonstrated that a double ended break of the reactor coolant piping is not physically possible.
4.
The lower containment purge valves utilize air to open, spring force to close actuators.
These actuators have l
proven to be highly reliable in routinely closing the valves following operation of the purge system during 4
~
t 4
- U.
S. Nuclear Regulatory Commission May 20, 1996 Page 3 l
outages.
Leak 1 tight closure is not necessary during l
the proposed purge of lower containment due to the l
small-radiological source term.
Leakage of 100% of the containment volume during the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> following i
a postulated design basis LOCA still resulted in acceptable offsite doses.
In summary, purging the Catawba Unit 1 lower containment following the upcoming steam generator replacement outage is considered ~necessary'and prudent in protecting plant personnel and would not pose a significant hazard to the public.
It is therefore requested that the NRC provide expeditious approval of the requested Technical i
Specification change.
If additional information is required, please call Robert Sharpe at (704) 382-0956.
i i
Very truly yours, l
c:
l W.
R. McCollum, Jr.
l Attachments xc:
S.
D. Ebneter l
Regional Administrator, Region II l
U.
S. Nuclear Regulatory Commission 101 Marietta Street, NW, Suite 2900 Atlanta, GA 30323 R.
J.
Freudenberger Senior Resident Inspector Catawba Nuclear Station P.
S. Tam j
Project Manager, ONRR
?
l R. E. Martin l
Project Manager, ONRR l
l p
w
~
_e rw-
...m.
--,e
.i-,---
Duke Power Company Catawba Nuclear Station, Unit 1 Response to Request For Additional Information
- 1) Performance Testing Could a diagnostic test measuring torque requirement and output be performed during refueling to simulate the design-basis differential pressure and flow (for example, using purge fans) ?
Licensee information indicates that closing torque is required at the zero-degree position.
Could a valve leak test be performed after the dynamic test to more
. closely approximate leakage under accident conditions?
Response
A diagnostic test measuring torque requirement with simulated design-basis differential pressure cannot be performed.
The differential pressure taken to exist across the containment penetrations following a design-basis LOCA (DBLOCA) was the lower value of 13.27 psid or the limiting value associated with choked flow through the penetrations.
The former value was found to be the peak blowdown pressure in any containment subcompartment adjoining a Containment Purge penetration.
The maximum differential pressure across the Purge containment penetrations when the Purge fans are in operation is only 2.3 inwd (0.08 psid).
The maximum pressure drop from the Purge containment exhausts to the suction of the Purge fans when these fans are in operation is 14.5 inwd (0.52 psid).
Therefore, the Containment Purge fans cannot be used to simulate the conditions taken to exist in the Purge Containment Penetrations during the blowdown phase of the DBLOCA.
A test can be performed to determine the spring output of the Purge containment isolation valves.
It is not necessary to simulate the conditions associated with the DBLOCA to perform this test.
Tests associated with Category A valves will be performed pursuant to ASME Section XI.
These include both stroke time tests and leak rate tests.(cf. Response to Question 2 below).
The required closure torque for the Purge containment isolation valves following a DBLOCA is negative for most of its travel cycle (80 through 10 inclusive).
The closure torque at 0 is the torque required to overcome seat friction and move the valve into its seat.
No spring torque is required to keep the valve within its seat.
For these reasons, it is determined that these tests on the valve are sufficient to provide assurance that the valves will close as required following the DBLOCA, 1
i e
l l
[
- 2) Inservice Test Requirements These' valves are classified as Passive Category A in the inservice testing (IST) program submitted to the staff.
l Will they be tested in accordance with ASNE OM-10 as Active j
Category A air operated valves prior to purging operations?
Will fail safe test and remote position verification be l
included?
Will leak testing be performed following fail I
safe test?
Response
The Unit 1 Containment Purge isolation valves which will be opened in Modes 4 and 3 are considered to be l
Active Category A valves-for the duration of this purge operation.
These valves will be tested as such prior to Unit 1 entry into Mode 4.
The tests will be conducted in accordance with Section XI of the ASME Code, (1983 Edition),
Subsection IWV (Inservice Testing of Valves in Nuclear Power Plants).
The tests to be performed on each affected valve will include a stroke time test, a fail safe test (cf. IWV-3415) and a remote position indication verification (cf.
IWV-3300).
Leak rate tests also will be performed before Unit 1 entry into Mode 4.
In addition, a leak rate test will be performed on each affected valve after the purging operation is completed but prior to entry into Mode 2.
