ML20128D140
| ML20128D140 | |
| Person / Time | |
|---|---|
| Site: | 05200001 |
| Issue date: | 12/02/1992 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20128D125 | List: |
| References | |
| NUDOCS 9212070156 | |
| Download: ML20128D140 (6) | |
Text
_
SAFETY EVALUATION OF LOW-PRESSURE PIP!WG SYSTIN DESIGN FOR THE INTERSYSTEM LOCA CONDITION IN THE ABWR 1
INTRODUCTION in SECY-90-016 dated January 12, 1990, the NRC staff resolved the intersystem loss-of-coolant accident (ISLOCA issue for advanced light water reactor plants by requiring that low-pre)sure piping system that interf ace with the s
reactor coolant pressure boundary be designed to ulthstand full reactor coolant system (RCS) pressure to the extent practicable.
The General Electric Company (GE) provided a proposed implementation of the issue resolution for the ADWR in a letter from J. Fox (GE) to C. Foslusny (HRC) dated October 8, 1992.
i in this safety evaluation, the staff has evaluated the GE proposal for implementing the ISLOCA resolution for the ABWR.
Specifically, the staff has evaluated the minimum pressure for which low-pressure systems should bo designed to ensure reasonable protection against burst failure should the low-pressure system be subjected to full RCS pressure.
In establishing the minimum design pressure, the following goals were used as the basis for selection:
(1)
The likelihood of rupture (burst) of the pressure boundary is low';
(2)
The.ikelihood of intolerable leakage of flange joints or valvo bonnets is reasonably low although some leakage might occur; (3)
Some piping components might undergo gross yielding and permanent deformation.
2 LOW-PRESSURE PIPING DESIGN To achieve the above objectives, the staff evaluated, first, on a qualitative basis, several possible ratios of the low-pressure system design pressure (P )
o to the RCS normal operating pressure (P ) to establish the margins on burst and ytold of the piping. TheresultsoIthestaff'sevaluationaredepicted in Tabie 1 for typical carbon steel (A106 Grade B) and sininless steel (SA312 Type 304) material and are discussed below for three ratios of the design pressure to the reactor vessel 3ressure (P /P ).
A margin of 1.0 or less d
y represents the condition where aurst or yielding is likely to occur.
1he higher the margin, the less likely it is for twrst or yielding to occur.
The low-pressure piping systems are assumed to be designed to the rules of the ASME Boiler and Pressure Vessel Code, Section 111, Subarticle NC/ND-3600 for Classes 2 and 3 piping systems.
- The NRC staff used a goal of approximately 10 percent failure probability (or conversely, a goal %f 90 percent survival probability) fo-rupture.
9212G '0156 921202 PDR ADOCK 05200001
^
Enclosure
Table 1 Harpins for Straight Pipe Material Temp P,,/ P, 3
5 S
S Hargins on i
'F ksi ks"i ks"1 ks"i Burst Yield SA106 Grade B 100 1/2 15 30 60 35.0 2.00 1.34 500
?/2 15 30 60 28.3 2.00 1.08 100 1/3 15 45 60 35.0 1.33 0.89 j
500 1/3 15 45 60 28.3 1.33 0.72 100 1/4 15 60 60 35.0 1,00 0.67 j
500 1/4 15 60 60 28.3 1.00 0.54 Haterial Temp P,/ P, S
S S
S Hargins on
't ksi hs"i k s"i ks"i Burst Yield SA312 Type 304 100 1/2 10.8 07.5 75.0 30.0 1.70 0.92 500 1/2 15.0 31.8 63.5 19.4 1.70 0.70 100 1/1 18.8 56.3 75.0 30.0 1.13 0.61 500 1/3 15.9 47.7 63.5 19.4 1.13 0.47 100 1/4 18.8 75.0 75.0 30.0 0.85 0.46 500 1/4 15.9 63.6 63.5 19.4 0.85 0.35 5
allowable stress per ASME Code, Section Ill for Class 2 piping S, a hoop stress at P - Py S/(P,/P,)
S, - ultimate tensile strength; from Section Ill, Table 1-3.1 and 1-3.2 5, = yielo strength; from Section III, Table 1-2.1 and 1-2.2 where F = 1.00 for SA106 Grade B, x (P /P,)/S Hargin on Burst Pressure - F'x S o
F - 0.85 for SA312 Type 304 Margin on Yield Pressure - 1.15 x S, x (P /P )/S d y Pinina Intearity at Pd/Pv - 1/2
-(ASME Code Service level D)
Whe. P /P is equal _ to one-half, the margins on burst and yield are equivaienttoapproximatelythoseoftheASMEBoilerandPressureVessel y
Code, Section !!! Service Level D condition.
