ML20069A664
| ML20069A664 | |
| Person / Time | |
|---|---|
| Site: | Cooper  |
| Issue date: | 06/12/1981 |
| From: | Delgaizo T, Kaucher J Franklin Research Ctr, Franklin Institute |
| To: | Huang Y Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20069A667 | List: |
| References | |
| CON-NRC-03-79-118, CON-NRC-3-79-118, TAC 11040 NUDOCS 8106170216, TER-C5257-013, TER-C5257-13 | |
| Download: ML20069A664 (20) | |
Text
.
1a l
TECHNICAL EVALUATION REPORT CONTAINMENT LEAKAGE RATE TESTING NEBRASKA PUBLIC POWER DISTRICT COOPER NUCLEAR STATION NRC DOCKET NO.
50-298 N RC TAC NO. 11040 FRC PROJECT C5257
,NRC CONTRACT NO. NRCA3-79-118 FRC TASK 13 Prepared by Franklin Research Center Author:
J.E. Kaucher The Parkway at Twentieth Street Philadelphia, PA 19103 FRC Group Leader:
T. J. DelGaizo Prepared for Nuclear Regulatory Commission Washington, D.C. 20555 Lead NRC Engineer:
Y. S. Huang June 12, 1981 This report was prepared as an account cf work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, or any of their employees, makes any warranty, expressed or impiled, or assumes any legal liability or responsibility for any third party's use, or the results of such use, of any information, apparatus, product or process
~
disclosed in this report, or represents that its use by such third party would not infringe privately owned rights.
N'
[
dgr
_ Franklin Research Center A DM: ion of The Franklin Institute The Benpman Frarmhn Parkway. Phsta Pt 191c3 (215) 448-1000
TER-C5257-13 l
CONTENTS l
l Seg.t ion Title Page 1
BACKGRCUND 1
2 EVALUATICN CRITERIA 2
3 TECHNICAL EVALUATION 3
' Exemption from Airlock Testing Requirements 3.1 3
3.1.1 Extrapolation of Reduced Pressure Leakage Test Results.
6 3.2 Bydraulic Testing of Feedwater Check Valves 6
4 CCNCLUSICNS.
9 5
REFERENCES 10 APPENDIX A - Conversion of Reduced Pressurd' Air Leakage Measurements to Equivalent Full Pressure Air Leakage s
t e
0 0"
e
. ~. ~ -. -.
TER-CS257-13 1.
BACKGROUND On August 5, 1975 [1], the NRC requested Nebraska Public Power District
-(NPPD) to review the containment leakage testing program for Cooper Nuclear Station and to provide a plan for achieving full compliance with 10CFR50, Appendix J, including appropriate design modifications, changes to te'chnical specifications, and requests for exemption from the requirements pursuant to 10CFR50.12, where necessary.
NPPD's response dated September 10, 1975 [2), included five requests for exemption f rom *the requirements of Appendix J.
On September 16, 1977 [3), the NRC issued Amendment No. 38 to Facility Operating License No. DPR-46, autho-rizing three of the five exemption requests. At the same time, the NRC requested that NPPD provide additional information regarding the two remaining exe=ption requests. This additional information was forwarded by NPPD on October 30, 1978 [4).
The purpose of this report is to provide technical evaluations of all outstanding requests for exemption from the requirements of 10CFR50, Appendix J, for Cooper Nuclear Station. Consequently, technical evaluations of the two remaining exemption requests of Reference 2', as amplified by Reference 4, are included.
W W
e a
e l
Nnklin Research Center
~~-
TER-C5257-13 2.
EVALUATION CRITERIA l
I Code of Federal Regulations, Title.10, Part 50 (10CFR50), Appendix J, Containment Leakage Testing, contains the criteria used for the evaluation of the exemption requests. The criteria are either referenced or briefly stated, where necessary, to support the results of the evaluations. Furtherm' ore, in recognition of plant-specific conditions which could lead to requests for exemption not explicitly covered by the regulations, the NRC directed that the technical review constantly emphasize the basic intent of Appendix J, that potential containment atmospheric leakage paths be identified, monitored, and maintained below established limits.
O.
=
m.
4e E
m eb O
4
~ ~ _ _.
r TER-C5257-13 3.