The status of the Containment Purge isolation valves which will not be opened in Modes 4 and 3 remains unchanged (Passive Category A).
Leak rate tests only will be performed on these valves prior to entry into Mode 4.
- 3) Butterfly Valve Performance Has the Fisher torque requirement prediction methodology for the specific butterfly valve, size, disc type, and aspect ratio used at Catawba been evaluated against butterfly valve testing, such as the EPRI butterfly valve test results and model, or INEL purge testing?
Have flow path characteristics such as elbows or tees been reviewed against assumptions in the analysis of the torque and self-closing performance under design-basis loss-of-coolant accident (DBLOCA) condi tions ?
Has performance of the elastomer seats been monitored?
When were the elastom:r seats last replaced?
If the assessment is for new valves, will deteriorated elastomers cause problems with the operation of the valves?
2
The Fisher report uses information obtained during a l
laboratory test of a 6" butterfly valve and appears to extrapolate the test results to a 24" butterfly valve.
ANSI B16. 41 Annex J recommends extrapolation proportionality limits of 50 to 200% of nominal pipe diameter.
Were tests of any larger valves used to verify the qualification calcula tions ?
Does the fact that Duke Power Company (DPC) uses a 7600 disc in a 9200 valve affect its analysis?
l Has the licensee considered the effect on torque requirements resulting from the purge system operation during a Design Basis LOCA with containment pressurization?
If the air operated valve vents pilot air inside containment, has the containment backpressure effect on closing torque margins been considered?
Response
The methodology used by Fisher to predict required torque for the specific butterfly valve size, disc type, and aspect ratio used at Catawba Nuclear Station has not been compared with either the EPRI butterfly valve performance prediction model or the INEL purge valve testing program.
However, the Fisher torque prediction methodology for the Catawba Containment Purge butterfly valves was developed through the use of scaling techniques and nondimensional variable analysis (disc aspect ratios, for example), making use of data from a series of controlled laboratory flow tests of representative butterfly valve models.
This approach basically mirrors the approach used by EPRI in developing the performance prediction program for butterfly valves, in which the results of the INEL purge valve testing program were incorporated.
While Fisher has provided same laboratory test data in the subject report for viewing purposes, the collective results are considered proprietary by the company.
Flow path characteristics, such as elbows and tees, were not considered in the analysis of required operating torque for the Catawba Containrent Purge butterfly valves following a postulated DBLOCA.
F'ow path disturbances such as these would affect only the a'rodynamic component of the total required operating torque.
For the subject butterfly valves, the analysis demonstrated (as expected, cf. Fisher, Engineering Response /NRC Questions, May 31, 1983), that for most of the valve travel cycle, the closing direction, the aerodynamic torque is negative (helps to close the valve) and dominant.
Thus, if any flow path disturbance were to cause the magnitude of the aerodynamic torque component to increase at any time following a postulated DBLOCA, the 3
A self-closing behavior of the butterfly valves would be enhanced and the margin would be increased.
- The elastomer seats on the Catawba Containment Purge butterfly valves are maintained in satisfactory condition through ongoing stroke and " Type C" leak rate tests every-refueling outage.
The elastomer seats are replaced on an "as-needed" basis, as dictated by their performance during the tests.
f I
Deteriorated elastomers will not affect the self-closing performance of the subject' butterfly valves, as the valve l
discs are out of their respective seats.
Deteriorated seats would be a concern primarily in the opening direction (during disc unseating).
However,-the subject analysis is focused on the ability of the valves to close.
Regardless, the satisfactory condition of the elastomer seats to perform the seating function of the valves will be verified through
" Type C" leak rate testing before the butterfly valves are operated in Mode 4 or Mode 3.
The cumulative effort-of the Fisher laboratory testing which L
was performed to develop the torque prediction methodology used in the calculations is not known, as information concerning the program is considered proprietary by the i
company.
The qualification analysis was performed by Fisher Controls, who originally supplied the Catawba Containment Purge butterfly valves.
The subject calculation performed by Fisher documents the applicable valve serial numbers.
l Therefore, all pertinent disc design _ considerations have been appropriately taken into account.
The liquid and gas sizing coefficients used to calculate the differential pressures across the valves following a DBLOCA are associated with the Type 7600 disc.
Fisher has stated that the sizing coefficients for the Type 7600 disc and the Type 9200 disc are identical.
The sample of laboratory test data supplied-by Fisher in the subject calculation is associated with a six inch test valve.
Again, the total magnitude of the Fisher laboratory testing program is not known to Duke Power Company.