For carbon steel pipe, this ratio will provide a margin of 2.0 on burst and 1.08 on yield for a pipe at 500 'F.
For stainless steel-piping, a ratio of one-half will 2
7
\\
provide a sufficient margin on burst (1.7).
However, a small amount of yielding 4 likely to occur with a margin of 0.70 at 500 'F.
No leakage of the pressure boundary is likely to occur at P,/P equal to one-half.
y As a result, a ratio of one-half will ensure the pressure integrity of the low-pressure piping systew with ample margin, figina Intandly_ at Pd/PV - 1/3 When the ratio, P,/P,, is reduced to one-third, the margins for carbon steel piping are lowered to 1.33 and 0.72 for burst anci yield at 500 'F, respectively.
For stainless steel piping, the margins are 1.13 and 0.47 for burst and yield at 500 'F, respectively.
At tiese margins, it is expected that burst failure will not occur in either carbon steel or stainless steel piping. However, significant amount of yiciding might occur in stainless steel piping at all temperatures and in carbon steel piping at 500 'F.
Wnere the carbon steel 31 ping is at a lower temperature, scme yielding might occur altlough to a lesser extent. The consequence of significant pipe yielding (without bursting) is that gross, permanent distortion might occur in the piping components thereby resulting in some leakage through flanges, or valve bonnets. However, it is not expected that such leakage would be uncontrollable or intolerable.
In summary, a ratio of one-third will ensure the pressure boundary of the low-pressure piping although a significant amount of pipe yielding and some leakage through flanges and valve bonnets is likely to occur.
Pipina Intendly at Pd/PV __11_4 At P,/P equal to one-fourth, the pressure integrity of carbon steel pi>ingbecomesquestionable,andforstainlesssteelpiping, itis licely that burst failure will occur.
Prior to bursting, the piping system would undergo gross plastic deformation, experience a significant amount of leakage at flanges, valve bonnets, and pump seals, and possibly lose some pipe supports due to the radial expansion of the pipe.
equal to one-fourth, the ability of the low-pressure Therefore, at P,/P,ithstand full RCS pressure is questionable for carbon piping system to w steel piping and unitkely for stainless steel piping systems.
The staff further evaluated, on a-quantitative basis, the survival probabilities of the low-aressure piping at various design pressures using the methodology described in NUREG/CR-5503, " Pressure-Dependent Fragilities for Piping Components," dated October 1990.
Calculations were performed by Idaho National Engineering Laboratory under contract with the NRC's Office of Nuclear Regulatory Research.
The INEL calculations led to results similar to the qualitative conclusions discussed above. A temperature of 350 *F was used in the calculations of the following survival probabilities. Using a temperature of 500 'F, the survival 3
e 9
D' h
probabilities decreases about 2-5 percent for the different materials and design pressures.
For carbon steel piping (SA-105 Grade B material), the survival probability was approximately 94 percent at the design pressure proposed by GE of 350 psig (roughly equivalent to P /P - 1/3). When the design pressure is increased to 400 psig (or approxima"tely P,/P, - 0.4), the survival probability increases to 99 percent.
for stainless steel piping (Type 304 material), the survival probability at the GE-proposed design pressure of 350 ptig (or approximately P /P,ly P /P, =
- 1/3) was d
about 70 percent. At a design pressure of 400 psig (or approximate d
1 0.4), the survival probability increases to about 87 percent.
Using the results described abovn, the staff finds that the GE-proposed design pressure of 350 psig is sufficient for carbon steel piping, but would result in a survival probability of approximately 70 percent for stainless steel piping. At a design pressure of 410 psig, the survival probability for riainless steel piping increases to approximately 89 percent. Therefore, the staff finds that a ratio of P /P 0.4 nearly achieves the staff's goal of a 90 percent survival probability,under ISLOCA conditions and provides a sound basis for establishing the design pressure of low-pressure systems interfacing with the RCS pressure boundary.
It should be noted, however, that the survival probabilities are based on the mininum wall thickness as calculated using Equation (3) in the ASME Boiler and Pressure Vessel-Code, Section !!!, Subsubarticle NC/ND-3640. The wall thickness thus calculated does not account for manufacturing tolerances or the use of the next heavier commercial wall thickness available which would increase the piping wall thickness and substantially increase the survival probability as well, furthermore, when standard weight piping wall thicknesses are.used (i.e., schedule 40 pipe up to and including 10-inch nominal pipe size), the survival probabilities at a design pressure 400 psig are above 99 percent for both carbon and stainless steel piping.
Currently operating boiling water reactor plants typically use, as a minimum, standard weight piping for the low-pressure pi)ing systems that-interface with the reactor coolant )ressure boundary. Tierefore, for the ABWR, the staff-concludes that tie use of standard weight piping will provide a bounding 1
design for the ISLOCA condition.