TECHNICAL EVALUATION 3.1 EXEMPTION FRCM AIRLOCK TESTING REQUIREMENTS
.In References 2 and 4, NPPD requested an exemption from th'e requirements of Appendix J to test personnel airlocks at intervals of no longer than 1 year at 58 psig (Pa), at least every 6 months at 3 psig, and af ter each opening at 3 psig. NPPD stated that to conduct airlock tests at Pa, strongbacks must be used, which can only be applied in a shutdown condition. Further, NPPD stated that frequent airlock tests at Pa increase the.isk of permanent deformation of the airlock doors and that yearly tests at Pa are sufficient to show physical ' integrity.
Evaluation. Sections III.B.2 and III.D.2 of Appendix J require that containment airlocks be tested at peak calculated accident pressure (Ps, at 6-month intervals and after each opening in the interim between 6-month These requirements were imposed because airlocks represent potentially tests.
large leakage paths which are more subject to human error than other contain-ment penetrations. Type B penetrations (other than airlocks) require testing in accordance with Appendix J at intervals not to exceed 2 years.
Appendix J was published in 1973. A compilation of a'irlock events from Licensee Event Reports submitted since-1969 shows that airlock testing in accordance with ' Appendix J has been effective in prompt identification of airlock leakage, but that rigid adherence to the af ter-each-opening require-ment may not be necessary.
Since 1969, there have been approximately 70 reported airlock l'eakage tests in which measured leakage exceeded allowable limits. Of these events, i
25 percent were the result of leakage other than that resulting from improper i
seating of airlock door seals. These f ailures were generally caused by I
leakage past door operating mechanism handwheel packing, door operating cylinder shaf t seals, equalizer valves, or test lines. These penetrations resemble other Type B or C containment penetrations except that they may be operated more frequently. Since airlocks are tested at a pressure of Pa every 6 months, these penetrations are tested, at a minimum, four times more
~~
nklin Research Center A Deae.on et The Frerwen wuende
j
?
i TER-C5257-13 frequently than typical Type B or C penetrations. The 6-month test is, there-fore, considered to be both justified and adequate for the prompt identifica-tion of this leakage.
Improper seating of the airlock door seals, however, is not only the most frequent cause of airlock failures (the remaining 75 percent), but also repre-sents a potentially large leakage path. While testing at a pressure of Pa af ter each opening will identify seal leakage, it can also be identified by alternative methods, such as pressurizing between double-gasketed door seals (for airlocks designed with this type of se:1) or pressurizing the airlock to pressures' other than Pa.
Furthermore, experience gained in testing airlocks since the issuance of Appendix J indicates that the use of one of these alter-native methods may be preferable to the full-pressure test of the entire airlock.
Reactor plants designed prior to the issuance of Appendix J of ten do not have the capability to test airlocks at Pa without the installation of strong-backs or the performance of mechanical adjustments to the operating mechanisms of the inner doors. The reason for this is that the inner doors are designed j
to seat with accident pressure on the containment side of the door, and there-fore, the operating mechanisms were not designed to withstand accident pres-sure in the opposite direction. When the airlock is press'urized for a local airlock test (i.e., pressurized between the doors), pressure is exerted on the airlock side of the inner door, causing the door to unseat and preventing the conduct of a meaningful test. The strongback or mechanical adjustments prevent the unseating of the inner door, allowing the test to proceed. The installation of strongbacks or performance of mechanical adjustments is time consuming (of ten taking several hours), may result in additional radiation exposure to operating personnel, and may also cause degradation of the operat-ing mechanism of the inner door, with consequential loss of reliability of 'the airlock. In addition, when conditions require frequent openings over a short period of time, testing at Pa af ter each opening becomes both impractical (tests of ten take from 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> to several days) and accelerates the rate of exposure of personnel and the degradation of mechanical equipment,.
A Nbranklin Research Center A C>wuman af The Fmnman innsame
TER-C5257-13 For these reasons, the intent of Appendix J is satisified, and the undesirable effects of testing af ter each opening are reduced if a satisfac-tory ' test of the airleck door s tals is performed within 3 days of each opening or every 3 days during periods t f frequent openings, whenever containment integrity is required. The test of the airlock door seals may be per' formed by pressurizing the space between the double-gasketed seals (if so equipped) or by pressurizing the entire airlock to a pressure of less than Pa that does not require the installation of strongbacks or performance of other mechanical adj us tmen,ts.
If the reduced pressure aitic3k test is to be employed, the results of the leakage test must be conservatively extrapolated to equivalent Pa test results. An evaluation of NPPD's proposed method of extrapolation of-these test results from 3 to 58 psig is discussed in Section.3.1.1 of this report.
