ML20138B107
| ML20138B107 | |
| Person / Time | |
|---|---|
| Site: | Comanche Peak  |
| Issue date: | 03/12/1986 |
| From: | Mccracken C Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20138B082 | List: |
| References | |
| OL, NUDOCS 8603200281 | |
| Download: ML20138B107 (13) | |
Text
o l
l UNITED STATES OF AMERICA NUCLEAR REGULATORY COPIMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of )
)
TEXAS UTILITIES ELECTRIC ) Docket Nos. 50-445 COMPANY, et _al._
) 50-446 (Comanche Peak Steam Electric )
Station, Units 1 and 2) )
AFFIDAVIT OF CONRAD E. McCRACKEN I, Conrad E. McCracken, do depose and state as follows:
Q1. What is your name and the purpose of your affidavit?
A1. My name is Conrad E. P'c 'racken. r The purpose of my affidavit is to ,
provide further explanation about core flow blockage due to fine paint particles as requested by the Licensing Board in its Memo-randum of September 18, 1985. Core flow blockage was discussed in Appendix L to Supplement 9 of the Safety Evaluation Report (SSER No. 9) .
Q2. By whom are you employed and what are your duties?
A2. I am currently employed as a Section Chief in PWR-B Licensing i l
Division . During the period in which I worked on Comanche Peak, I was employed by the U.S. Nuclear Regulatory Commission in the Division of Engineering, Office of Nuclear Reactor Regulation. I was Section Chief of the Chemical Technology Section . My duties included evaluating compliance of applicants with the Protective 8603200201BgO 445 PDR ADOCK 0 PDR G
_ _._ . _ . . _ . . _._ - _ _ . _ . - _ _ _ - ~ , - - _ . _ . . _ . _ . _ ~ . _ . . _ . .
Coatings Criteria of Standard Review Plan (SRP) Section 6.1.2.
Additionally, effective November 1984 I was assigned as Group Leader of the Comanche Peak Coatings Technical Review Team (TRT).
Q3. Have you prepared a statement of professional qualifications?
A3. A statement of my professional qualifications was previously forwarded as an attachment to my affidavit submitted to the Licensing Board on October 9, 1985. That statement is accurate as modified by my response to Question '2.
Q4. If an applicant elects to apply qualified coatings inside of the reactor containment building, how does the Staff determine that the coatings are capable of withstanding design basis accident (DBA) conditions?
- A4. Coatings inside of containment are considered to be qualified if they are applied consistent with the positions of Regulatory Guide 1.54 and ANSI N 101. 2, and proper documentation exists to so demon-strate. Additionally, standardized specimens of the same coatings systems as are used in containment are prepared anc' tested under i
DBA conditions by independent laboratories according to ANSI !
N101.2.
Q5. What are the test conditions specified by ANSI N101.2?
A5. The test conditions duplicate those of a larga-break loss of coolant accident (LOCA) because this induces the maximum radiation ,
temperature and pressure conditions within containment.
(
v- , e ~ 'n,, ,, - - -- + + - - - - - ,,---,,m-,, ,.m..e,,-,-.r-- m,- , , , , - - - e ,..m ,
Additionally, the containment pressure results in initiation of the containment spray systems, so testing includes the effects of containment spray solution. i Q6. Will the coatings within the reactor containment building of CPSES Unit No. I withstand a large-break LOCA?
A6. As stated in SSER No. 9, Appendix M, Section 3.4 (p. M-10) , a relatively large percentage of coatings at CPSES are unqualified.
This occurs because adequate traceability (documentation) does not exist. However, as stated in Appendix L, Section 2.1.1.1 (p. L-3),
the coatings applied at CPSES are of the same generic type as have been qualified by other nuclear plants for the DBA environment.
Additionally, the CPSES backfit test program adhesion testing, while not restoring traceability, provides a useful measure of overall adhesion strength and coating thickness. ( Appendix M, p. M-44) f Based on the adhesion test results, coatings thickness measurements, and the fact that nuclear grade coatings were applied, reasonable assurance exists that a significant quantity of the coatings at CPSES would withstand a large-break LOCA.
Q7. Based on your response to Question 6, is the Applicants' assumption in Appendix L, Section 2.1.1 (p. L-2) that all coatings will fail and form debris conservative?