Regardless, Fisher used scaling and nondimensional analysis techniques similar to those used by EPRI in extrapolating l
the laboratory test data across various butterfly valve i
sizes.
Additionally, in the EPRI MOV Performance Prediction Program Butterfly Valve Test, EPRI concludes that flow and torque coefficients for butterfly valves are true nondimensional parameters and as such do not depend on flow rates, valve differential pressures, or valve sizes.
i i
1 4
l E
l l
f l
The effects of the operation of the Containment Purge Ventilation System on the analysis of the torque requirements was not considered.
The differential pressures across the Purge Containment penetrations are very small compared to the differential-pressures taken to be across i
these penetrations immediately following a DBLOCA (cf.
l response to Question 1 above).
Therefore, it is judged that l
the prior operation of the Containment Purge Ventilation System would have no significant adverse effect on the calculation of DP across the valves and the resulting l
analyses of torque requirements.
The inboard Purge containment isolation valves are vented'to containment.
Both the pressurized side and'the unpressurized (spring) side of the actuators of the Containment Purge valves are vented to the ambient.
Therefore, no contribution should be made to the pressure differential across the piston as a result of the increase in containment pressure.
- 4) Actuator Qualification and Capability It appears from the information provided that the valves' l
control systems are not engineered safety features (ESF).
How does the fact that the operators and controls are not ESF grade affect assumptions used in determining reliability under DBLOCA conditions?
Are the valves capable of fail-safe closing in the event that power and controls to the valve actuators are lost?
Is the actuator rating adequate for its requirements?
It appears that the torque requirement exceeds spring torque output at greater than 50 degree open position.
Discuss reliance on flow to enable the valve to close.
What is the assurance that sufficient flow will occur to assist in the closure of the valve?
Valve testing at INEL indicates that supersonic flow existed downstream of the valve during most of the valve cycle.
This could affect the linearity of calculations using differential pressure (reference NUREG/CR-4141).
Discuss the reliability of the 80-degree open limit.
Will limit stops be installed if not already in place?
If not, how will Duke control the limit ?
Fisher recommends that both valves be limited to 70 - 80 degrees open, but the DPC information appears to limit only the outboard valve.
i i
l 5
What are the results of the component structural integrity analysis?
Response: The actuator vent solenoid valves on the CNS containment purge butterfly valves, and the associated electrical controls which cause them to close on a " Phase A Containment Isolation" (Sr) signal, are Class 1E: Nuclear Safety Related.
On a loss of either control power or instrument air, the containment purge butterfly valves will fail to the closed position.
The closing torque requirements for the inboard and the outboard butterfly valves are all negative for all open angles up to 80 degrees under the postulated DBLOCA conditions.
At 90 degrees (full) open, the inboard valves require 7,103 in-lbf of torque to initiate closure, and the outboard valves require 15,188 in-lbf of torque.
The Bettis actuators are rated for 12,930 in-lbf of torque at the 90 position.
It is clear that the inboard valve requirements are all well within the capabilities of the Bettis actuator.
The outboard valves will be limited to an opening angle of no more than 80, at which the torque requirements all become negative (until seating).
With this stroke limitation on the outboard valves, the operator rating is adequate to ensure closure of both the inboard and the outboard butterfly valves.
For the inboard Containment Purge butterfly valves at Catawba Nuclear Station, the net stem torques required for valve closure under DBLOCA conditions are all within the capacities of the Bettis actuators, as demonstrated through the torque prediction analysis and stated by Fisher in Item D.2.c (pg. 6 of 7) of the qualification report.
For the outboard butterfly valves, limiting open travel to 80 or less will ensure sufficient actuator torque to complete valve closure, as stated by Fisher in Bullet 6 (pg. 1 of 1
of the summary of the qualification report).
The appearance that "the torque requirement exceeds spring torque output" is based en an analysis of the torque required to open the valves.
This is apparent from comparing Report FOR-49, Rev A, Item B.3.a with Engineering Response /NRC Questions, May 31, 1983, (pg. 10/20).
In the latter document, Fisher has stated that "At most open angles flow tends to close the valve so the operating torque is i
that required to open the valve farther or hold it open."
The qualification report demonstrates only that DBLOCA flow / pressure conditions will provide a significant degree of assistance to close both the inboard and outboard 6
.= _
butterfly valves.
Fisher was advised that "A - sign is to indicate that the NET effect of a differential pressure condition at (that) valve disc angle is to close the valve without the help of the actuator."