NPPD contends that the requirement to test the airlocks at Cooper Nuclear Station at Pa every 6 months is excessive, since the installation of the strongback necessary to perform the test requires shutting down the reactor to gain access to the containment. NPPD proposes to perform an airlock test at Pa once per year, at reduced pressure (3 psig) every.6 months, and at 3 psig af ter each opening. In view of the above discussion, this proposal is unac-ceptable because it does not meet the requirements of Appendix J nor does it satisfy the objective of the regulation.
Since NPPD submitted its request,,the NRC has revised Section III.D.2 of l
Appendix J, ef fective October 22, 1980. Essentially, the revised rule requires testing of airlocks as follows:
1.
Every 6 months at a pressure of Pa (and af ter periods when the airlock is opened and containment integrity is not required).
2.
Within 3 days of opening (or every 3 days during periods of frequent opening) when containment integrity is required, at a pressure of Pa or at a reduced pressure as stated in the Technical Specifications.
NPPD should establish an airlock testing program to conform to the requirements of the revised Section III.D.2.
No exemption from the require-ments of Appendix J is necessary.
N a c-o.,earch Center nktin Res
,.ma
1 1
2 i
TER-CS257-13 3.1.1 Extrapolation of Reduced Pressure Leakage Test Results In Reference 4, NPPD stated that the results of reduced pressure airlock tests (3 psig) and also reduced pressure bellows leakage tests (5 psig) are extrapolated to 58 psig using the criteria of ASME section XI, Winter 1976
~
Addendum, Article IWV-3000, " Test Procedure," paragraph IWV-3420, which states:
When leakage tests are made in such cases using pressures lower than function maximum pressure dif ferential, the observed leakage shall be adjusted to function maximum pressure differential value by calculation appropriate to the test media and the ratio between test and function pressure differential assuming leakage to be directly proportional to the pressure differential to the one-half power.
Evaluation. This correlation, nameAy that the Leakage results are pro-portional to the ratio of the test pressures tc~the one-half power, is appropriate when the characteristic of the leakage is essentially orifice-a l
like. However, when the flow characteristic of the leakage approaches capillary-like flow, this correlation becomes less conservative. As can be seen by applying equation A-3 (Appendix A to this report) for capillary-like 2
flow, the correlation proposed by NPPD is less conse'c'vative by a factor of
~
11.9 for the airlock test and 8.7 for the be'llows test. Although the actual leakage path characteristic is some unknown combination of orifice and capillary-like flow, the correlation pecposed by NPPD, particularly for the.
situation in which,the reduced pressure.is a small percentage of the full pressure test, is unacceptably non, conservative. It is recommended that equation A-3 be used to correlate leakage results as follows:
I ma.
(Pa + Pat) 2 - (Pat) 2 Et (Pt + Pat) 2 - (Pat) 2 (Note s is in terms of mass flow rate and Pat is atmospheric pressure.)
3.2 HYDRAULIC TESTING OF FEEDWATER CHICK VALVES 1
4 In Reference 2, NPPD requested an exemption frem the requirements of Appendix J to test the feedwater check valves with water as a test medium in i
i nklin Research Center A Comesen of he Frwwen rusame
TER-C5257-13 lieu of air or nitrogen.
In Reference 4, NPPD provided analyses to demon-
]
strate that the feedwater check valves remained water covered following a postulated accident and that feedwater system check val"s leakage following a LOCA will not exceed that established by 10CFR100.
Evaluation. Sections II.H.4 and III.C.2 of Appendix J require that containment isolation valves in main steam and feedwater systems of direct-cycle boiling water reactors be tested with air or nitrogen as a medium.
Section II.B of Appendix J defines containment isolation valves as those valves relied upon to perform a containment isolation function. It is clear that the feedwater check valves are relied upon to perform a containment isolation' funct' ion, and therefore, Appendix J requires that they be tested with air or nitrogen.
For operating reactors designed or constructed prior to the issuance of Appendix J, the substitution of a hydraulic test for the required pneumatic test may be an acceptable exemption from Appendix J where the hydraulic test is used to demonstrate that the valves will remain water covered throughcut the post-accident period. By using the hydraulic test to demonstrate this fact, the possibility of leakage of containment atmo'phere is eliminated.
s Th;refore, a determination of the pneumatic leakage rate is unnecessary since the valves are not being relied upon t_o isolate air leakage.
~
NPPD's submittal demonstrating that the valves will remain water covered
[4], however, f ails to demonstrate that they will be water covered throughout the post-accident period.
