ML20129F286
| ML20129F286 | |
| Person / Time | |
|---|---|
| Site: | 05000447 |
| Issue date: | 04/20/1984 |
| From: | GENERAL ELECTRIC CO. |
| To: | |
| Shared Package | |
| ML20127A304 | List: |
| References | |
| FOIA-84-175, FOIA-84-A-66 NUDOCS 8506060729 | |
| Download: ML20129F286 (89) | |
Text
__
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.150 Does the GESSAR 11 seismic hazard curve represent the best estimate (p. 15)? Describe the. procedures used t6 generate the GESSAR II hazard curve and the other curves shown in Figure 2.2 of your report.
Response
74 l 0 h : #.
~.a, 3:*;
33,,,
3t:;;f h s%5fhfi 4 0b'C
^
~~
6 N
KWH:rf/G01031*-1 2/28/84^
r GENERAL ELECTRIC COMPANY PROPKIETARY INFORMATION Question 720.151 (a)
In Section 2.3.3 it was indicated that the GESSAR-II seismic hazard curve was constructed as an " arbitrary bounding curve" of four best estimate curves from plant specific studies.
How does your " arbitrary boundary curve" take into account the possibility that hazard curves calculated for other plants in the eastern U.S. may exceed the GESSAR-II seismic hazard curve.
(b) Provide the supporting evidence and calculations used to show that the GESSAR-II seismic hazard curve as stated in your report (p. 16) would bound "more than 80% of the potential GESSAR-II sites."
Response
(a)
It is expected that the Limerick, Zion, Oyster Creek and Indian Point seismic hazard curves are representative of eastern U.S.
sites and provide a reasonable representation of potential GESSAR sites.
The curve does not account for hazard curves for plants in the eastern U.S. that may exceed the GESSAR-II seismic hazard curve.
However, as noted in the discussion of the intended implementation of the GESSAR Seismic Analysis, a site-specific hazard curve will be required for comparison to the GESSAR II curve to determine the applicability of the generic analysis.
(b)'
~'
A C.8 KWH: rf/G01031*-2 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.152 Provide the supporting bases for extrapolating the existing seismic hazard curves to obtain as stated in your report on page 13 "a
realistic median-centered upper-bound curve" beyond 0.8g (p. 15).
Response
In terms of the GESSAR II Seismic Analysis, the effective peak ground accelerations (EPGA) from 0.8g to 0.95g are approximately equal to 1.0g to 1.2g in peak ground acceleration.
In this acceleration range, strong motion data are in general very minimal.
As a result, any estimates on exceedance probabilities and the upper bound cut-off on ground acceleration in this range are highly conjectural.
However, if the seismic hazard functions for Limerick, Zion, Oyster Creek and Indian Point are indicative of the state-of-the-art knowledge in,this area, the j_'
upper bound cut-off selected for the GESSAR 11 curve at.'
.j, appears reasonable to GE.
41-It
3-d
$fd KWH: rf/G01031*-3 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.153 Please elaborate on what is meant on page 12 of your report, "the seismic hazard capability, represented here as a seismic hazard curve, is unique."
Response
The " capacity" approach leading to the sentence quoted from page 12 can be illustrated by the comparison to the analysis of a structural system for a monotonically increasing loading wherein the loading is applied until the system response has exceeded a given set of failure criteria.
The limiting load, in this example, defines the system capacity and is unique for the system for the load considered.
If, in the example, the "monotonically increasing loading" were replaced by its equivalent in the seismic hazard curve, the seismic hazard capability would be unique for the system.
~
E #
KWH: rf/G01031*-4 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.154 It has been indicated that the individual p values were not estimated; however, the combined p for the overall median factor of safety was established on the basis of engineering judgment.
Therefore, to better understand this process and its adequacy, provide a detailed analysis for one of the p values by examining individual components with a comprehensive discussion of the site-specific effects.
Response
1 r:/,
.)a ris,
4 a r.-s KWH:rf/G01031*-5 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.155 Describe the procedure by which the specific structural components were selected to evaluate the fragility of the seismic Category I structures (for example, how the flexural capacity of the auxiliary building was derived).
Response
The procedures for the evaluation of the fragility of the seismic Category I structures are described in GESSAR II Section 3.2.3.
For example, the procedure for the evaluation of drywell wall and shield building are given in pp. 36-39.
See also response to question 720.161.
,oJ (* I KWH:rf/G01031*-7 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.156 Discuss the possible effects of the interaction between the buildings.
Also, include interaction between seismic Category I and non-Category I structures.
Response
4a b:I l
l KWH:rf/G01031*-8 2/28/84
u kALLRL919Mifh%vLLh/
PROPRIETARY INFORMATION Question 720.157 Discuss the plant-specific featurer, and structures which may need to be evaluated as far as the structural fragility investigations are concerned.