This does not imply that reliance is being made upon this valve behavior in order to complete the closing function.
The Bettis actuators are available to provide stem torque to close the valve (required to initiate closure of the inboard valves and to seat the inboard and outboard valves).
The analysis does provide a basis to conclude that the valves will close within the time recorded in the valve stroke tests.
A computer generated analysis was performed in order to generate the postulated differential pressures across the inboard and outboard butterfly valves under DBLOCA conditions.
The analysis took into account various flow characteristics, such as choked and sonic flow conditions, in determining the respective differential pressures.
For valve disc angles at which choked flow conditions in the Purge containment penetrations were predicted, differential pressures across the valves were calculated based on limiting the flow rates to the sonic limit.
This approach is based on a review of the paper of Buresh and Schuder, presented at the 18th Annual ISA Conference and Exhibit, September 9-12, 1963 in Chicago.
Based in part on this
)
analysis, the qualification report prepared by risher represents the most comprehensive and detailed analysis performed on the butterfly valves to date.
New Bettis air actuators will be installed on the outboard butterfly valves which will be used to perform the subject containment purge.
These actuators will be equipped with travel limiters to limit the full open position.
The travel limiters are a part of the standard specification options for Bettis actuators, and are not a special design consideration for CNS.
Limiting the open direction travel of the outboard butterfly valves to 80 is consistent with Fisher's recommendations (FOR-49, Rev. A).
Fisher did not recommend limiting the open direction travel of the inboard butterfly valves.
Implementation of the 80 open direction travel limiters will be controlled by Engineering Instructions in the modification package which installs the new actuators on the outimard butterfly valves.
Additional I
verification will be provided through Post Modification Testing requirements.
In Bullet 1 (page 1 of 1) of the qualification report's summary, Fisher states that " critical valve components can withstand the specified differential conditions without approaching yield."
7
(
r S) Emergency Procedures What is the contingency plan to mitigate consequences if l
valves do not close (such as emergency operating
{
procedures)?
How long would it take to access the valves and close at least one in each line?
What is the risk of a l
design basis LOCA over this period?
l Response: The Containment Purge Ventilation Isolation Valves are designed to " fail closed" on an Sr signal.
In turn, this signal is initiated on generation of a " Safety Injection" signal.
Should a safety injection event occur, the site emergency procedures (EP's) direct the operators to verify proper response to the Sr signal.
If the proper response is not obtained, the operators are directed to manually close the valves which have not closed in response to the Sr signal.
These procedure steps enable the 1
operatcrs to recover from a control failure (i.e.,
failure to generate the Sr signal or failure of the valve controls to initiate closure on receipt of an Sr signal) from within the Control Room.
It has been determined that the valves will fail closed on loss of either instrument air or power (cf. the reeponse to Question 4).
Each containment penetration associated with the. Containment l
Purge Ventilation System is equipped with two valves.
The controls which cause each valve to fail closed on the Sr signal are associated with a separate and redundant Class 1E train.
The analysis of the response of the valves to a DBLOCA, reported in the Duke Power Company (DPC) submittal of January 26, 1996, was based on the assumption that one valve in each penetration remained " wide open" (a single failure associated with the minimum safeguards assumption).
i The analysis determined that even if one Purge Isolation valve in a containment penetration remained open as a result of a single failure, the second valve would close following a DBLOCA.
The manufacturer of the valve has recommended that the travel of the outboard valves be limited to 80 to ensure that these valves will close.
DPC will replace the actuators for the outboard valves with those equipped with travel stops in order to limit travel to 80 or less.
If a Purge isolation valve fails to close following receipt of an Sr signal, there are no operator actions that can be taken to manually close the valve.
The inboard Purge isolation valves are inside containment while the outboard valves are in the annulus.
Neither are accessible following 8
a LOCA.
In addition, the valve assemblies are not equipped l
with handwheels for manual valve closure.
Risk of the DBLOCA is a composition of the probability of the DBLOCA sequences resulting in radiological consequences and the associated consequences.
In answering the question "What is the risk of a design basis LOCA over this period?"
l both the probability and consequences of a DBLOCA occurring at Unit 1 in Modes 4 or 3 immediately after the steam generator replacement outage with its Containment Purge System in operation were evaluated.
The probability of the occurrence of a DBLOCA is not increased either with the replacement of the steam generators or with the proposed i
operation of the Unit 1 Containment Purge System.
The work to replace the steam generators will be performed pursuant to Section XI of the ASME Code.
Welds will be inspected pursuant to the ASME Code and a leak test will be performed in accordance with Code Case N416-1 as approved by the NRC.