3 VALVES IN LOW-PRESSURE SYSTEMS for the valves in the low-pressure pi)ing systems (excluding the pressure isolation valves),.the selection of tie valve class rating is a primary factor for designing-against full RCS pressure.
for example, ANSI B16.34 valves are supposed to be shop-tested to 1.5 times their 100 'f rated pressure.
This would mean for a Class 300 A216 WCB (cast carbon steel ) valve, the test pressure is 1.5 x 740 - 1110 psig, for Class 150, the valve test-pressure is-1.5 x 285 - 427.5 psig.
Clearly, the Class 300 valve that is tested to a pressure-of 1110 psig would-be expected to withstand an RCS normal operating pressure of 1040 )sia (or 1025 psig). However, it should not be assumed that the valve in t1e low-pressure system would be able to operate with this full RCS pressure across 4
e w-ye
-r-
-'m a--
e
-I d-er-m.r y*w 9e
'W r'$- - + ' +. -
hm-a w--
e y
e e
the disk.
Therefore, the staff finds that a Class 300 valve is adequate for ensuring the pressure of the low-pressure piping system under full RCS pressure (i.e., 1025 psig) but no credit should be taken to consider these valves operab'e under suc.i conditions without further justification.
4 OTHER COMPONENTS IN THE LOW-PRESSURE SYSTEMS for other components in the low-pressure systems, such as pumps, tanks, heat exchangers, flanges, and instrument lines, the staff finds that establishing an appropriate safety factor involves several complicating factors related to the individual component design.
These factors include requirements for shop hydrotests, the method to determine the pressure class rating of the component, the s)ecific material used for bolting and the bolt tension applied, or whetier the component is qualified by test or analysis.
The staff is of the opinion that designing such components to the GE-proposed design pressure of 350 psig (or approximately one-titrd of the normal operating RCS pressure) might not provide sufficient margin to burst failure because of the uncertainties in their design as discussed above. Therefore, the staff's position is that the remaining components in the low-pressure systems should also be designed to a design pressure of 0.4 times the normal operating RCS )ressure (i.e., 410 psig). The staff believes that the margins to burst for tiese remaining components are at least equivalent to that of the piping at its minimum wall thickness since these components typically have wall thicknesses greater than that of the pipe minimum wall thickness.
5)ecific components that cannot meet these guidelines should be identified to tie staff for review on a case-by-case basis by the COL applicant as a part of its COL applicction.
5 CONCLUSIONS On the basis of the above discussions, the staff finds for the ABWR that the design pressure for the low-pressure piping systems that interface with the RCS pressure boundary should be equal to 0.4 times the normal operating RCS r9 m re of 1025 psig (i.e., 410 psig), and the minimum wall thickness of the Ws resse:11 piping should be no less than that of a standard weight pipe.
% '5 sign is to be in accordance with the ASME Boiler and Pressure Vessel Section 111, Subarticle NC/ND-3600. Furthermore, the staff will e
3; anue to require periodic surveillance and leak rate testing of the M essure isolation valves per Technical Specification requirements as a part of the inservice inspection program.
As stated in SECY-90-016, for those low-pressure systems for which designing to withstand full reactor pressure, as discussed above, is not practical, the desion 6,. 'd provide (1) the capability for leak testing of the pressure l
iscl e : '
ives, [2) valve position indication that is available in the i
cont 6f when tsolation valve operators are deenergized, and (3) high-pressure 4 arms to warn control room o>erators when rising RCS pressure approacus the design presure of attacied low-pressure systems and both isolation valves are closed.
5
=
h
. ~.. -. -. - -.. -
A Using these minimum design guidelines, the staff concludes that:
(1) the likelihood of the low-pressure piping rupturing under full RCS pressure is low, (2) the likelihood of intolerable leakage is low under ISLOCA conditions although some leakage might occur at flanges and valve bonnets, and (3) some piping components might undergo gross yielding and permanent deformation under ISLOCA conditions.
Because of the likelihood of piping system distortion, the staff finds that A visual inspection of the low-pressure piping system after it has been subjected to the full RCS pressure is necessary to identify those components that might require replacement, repair, or further detailed analysis prior to restart, The minimum design pressure and pipe wall thickness for the Icw-pressure piping systems should be identified as a Tier I commitment for the ABWR. An ITAA. should be developed to ensure that the design specifications for the low-pressure piping system adequately define the pressure design requirements and minimum pipe wall thicknesses for protection against the ISLOCA condition.
Contingent upon GE revising its SSAR and submitting Tier 1 commitments and ITAAC as identified above, the staff concludes that there is reasonable.
assurance that the low-pressure piping systems interfacing with the reactor coolant pressure boundary are structurally capable of withstanding the consequences of an intersystem loss-of-coolant accident.
c
.6 4