Provide the interface requirements for plant-specific application (for example, interface requirements for the steel vessel containment which will be designed by the future applicant referring to GESSAR).
Response
Interface requirements for plant-specific application in support of PRA.
requirements and conclusions will be specified in appropriate interface documentation between GE and future applications.
For example, the relatively simple requirement that the suppression pool region not be the first locus of containment failure is readily communicated by GE to the future applicant.
See also the response to question 720.174.
il, Jo olal (Y$
KWH:rf/G01031*-9 2/28/84
n.
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.158 The relationship given in the Section 3.2.3.1.3 between the ductility of the dampling.
Similarly, it is not clear whether the interaction between duration, damping and the inelastic energy absorption has been considered.
Provide discussion to indicate how individual factors of safety and the logarithm of the standard deviation accounts for interrelationship between these parameters.
Response
)L**
' if ;'h [i
!\\
i.
I e
Y
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.159 The general fragility model is based on the simple assumption of the multiplicative model which implies independence between individual factors of safety.
As this method has been unverified, discuss how did you assure that the overall factor of safety resulting from that approach and associated standard logarithm deviation are reasonabic.
Response
As the state-of-the-art approach, the multiplicative model has been used by a number of PRA analyses, including Limerick, Zion, Oyster Creek and Indian Point.
The GESSAR II results have been compared with results from these studies and have been found to be consistent.
It is believed that this approach adequately employs current methodology.
{
.,r s.
KWH: rf/G01031*-15 2/28/84
.~. _ _ _ _ _ _ _ _ _ _
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.160 It should be noted that duration effects are dependent upon the magnitude of the earthquake, how did you consider this relationship.
Response
3-d.,
l l
t t
i-l 6
l I
l t
I KWH: rf/G01031*-16 2/28/84 L
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.161 Provide detailed calculations supporting fragility values presented in Tables 3-2 through 3-19.
Response
V(1
-g t
4 6
t
.'\\
KWH:rf/G01031*-17 2/28/84 J
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.162 The applicant should discuss the considerations given to the pertinent geotechnical parameters in developing factors of safety associated with fragility analyses.
Specifically, the applicant should address thi contributions of the geotechnical parameters to the degree of uncertainty associated with each factor of safety.
Response
r h
i' 5
I l
g_
ln b b O.L \\
tj fr L',
l l
KWH:rf/G01031*-20 2/28/84
GENERAL ELECTRIC COMPANY k
PROPRIETARY INFORMATION Question 720.163 The applicant should discuss the uncertainties associated with the following potential site-specific external-events-related failures on the core-melt probabilities:
(i) liquefaction of site soils, (ii) differential settlements affecting structures, systems and component failures, (iii) consideration of dynamic earth pressures beyond those considered in the design, and (iv) slope failures affecting performance of ultimate heat sink functions.
The seismic loading to be considered in the analyses of items mentioned above should reflect loading beyond the seismic input considered in the GES,SAR II design (up to 0.95g).
The app should also justify the applicability of these analyses on.licant a generic basis to all potential sites meeting GESSAR 11 design criteria.
Response
\\T, sh '
l t
t
\\
)
KWH: rf/G01031*-21 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.164 It is not clear that the applicant has considered the possible failure of non-safety related structures or equipment which could impact on safety-related items.
The applicant is requested to provide a description of the procedure of evaluating the effects of failures of non-safety related structures or equipment on safety-related structures and safety-related piping systems and components and their supports, which could lead to a possible core meltdown.
f
Response
The design basis for safety related items is to ensure that the failure of non-safety items cannot impair the function of the safety related items.
For example, mechanical and divisional separation help to preclude the total loss of any required safety function.
p'f g/
a:.'
l KWH: rf/G01031*-23 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.165 It is not clear that the applicant has considered the possible failure of piping systems and components due to differential movement or tilting of structures.. In addition, it is not clear how the applicant has considered the failure modes of buried piping which may be influenced by geotechnical parameters, e.g., soil amplification.
The applicant is requested to confirm that these effects which could
. eventually lead to a core meltdown, have been considered and included in its evaluation.
Alternatively, provide the basis for not considering them.
Response
(i)
\\
q,
(ii) There is no seismic Category I pipes buried in soll in the standard GESSAR design and therefore the effects stated will not be considered.
N, KWH: rf/G01031*-24 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION 1
Question 720.167 There are several essential piping systems which are required to deliver rated flow and be designed to retain dimensional stability when stressed to the allowable limits associated with the emergency and faulted conditions, e.g.,
the functional capability of the l
piping is required to be demonstrated.