The probability of occurrence of the following events was estimated based on information available in the Catawba Probabilistic Risk Assessment (PRA):
- 1) a DBLOCA (the initiating rupture only), and
- 2) a DBLOCA followed by inadequate core cooling conditions (complete loss of both trains of the Emergency Core Cooling System (ECCS) either in the injection phase or the recirculation phase).
The time in which the Unit 1 Containment Purge Ventilation System is to be operated.during Modes 4 and 3 is estimated to be 5 days or less.
(This includes time to make any necessary adjustments to the shims following the initial
" hot" inspection.)
The DBLOCA is a double ended guillotine rupture of a main reactor coolant pipe segment.
As such, it is associated with the class of large break LOCA's (LBLOCA's, determined in the Catawba PRA to be any break in reactor coolant piping of equivalent diameter of at least 5").
The anticipated frequency of the DBLOCA has been estimated to be 1.75x10-*/rxr-yr or less, based on an extrapolation of no LBLOCA's in experience of the U.S.
nuclear industry to date.
The associated probab.tlity of a l
DBLOCA over a 5 day period is estimated to be 2.4x10-' or less.
From a review of the Catawba PRA, the probability of a DBLOCA followed by inadequate core cooling conditions is estimated to be 3.3x10-' or less.
Both offsite doses and doses to the Control Room operators l
were calculated based on the assumptions described in the 9
l1.--
l January 26, 1996 submittal.
The " source term" used for L
these analyses were calculated in accordance with Regulatory Guide-(RG) 1.4 and TID-14844.
The offsite doses were determined to be within the limits of 10CFR100.
The doses to the Control Room operators were determined to be within l
the limits of Standard Review Plan Section 6.4.
(Duke Power l
committed (May 6,.1996 letter from W. R. McCollum) to provide the Control Room operators at Catawba Nuclear Station with eye protection if beta activity or noble gases are detected in the Control Room following an accident with L
the Unit 1 Containment Purge System initially in operation.)
This analysis demonstrates that under licensing basis assumptions, the consequences of a DBLOCA at Unit 1 while j
l its Centainment Purge System is in operation is comparable l
to the risk of a DBLOCA at Unit 1 with it initially in Mode L
1 and with its. Containment Purge isolation valves sealed closed pursuant to TS 3/4.6.1.9.
As noted above, both the source terms and the radiological consequences were analyzed in accordance with RG 1.4 and TID-14844.
As reported in the submittal of January 26, 1996, DPC also considered the proposed evolution from a l
risk-based perspective.
A separate series of analyses was performed to determine the time dependent fuel temperature,
)
clad temperature, and source terms following a DBLOCA.
As an initial condition, the core was assumed to be uncovered j
and without flow from the ECCS.
In one analysis, an adiabatic boundary condition was applied to the clad while i
heat transfer between the fuel and clad was computed, including heat generated from zirconium-water reactions.
The results of that analysis are as follows: the operators have at least 71 minutes to recover cooling water flow to the core before significant amounts of fuel-clad gap j
activity is released from the fuel pins (beginning at 1200 F).
They have at least 100 minutes before a significant 4
amount of activity is released from the fuel itself.
A l
second analysis was performed in which credit was taken for L
natural convection to steam or air.
It was shown that under these conditions the fuel would not reach the lowest temperature associated with either fuel damage or release of any significant amount of activity (1200 F).
In an additional analysis, it was determined that the flow rate of l
water to the core needed to make in) loss of coolant from boiling is 12.5 GPM or less.
This is well within the capacity of any one ECCS pump.
l It is concluded that the risk of a DBLOCA at Unit 1 with it l
initially in Mode 4 or 3 with its Containment Purge
(
Ventilation System in operation is insignificant.
The risk l
may be compared favorably to the risk of a DSLOCA at Unit 1 i
10
over the same period with it initially at power or in startup and with its Containment Purge isolation valves sealed closed.
- 6) Possible Alternative Procedures On what basis did DPC " determine" to inspect the new supports in a hot condition rather than cold?
It appears that the current inspection requirements would allow the inspections in the cold condition to be acceptable.
Is a i
" hot" inspection a concern that overrides placing the plant in a condition that is only conditionally qualified (i.e.,
allowing the 24" purge valves to be open during some period of time in Modes 4 and 3) ?
Could the " hot" inspection be done at any other time and accomplish the same purpose?
Could the building be cleared of toxic gases before changing modes such that the purge valves could be closed?