A7. Yes.
l l
l 1
i
Q8. SSER No. 9, Appendix L (pp. L-1 through L-6) provides a discussion of the potential for debris blockage of the containment building emergency sump screens. Please summarize the major points of that discussion.
A8. In the . event of a LOCA that is large enough to deplete the refueling water storage tank. inventory, continued core cooling is provided by recirculating water from the containment emergency sump through the core. This recirculation would normally be accomplished by operation of the residual heat removal (RHR) system for which the pumps take their suction from the containment building emergency sumps. Screens are erected around the emergency sumps to protect the RHR pumps and decay heat renioval systems from debris which can be transported during a LOCA. Excessive blockage of these screens could restrict flow to the RHR pumps' suctions which would diminish flow through the decay heat removal systems. To properly operate , the RHR ~ pumps must have an adequate net positive suction head (MPSH) ~ margin, i.e. an adequate supply of water. The emergency sump fine debris screens are sized at 1/8-inch (0.125 inch) to protect the emergency core cooling and containment spray systems from blockage. (Sect. 2.2) Particles of 1/8-inch or larger are capable of blocking the fine debris screens.
For this analysis all coatings within the containment are assumed to fail and form debris. For the purpose of estimating paint particle transport and fine debris screen blockage, the debris is assumed to consist of 1/8-inch particles. This is conservative because all
particles are assumed to be large enough to cause fine screen blockage yet small enough to enhance transport.
Post-LOCA recirculation flow field analyses demonstrated that containment water velocities remote from the ' emergency sump locations were too low (i.e. , less than 0.2 feet /second) to result in de'bris transport to the emergency sump screens. Therefore, the quadrant of concern was narrov.'ed to the 60c-0-315o azimuthal area which encompasses the emergency sump location. Recirculation flow velocities in some parts of the azimuthal area of concern and near the sump were calculated to be low and would not support transport '
of most of the debris to the screens. Ilowever, the Applicants conservatively calculated blockage estimates by assuming that all the particles in the 600-0-315o region would be transported to *the screens and deposited at a 45-degree angle of repose . (Sect. 2.1.3).
Consideration was also given to coated surfaces in' the immediate vicinity of the sump screens (Sect. 2.1.1) to assess the potential for screen blockage from failure of local coatings which did not have to he transported. Flow field calculations modelling the sump screen and cover plate demonstrated that the small particles (less than, or '
equal to 1/8 inch) would transport, but the presence of the sump
~
screen cover plate overhang results in a velocity distribution that would maintain an open area of at least 24 square feet (a 2-inch band) at the top of the debris screen. In both of the above cases the Applicants predicted that adequate NPSII margin remains , i Therefore , the estimated blockage of the emergency sump screens i
- _ - - -_ . - . - - , - . . . - , - - - - . - . - - , . , - - - - , - - - - - - - - - , - - - - - - - - ~ ~ ~ ~
. would not result in loss of NPSH margin even with the assumption that all coatings fail in.the conservative manner assumed.
The Staff independently reviewed and verified the Applicants' calculations and assumptions and found them. to be conservative and consistent with the guidance and methodology developed by the Staff in conjunction with the work for resolution of Unresolved Safety Issue (USI) A-43, Containment Emergency Sump Performance (Sect. 2.1).
Q9. In Appendix L (pp. L-7 through L-13) there is a discussion of both small-break and large-break LOCA issues. If the large-break LOCA is the design basis for coating survivability why is a. small-break discussed?
A 9 .~ As discussed in response to Ouestion 8, during a large-break LOCA core cooling is provided by the RHR pumps because the reactor coolant system (RCS) is rapidly depressurized. However, during some small-break LOCA scenarios the RCS may remain pressurized above the RHR system capabilities. There fore , core cooling would be provided by the high pressure injection pumps. The small-break LOCAs which are discussed in Appendix L include only a small spectrum of breaks which are large enough to deplete the refueling water storage tank and result in containment spray initiation and the need to recirculate from the sump, yet small enough that the RCS remains pressurized. Under these conditions paint debris may have been generated, transported into the reactor vessel via the high i
i
)
pressure injection pumps on cold leg recirculation and be available for transport to . the core if RCS flow velocities are increased by starting a reactor coolant pump (RCP). This small-break LOCA-scenario formed the basis for the Staff's concerns as expressed in SSER No. 9, p. L-8, which was noted by the Licensing Board in its September 18, 1985 Memorand um .