In fact, this analysis demonstrates that at the s
average leakage rates of these check valves experienced at Cooper Nuclear Station (8.3 f t /hr), the initial water inventory in a feedwater line at the start of an accident will be depleted after 421 minutes. At this time, unless reactor water level has been restored above the levei of the feedwater nozzles or the piping has been otherwise refilled, the check valves will be relied-upon to prevent the leakage of containment air. This situation may be miti-j gated by cooling water being injected by the HPCI or RCIC systems, which are initiated at the start of the accident. However, a single active failure in either of these systems could result in one of the feedwater lines being water i
dhrank!!n Research Center
% v w r.~
u.
TER-C5257-13 filled only by its initial water volume, which would be rapidly depleted by a combination of flashing to steam and the average leakage rate of the check valves.
Consequently, NPPD's proposal to test these valves with water in lieu of l
air or nitrogen i i not acceptable. The feedwater check valves should be l
pneumatically tested, with the leakage results added to the total pnetimatic leakage of the local leakage rate tests to determine acceptability in accor-dance with Section III.C.3 of Appendix J.
However, if liquid leakage limits are established which demonstrate that the valves will remain water covered for 30 days following a LOCA, hydraulic testing with acceptability based on these limits wotild be acceptable as an exemption to the pneumatic testing requirem~ents of Appendix J.
-~
u 4
~
I
-B-
~
, ac/ Franklin Research Center N
A DPanson ed The Fransen m
TER-CS257-13 4.
CONCLUSIONS d
Technical evaluations of requests for exemption f rom the requirements of Appendix J for Cooper Nuclear Station, submitted in Reference 4, were conducted. The conclusions are summarized below:
o NPPD's proposal to test containment airlocks annually at a pressure of Pa, every 6 months at a pressure of 3 psig, and af ter each opening at a pressure of 3 psig is not acceptable. Airlocks should be tested in accordance with the requirements of Section III.D.2 of Appendix J, revised October 1980.
o NPPD's proposed method for correlating reduced pressure leakage rates to full pressure leakage rates is not sufficiently conservative. A correlation assuming capillary-like flow characteristics should be
'used.
o NPPD's proposal to test feedwater check v.sives with water in lieu of air or nitrogen as a test medium is not acceptable because these valves may be exposed to containment atmosphere during the post-accident period. These valves must be tested in accordance with Appendix J unless they meet liquid leakage limits which demonstrate that they will remain water covered for 30 days following an accident.
u
~
.i r-N~ j Franklin Research Center A Ceamman af The Franean wasmae
TER-CS257-13 5.
REFERENCES
'l.
K. R. Goller (NRC)
I4tter to J. M. Pilant (NPPD)
August 5, 1975 2.
J. M. Pilant (NPPD)
Letter to K. R. Goller (NRC)
September 10, 1975 3.
- v. Stello, Jr. (NRC)
Letter to J. M. Pilant (NPPD)
September 16, 1977 4.
J. M. Pilan't (NPPD)
Letter to T. A. Ippolito (NRC)
Octoba-30, 1978 p
e WD f
5 s
sL P
j.
i l
t i
l i
I f
a i
l l
- h. Du rr.nxiin,w a.ren center 9 aese i
aw.
me
I APPENDIX A CONVERSION OF REDUCED PRESSURE AIR IIAKAGE MEASURDfENTS TO EQUIVALENT FULL PRESSURE AIR LEAKAGE JULY _17, 1980 I.
I l
Dr. G. P. Wachtall i
I i
Franklin Research Center l
F k
i t
l s
I i
l i
-..., _ _ _ _ _ _ _. _ _ ~,., _ _.... _ _ ~. - _____._
s l
APPENDIX A.
AIR TO AIR LEAXAGE CONVERSION In pneumatic leakage testing in which application of Pa psig is called for by Appendiz J, it is sometimes necessary to request an ex-e=ption that permits pneumatic testing at a lower pressure, Pt psig.
The leakage rate, Lt, measured under test conditions must then be, con-verted mathematically to the leakage rate, La, that would occur if the pressure were equal to Pa.
It is essential that the conversion be con-servative. That is, the calculated value of La must not be lower than the actual leakage rate at Pa would be.
On the other. hand, the conver-sion should. not be more conservative than necessary in the light of available data, because excessive conservatism could frequently result in 'the interpretation that a given leak exceeds its
=vN
allowable limit when in fact it would not exceed that limit if Pa were actually applied.