It is not clear thst the applicant has considered the functional capability in its evaluation of piping failure modes.
The applicant is requested to assess its evaluation given the above concern.
Response
/i
?
t I
k L
A response to this question was provided on the Limerick docket (reference J. S. Kemper (PECO) letter to A. Schwencer (NRC), subject:
Limerick Generating Station, Units 1 and 2 Responses to NRC Questions on the Severe Accident Risk Assessment (SARA}, August 29, 1983 Docket Nos. 50-352 and 50-353.
That response is also applicable to GESSAR and is reproduced in the following paragraphs.
A lower bound on functional failure can be derived utilising the available data base on collapse of piping. subjected to ' static loadirg and considering no beneficial ef fect from inelastic energy absorptions.
Stainless steel was shown in the Limerick SARA to be the governing material for piping collapse and will be used as an example.
i Reference 1 summarises available test data on piping components subjected to static loading.
The data base for straight pipe sections consists of several sources, References 2-7.
A review of the data base performed by SNA engineers during the seism,1c Safety Margin Research Program, Reference 8, indicates that the of factive shape factor at collapse for straight pipe ranges from i
1.4 to greater than 2.8 where the shape factor is defined as the calculated elastic stress at collapse divided by the stress at j
first yelid of the outer fiber.
Collapse is defined as in Appen-din II of the ASME code as the point where deformation is two times the deviation from linearity taken from the test load-i deformation diagram.
The shape factor is dependent upon the D/t ratio, material and strain hardening esponent.
Note that at the defined value of plastic collapse that ovaling was not of suffi-cient magnitude to impose a significant flow restriction.
For most of the data, the of fective shape factor at collapse ranged from 1.4 to 1.67 for D/t less than 50, with 1.5 being a best estimate value of the data base.
The effective shape factors
[
were derived relative to the actual material yield.
Average yield is about 1.25 times code-specified yield, thus the best i
KWH: rf/G01031*-26 i
2/28/84
estimate shape factor at. collapse relative to code-specified yield is 1.45 x 1.5, or 1.875 times code yield.
Por stainless steels, the allowable stress, S, is % of the ulti-h mate strength, or from 0.625 to 0.9 times the yield strength i
depending upon the design temperature.
Jn Limerick SARA, the average allowable stress S, for stainless steel pipe was con-h sidered to be about 0.72 of the code specified yield strength.
l The ratio of the collapse load to the code Level B Service
- allowable load of 1.2 Sh is:
Pe 1.87
= 2.16 Ip " 1.2(0.72)
As discussed in pages 5-24 & 5-25 of Appendix B of SARA, assump-tions were made on ranges of static stress due to pressure plus deadweight and seismic stress from the OBE.
(Note that the OBE governs piping design in the absence of hydrodynamic loads).
The range for the static load was assumed to be 0.3 to 0.6 times the j
allowable stress of 1.2 Sh and the range of OBE loading was assumed to be 0.2 to 0.65 times the allowable.
Using properties of the lognormal distribution, the median values are 0.42 and O.36 times the design allowable, respectively.
The median strength factors (fs) can then be calculated as:
f,
C/Po - P /P N
D
, 2.16 - 0.42
,4,g3 g
P
/P 0.36 OSE D
where PC is the median collapse load or stress, PN is the normal operating load or stress, Po is the code design allowable load or stress, and POBE is the seismic load or stress associated with the OBE.
The factor relative to the SSR is then 4.83/2 or about 2.4.
Thus, a lower bound'on collapse, considering 32 beneficial effects of inelastic energy abosorption is 2.4 times the SSE.
Accounting for conservatisas in the piping response analysis and structural analysis used to generate instructure response spectra, typical calculated equipment and structural response factors in the reactor enclosure were 2.68 and 1.72, respectively.
The median ground acceleration capacity can then
, be calculated to bes 2.4 x 2.68 x 1.72 x 0.15 = 1.66 g where 0.15 is the specified SSE peak ground acceleration.
- As defined in ASME Boiler & Pressure Vessel Code,Section III.
t
REFERENCES 1.
Rodabough, E. C. and.S. E. Moore, " Evaluation of the Plastic Characteristics of Piping Products in Relation to ASME Code Criteria," NUREG/CR-0261, ORNL/SUB-2913/8, Battelle Columbus Laboratories, July, 1978.
2.
Gerber, T.
L., " Plastic Deformation of Piping Due to Pipe-Whip Loading," ASME Paper No. 74-NE-1.
3.
Pranzen, W.E., and if. P. Stokey, "The Elastic-Plastic Be-havior of Stainless Steel Tubing Subjected to Bending, Pres-sure and Torsion," Second International Conference on Pres-sure vessel Technology, San Antonio, Texas,1973; Published by ASME, New York.