Q10. In a small-break LOCA will coating failures be as extensive as for a large-break LOCA?
A10. No, the radiation , temperature and pressure conditions inside of containment are lower for a small break than for the design basis large break. Th erefore, lower coating failure rates can be predicted.
Q11. Were these lower coatings failure rates for a small-break LOCA taken into account in the estimates of debris provided for the small-break LOCA?
All. No, the analysis supporting the debris estimates were based on the very conservative assumption that all coatings fail. (See Questions 4 through 7 and the responses thereto.)
Q12. Assuming that all of the coatings fail and 10 cubic feet of deb'i r.
passes through the sump screens ( Appendix L, Sect. 2.5.2), would
- all of that debris be available to produce core flow blockage?
A12. No. Debris which could be generated by coatings failure would be in a variety of sizes. The low water velocities within the l
l .
containment floor pool after ' a LOCA do not generally support transport of particles .as large as 1/8-inch (Sect. 2.1.3), because flow velocities are insufficient. - The maximum debris size passing through the screen is limited to 0.125 inches by the fine debris I
screens. The minimum size particles would ~ bd in the micron size
. range, from chalking or corrosion of the inorganic zine coatings.
3
( Appendix L, Sect. 2.1.1.1) The minimum flow channel dimension in the core is' O.040 inches at the spacer grids. (Sect. 2.6.1)
Therefore, the particles of concern for core flow blockage are only those between 0.125 inches and 0.040 inches which would represent only a portion of the total. Particles of less than approximately 0.040 inches would be transported through the core with the recircu-1 lating fluid if sufficient flow velocity exists to transport them.
Q13. Are the flow velocities from the RCPs sufficient to sweep particles up to 0.125 inches into the core?
A13. Yes. Ifowever, the operators are instructed to trip the RCPs in the event of a small-break LOCA in order to minimize the primary inven-tory loss through the break. At Comanche Peak, the operators are instructed not to restart the RCPs if containment spray has been actuated and transfer to cold leg recirculation performed, except in '
the event of an inadequate core cooling -(ICC) event, because actuation of the containment sprays could result in transportation of damaged paint particles to the sump and cold leg recirculation from the containment building sump using the high pressure injection f .
-. - _ - -- ,,---, .. . . - , - , . , _ _ - - . - . . . - . ~ , _ _ . - . ~ , , , _ . . . . _ . . - - , _ , __._,,--.-s
pumps could introduce paint particles from the sump into the reactor vessel lower plenum.
Q14. Under normal post-LOCA conditions, without a RCP running, wili l
flow velocities be sufficient to transport particles of greater than i 0.040 inches which may have collected in the reactor vessel lower i
plenum to the core?
A14. No. In the case of a large-break LOCA, RIIR flow velocities entering the core would not support particles large enough to result in core blockage. (Sect. 2.6.1) In the case of a small-break LOCA, utilizing high head injection ' pumps, flow velocities through the lower plenum into the core would be less than for RHR.
Therefore , flow blockage during either a large- or. small-break LOCA during the recirculating phase is. highly unlikely.
Q15. V! hat is the probability of an operator inadvertently starting a reactor coolant pump (RCP) during the high pressure recirculating phase and sweeping paint particles into the core?
4 A15. This would be considered an operator error. Applicants are not generally required to assume operator error if it is contrary to specific steps in the emergency response guidelines (ERG), as noted in the accompanying Affidavit of Bernard Mann in response to Question 4.
i ;
4 i
. _ _ . . . _ . - . - . , _ , _ . . _ . , _ , _ , _ . .. . _ . - _ _ _ . . - - _ . _ _ _ - - _ _ , _ . _ . . _ , - . _ _ _ _ . , , . . . . ~ . _ . . . . _ _ , _ . , . .
. Q16. Granting that the Staff would not generally consider an RCP restart due to operator error under these conditions, what is the likelihood of it occurring?
A16. Inadvertent restart of a .RCP is highly unlikely because it would involve multiple operator errors by more than one operating crew member, as noted in the accompanying Affidavit of Samuel D. MacKay in response to Question 4.