The meaning of the expression "if Pa were actually applied" should be carefully considered. The assumption is made that the geometry and dimensions of the leakage path would be the same with Pa applied as with Pt applied, or that any changes in geometry vould not increase the leakage rate. 'In the case of airlock doors in which Pt is applied in I
the reverse direction, opposite to the direction in which I'a would be
, applied under function conditions, the use of the reverse direction of application of pressure is ex,pected to tend to open the seal and increase the leakage rate. Under function conditions, in which pressure is 1
applied in the forward direction, the seal should be improved if, it changes at.all. The expression "if Pa were actually applied" in this case means "if Pa were actually applied in the forward (normal for function) direction." In the case of valves and other penetrations, it is essential that increasing the applied pressure from Pt to Pa not change the geometry se as to increase the leakage rate.
For example, increasing the pressure on a closed valve should tend to improve its sealing at the surfaces that provide the seal, and also in cny other A-1
potential leakage paths such as valve stem or packing that may have a connection to the applied pressure. Such other potential leakage paths are of course absent in valve designs in which the stem and packing have a connection only to the downstream side of the valve.
Reference 1, which is AS?!E Code,Section II, paragraph IW-3423 (e),
states the following rule for tests at less than function differencial pressure:
" Leakage tests involving pressure differentials lower than function pressure differentials are permitted in those types of valves in which service pressure will tend to diminish the overall leakage channel opening, as by pressing the disk into or onto the seat with greater force. Cate valves, check valves, and globe-type valves having function pressure dif ferential applied over the seat, are examples of valve applica-tions satisfying this requirement. When leakage tests are made in such cases using pressures icwer than func-tion maxi =um pressure differential, the observed leak-age shall be adjusted to function runrimu.a pressure differential value. This adjustment shall be made by calculation appropriate to the test media and the ratio between test and function precsure differential, assuming leakage to be directly proportional to the pressure differential to the one-half power."
In the discussion below, it is shown that if (a) the test mulium is air, (b) Pa is appreciable compared,to one atmosphere, and (c) the leakage path is such as to produce laminar viscous flow (i.e., capillary-
, like rather than orifice-like), the calculation appropriate to this test medium yields a'substantially higher calculated value of Pa than would' be obtained by assu=ing leakage to be directly proportional to the pres-sure differential to the one-half power.
For air flow through an orifice, assuning uniform flow velocity over the orifice area, the mass flow rate per unit orifice area is ov, where o is the density of air in the orifice and v is velocity in the orifice. Assuming that the discharge pressure is Pat = 1 atmosphere and the source pressure is Po, where Po and Pat are both absolute pressures, ov is given by 2
(pv)
-1 G
(A-1)
=
y-1 R,T Pat A-2
l vhere y = 1.4 is the specific heat ratio for air, g = 32.2 ft/sec is the acceleration of gravity, T is source (upstream, at Po) temperature
(*R), P is absolute pressure (psf), R, = 53.26 ft-lb/lb*F is the gas constant for air and G is given by y-1 Y-1
'Pe '
Y Y
x x
_1
( _2) g
,Pa c,
'p a
, Pat ~k Lo Pe Pe = Pat for subsonic flow Pe = 0.5283 Po for clicked flow Choked flow occurs when Y
Pat y+1 y-1
= 0.5283 Po 2
4 is propor).ional to ov// Po-Pat.
Values of 4 are listed in Tabie A-1.
6, the limiting value of 4 for small (Po-Pat), is o
/ (y-1) /y.
0.5345.
=
In Table A-1, inspection of 4/ %
shows the accuracy of the assu=ption that for an orifice-like leakage flow resistance, leakage mass flow rate is proportional'to pressure difference to the one-half power. For example, if Po
- 60 psig (Po-Pat = 60 in Table A-1),
4/ % = 1.210.
Extrapolation of mass flow rate measured with Pg =
15 psig to mass flow rate predicted for Pa = 60 psig will. underestimate the mass flow rate by the factor 0.968/1.210 = 0.80, or 20%.
The foregoing argument tacitly assumes that the orifice coefficient is = 1.0.
However, the same conclusion concerning extrapolation from low values of Pt to high values of Po can be drawn if the orifice coef-ficient is assumed to be constant, i.e.,
independent of Po.
Consequently, A-3
Table A-1.
/li for Various Values of Po - Pat for Orifice.