How was this issue addressed at other plants replacing steam generators which do not appear to have requested a similar TS change?
In that this request relates to NUREG-0737 Item II.E.4.2, discuss the particular hardship if purging is not allowed in Modes 3 and 4.
Response
The gaps for the Reactor Coolant System supports are shimmed at cold conditions based on predictions of the thermal movement of the Reactor Coolant System piping and components.
Plans call for checking these gaps at 340, 450 and 557 F during the initial heat-up following steam generator replacement.
If any of the support gaps are found out of tolerance, the schedule provides for cooling the unit down to nearly cold conditions for shim adjustment followed by a second heat-up Verification of support gaps is not the only activity in containment that necessitates personnel entry.
Personnel will also be required to perform leak checks on a number of piping system welds per ASME Code Case N 416-1.
- Also, during the heatup process, personnel routinely enter containment to perform surveillances.
l The heat up of the Reactor Coolant System, connected systems and the containment building results in the production of thermal decomposition gases as discussed in the January 26, 1
11 l
s 1
1996-submittal.
These gases are not produced prior to heat j
up to Mode 4 and above.
The Catawba steam generator replacement has two primary factors which increases the risk of respiratory hazards verses other previous steam generator replacement projects.
These factors are 1) the steam generators and interfacing piping will be entirely covered with new fiberglass thermal insulation,.and b) the Catawba ice condenser containment design provides a relatively small ventilation-volume within which respiratory hazardous gases are concentrated.
First, the new fiberglass insulation covering the steam generators and interfacing piping is a major source of respiratory hazards.
Analysis has been performed to identify and quantify the thermal decomposition gases from fiberglass insulation.
Much of this analysis has been conducted as a result of respiratory incidents.such as a fiberglass insulation replacement at Calvert Cliffs Unit 2.
This unit replaced the insulation on the reactor coolant system piping with the same type of insulation that has been purchased b3 Duke Power Company for Catawba Unit 1 and McGuire Units 1 and-2.
They were not replacing steam generators.
Following heat-up from their End of Cycle 8 refueling outage, Baltimore Gas and Electric (BG&E) experienced off-gassing from the insulation and the heat-up of fluid residues which resulted in high levels of carbon monoxide inside the containment. building.
By letter dated April 6, 1991, BG&E requested a Regional Waiver of Compliance to allow purging of the containment in Modes 4 and 3.
A Temporary Waiver of Compliance was subsequently granted by Thomas T. Martin's letter of April 9, 1991.
Second, the Catawba ice condenser containment design with a relatively small ventilation volume concentrates respiratory hazardous gases relative to other units such as Calvert j
Cliffs.
The volume of the Catawba Unit 1 lower compartment (237,000 cubic-feet) will concentrate similar respiratory hazardous source to more than 8 times the concentration relative to a Calvert Cliffs Unit 2 (2,000,000 cubic feet).
Duke Power Company recognizes the respiratory hazard of the steam generator replacement outage and is seeking engineering controls to minimize exposure to personnel.
This includes operation of the Containment Purge Ventilation System during modes 4 and 3 and avoiding the use of respiratory personnel protection equipment.
The use of engineering controls to avoid use of respiratory protection 12
l i
is in compliance with permissible practice as stated in respiratory protection standards such as OSHA 1910.134.
1 I
As discussed in the January 26, 1996 submittal, the initial heat-up of the steam generators and interfacing piping systems is expected to produce thermal decomposition product j
gases which could cause a respiratory hazard to personnel.
It is expected that the concentration of respiratory hazardous gases inside containment will exceed personnel.
' allowable exposures.
If the Containment Purge Ventilation l
System is not operated during heat-up (Modes 4 and 3), the concentration of respiratory hazardous gases will buildup in containment.
Performance of necessary tasks'in containment (checks of support gaps and system leak checks) would require use of respiratory personal protective equipment (PPE).
Use of PPE's in containment would limit working time, would make'it difficult to reach tight spaces in lower containment, and would increase heat stress on personnel-already experiencing lower containment temperatures approaching 100 F.
- 7) Other Valves Discuss the butterfly valves used for incore instrumentation room penetration and their status related to the technical specification amendment.
In one place, DPC states that the i
valves will remain closed, but in another place, it says i
that the capability to close the 12" valves is bounded by the 24" valve.
Does Fisher agree with this statement ?
Response
The 12" valves in the penetrations to the Incore Instrument Room at Catawba Nuclear Station will not be used during the proposed containment purge.
1 t
l l
u
--..n