Q17. In SSER No. 9, at L-10 the Staff indicates that the introduction of paint debris into the core due to an RCP restart following inadequate core cooling (ICC) is unlikely at CPSES. What is the basis for that statement?
A17. First , ICC is in itself a low probability event. See accompanying Affidavit of Bernard Mann in response to Questions 5 and 6.
Second, a very specific set of conditions must exist to have initiated the event of concern:
- 1. As discussed in response to Question 9, the break size must be large enough to initiate containment spray yet small enough to maintain the RCS in a pressurized condition. Additionally, the RCS must remain pressurized long enough to deplete the refuel-ing water storage tanks - so that high pressure recirculation utilizing the high pressure injection pumps is required.
- 2. The high pressure injection pumps must operate in the recircu-lation mode for a long enough time to introduce paint debris but
then fail, to induce ICC (i . e . , as long as the high pressure 4
injection pumps continue to function ICC should not occur).
- 3. If the high pressure injection pumps do fail, the operating crew is directed by procedure to depressurize and initiate RHR cooling. This backup cooling method must also fail in order for an ICC event to occur.
Q18. Assuming that an RCP is somehow restarted what is the potential for paint debris blockage of the core?
A18. As noted in Appendi:: L, Sect. 2.6.2, a bounding calculation shows that if the maximum estimated debris of 10 cubic feet existed in the i
reactor vessel lower plenum at particle sizes in excess of 0.040 inches it would form a 1-inch thick porous layer at the first grid if evenly distributed. Based on the discussions in response to 2
Q10, 011 and Q12, this is a very conservative estimate. Under these conditions it is reasonable to assume that a "small residual flow" would exist (Sect. ll.G.2), sufficient to provide core cooling.
This is particularly true in the case of a PWR such as CPSES J because of the open lattice fuel bundle design (i.e. , no impediments i
exist to 360o lateral cross-flow once water enters the core area).
l Additionally, the postulated core blockage could not exist until the
- sequence discussed in response to Q17 had occurred. This would be at least a couple of hours into the event and decay heat would be approximately 1% of that generated at power, for which a relatively small flow rate is sufficient to provide cooling. The flow rate i
necessary to provide cooling would represent a fraction of that provided by a reactor. coolant pump. Therefore , significant paint debris blockage of the core is unlikely.
Q19. The Licensing Board has noted that in SSER No. 9 at p. L-12 the Staff concludes that flow blockage must be extensive to result in fuel damage. What is the basis for the Staff's statement?
A19. The Staff has based its qualitative conclusion that flow blockage must be extensive in order to cause fuel rod damage on the boiling water reactor (BWR) data reported at L-12. That data indicates that greater than 79% flow blockage must occur to initiate loss of nucleate boiling. Based on this data the Staff qualitatively considers blockage of greater than 79% to be extensive. This data has been generated for a single BWR fuel bundle at full power. BWR fuel bundles are enclosed in individual channel boxes such that open lateral flow between bundles does not exist, as is the case in PWRs.
Therefore, the BWP. data represents a qualitative worst case for the amount of flow blochage necessary to result in individual bundle overheating (i.e. , if an individual PWR bundle were equally blocked, lateral cross-flow from other bundles would provide additional cooling, resulting in less overheating.) .The BWR data demonstrated that a blockage of greater than 79 percent was necessary to cause loss of nucleate boiling and greater than 95 percent to induce fuel cladding to melt for a fuel bundle at power. For the CPSES case only one RCP would be running, resulting in lower available core flow than for the BWR full power case. However, qualitatively this
would be more than compensated for by lateral cross-flow from other fuel bundles which are not obstructed. Additionally, the decay heat at CPSES would be equal to approximately 1% of full power, which is significantly less than the BWR data which is based on full power.
This results in proportionately lower cooling requirements at CPSES.
Therefore, we qualitatively concluded that average core blockage 4
must be extensive (i.e., greater than 79 percent) to cause fuel rod damage.
J . u !
Mb Conrad E. McCracken L Jn- Q Subscribed and sworn to before me this f ? h day of WirCh 1986 Notary Public kDb6 b My Commission expires: /
i 0
~ ._ .. _ _ _ _ _ . . _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ . . _ . . _ . _ . -
_ _ _ _ _ _ _ . .