(Pat taken = 15 psia.)
psi d
0.01 0.5345 1.000 1
0.5332 0.998 5
0.5282 0.988 13.3 0.5185 0.970 13.4*
0.5184 0.970 15
- 0.5176 0.968 20
- 0.5230 0.978 25
- 0.5346 1.000 30
- 0.5490 1.027 35
- 0.5648 1.057 40
- 0.5811 1.087 45
- 0.5977 1.118 50
- 0.6143 1.149 I
55
- 0.6307 1.180 60
- 0.6470 1.21 0
- Choked flow for leakage paths that are known to be entirely orifice-like, the assump-cion that leakage mass flow rate is proportional to pressure difference to the one-half power gives a reasonably accurate correlation, underesti-60 psig. To mating the leakage mass flow rate by at most 20% for Pa
~
correct the underestimate, the factor (4/%),/(4/%)g has to be applied, where a and t mean Po = Pa and PC, respectively. References 2, 3, and 4 discuss the conversion formulas to be applied for various fluids (e.g, air and water) for various types of leakage path. For viscous flow of a gas, the mass flow rate from a source at absolute inlet pressure P to absolute y
2 2
outlet pressure P is proportional to (P -P2 ).
The proportionality 2
y factor is C/uT, where C is a function of geometry, T is absolute tempera-ture, and u is viscosity (which is a function only of temperature).
Assuming that test pressure Pt psig is applied at the same te=pera-ture as that at which function pressure Pa psig is applied, and assuming A-4 t
I
1 1
1 I
further that the downstream pressure is one at=osphere, Pac psia, then the ratio of the mass flow rates is ma (Pa + Pat) - (Pat) 3)
5e (Pt + Pat) - (Pat)
If the temperatures are not the same, the right side of Equation,(A-3) has to be m*ltiplied by u(Tt) Tt (A-4) p(Ta)* Ta Assuming t' hat Tt = Ta, Table A-2 shows the ratio Ea/Et for various values of Pa and Pt, along with values of (Pa psig/Pt psig)1/2 Pat h taken to be 15 paia in calculating ma/Et.
Table A-2.
5a/5t for Various Values of Pa and Pt.
(ha/ht) ha/ht (Pa/Pt)l/2 (Pa/Ptl l/2 Pt Pa=50 55 60 50 55 60 50 55 60 (psig)
(PIIO) 5 22.86 26.71 30.86 - 3.16 3.32 3.46 7.2 8.1 8.9 15 5.93 6.93 8.00 1.83 1.91 2.00 3.2 3.6 4.0 25 2.91 3.40 3.93 1.41 1.48 1.55 2.1 2.3 2.5 35 1.76 2.05 2.37 1.20 1.25 1.31 1.5 1.6 1.8 45 1.19 1.39 1.60 1.05 1.11 1.15 1.1 1.3 1.4 In all cases, the assumption that mass flow race is proportional to pressure differential to the one-half power is unconservative for purely viscous flow. For Pa = 60 psig and Pt = 5 psig, it is unconserva-tive by a factor of 8.9.
RECOMMENDED PROCEDURE Any one of the fo11cving procedures, A, B, or C should be adopted.
A-5
O A.
Test Program i
An extensive test program, covering several components of each type for which a correlation from Pt to Pa is sought, should be per-formed, in which sufficient experimental data showing the relation between Pt and leakage mass flow rate are obtained to permit a con-servative empirical correlation to be established. Care must be taken to ensure that experimental orifice-like leaks are not used to repre-sent actual, potentially capillary-like or viscous leaks.
B.
Conservative Theoretical Correlation Use Equation (A-3) as the correlation formula, including the factor (A,4) if necessary.
C.
Measure Leakage Characteristic For a given penetration, several values of Pt may be applied, so that an e=pirical correlation can be established. A statistical analysis of the data would be required to ensure at a 95% confidence level, that the predicted value of $a is not exceeded by the actual value of Ea.
REFERENCES l.
ASMI Code,Section II, paragraph k$'V-3423(a).
~
2.
Amesz, J., " Conversion of Leak Flow-Rates for Various Fluids and Different Pressure Conditions," 1966, EUR 2982.e, ORGEL Program,
~~
Ispra Establishment, Italy.
3.
- Maccary, R.R., DiNunno, J.J., Holt, A.E., and Arlotto, G. A., " Leakage Characteristics of Steel Containment vessels and the Analysis of Leakage Rate Determinations," May,1964, Division of Safety' Standards, AEC, TID-20583.
4.
Cottrell, Wm. B., and Savolainen, A.W., editors, "U.S. Reactor Con-tainment Technology," ORNL-NSIC-5, Aug.1965. Chapter 10 " Performance
' Tests," R.F. Griffin and G.H. Dyer. Sections 10.4.5 and 10.4.6 adapted from Reference 3.
~
s e
A-6
_ _ _, _ _