4.
Del Puglia, A. And G. Nerli, " Experimental Research on Elasto-Plastic Behavior and Collapse Load of Statically In-determinate Space Tabular Beams," 2nd International Con-ference on Structural Mechanics in Reactor Technology," Ber-lin, Germany,1973, Vol. 2, Part F.
5.
Sherman, D. R. and A. M. Glass, " Ultimate Bending Capacity of Circulac Tubes," Proc. Offshore Technology Conference, Dallas, Texas, 1974, OTC Paper No. 2119.
6.
Jirsa, J. O., F.
H. Lee, and J. C. Wilhoit, "Ovaling of Pipelines Under Pure Bending," Proc. Offshore Technology Conference, Dallas, Texas, 1972, OTC Paper No. 1569.
7.
Sorenson, J. E., R. E. Mesloh, E. Rybicki, A. T. Hopper, and T. J. Atterbury, " Buckling Strength of Of fshore Pipelines",
Battelle-Columbus Labs.
Report to the Offsbore Pipeline 3
Group, July 13, 1970.
i 8.
Kennedy, R. P., R. D. Campbell, G. Hardy, B. Banon, "Sub-systen Fragility, Seismic Safety Margin Research Program (Phase 1).
NUREG/CR-2405, DCRL-15407, February, 1982.
GENERAL ~ ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.168 There are several safety systems connected to the reactor coolant pressure boundary that have design pressure below the reactor coolant system (RCS) pressure.
There are also some systems which are rated at full reactor pressrue on the discharge side of pumps but have pump suction below RCS pressure.
In order to protect these systems from RCS pressure, two or more isolation valves are placed in series to form the interface between the high pressure RCS and the low pressure systems.
Feilure of these isolation valves (from either excessive leakage or rupture) will allow the high pressure reactor coo.lant to communicate with the low pressure piping outside of containment.
Rupture of the low-pressure system would result in loss of coolant outside of containment and possible core meltdown.
It is not clear how the applicant has considered the above concern in its evaluation of valve failure modes.
The applicant is requested to assess its evaluation given the above concern.
Response
,C 4
.'r 2n5r:
1 i
I I
4 1
\\
KWH: rf/G01031*-27 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.169 W
High fragility values of structures and components
~ 3, are indicated.
'These values should initiate an overall failure mode such as foundation failure, a ground stability problem etc.
There should be a cap on structures and components fragility level that is consistent with an overall failure mode.
Response
The use of high fragility values is not intended to suggest that these values will actually be realized in the event of an extraordinary earthquake.
Clearly there will be some other structure or component which will limit overall behavior.
These have been identified in the analyses: the relative values provide an index of resistance and potential failure mode.
Potential foundation failure or ground stability problems have not been included because they depend on site specific information for analysis.
KWH:rf/G01031*-29 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.170 Fragility estimates for components whose seismic failure could lead to a core melt should be examined in some detail.
This could be accomplished by looking at similar components from an actual p'lant/
reference pic9t.
In addition, sensitivity of the seismic contribution to core melt due to a shift in the capacities of those components that are very near the capacities of the critical components should provide greater insights into the robustness of GESSAR design, for example shroud support, CRD Guide Tube, Hydraulic Control Unit, Electrical Power etc.
A detailed look at fragility estimates should highlight the local response values in terms of accelerations and corresponding frequency range of critical / controlling components such as relays, contactors, breakers, control logic elements etc.
Response
See the response to Question 720.161.
w
-IY d
KWH: rf/G01031*-30 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION l
Question 720.171 In order to understand the influence of earthquakes on the course and consequence of a severe accident it should be clear in our minds whether or not the earthquake happens before or after the accident.
l If the accident occurs after an earthquake, we should consider the effect of the earthquake on the consequences of the accident, for example the failure of the stack may cause ground level release of radioactivity, buckling of steel containment can affect its ability to prevent leakage through large penetrations.
Should the accident occur prior to the' earthquake, fragility levels for certain parts of containment can be substantially affected because of the accident loading.
Provide an assessment of the impact of aftershocks following severe earthquake on core damage frequency and risk.
Response
The seismic event analysis considered seismic-initiate core damage events.
Therefore component and structural failures were assessed based on the first shock and the., core damage progression was calculated assuming these failures. f" lc 36 ab N
KWH: rf/G01031*-31 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.172 Assumptions regarding leak path and behavior of locks, seals and 0-rings are very much influenced by time at temperature.
- Also, certain types of details can be more prone to deterioration due to temperature.
Inflatable seals are generally exposed to local temperature effects such as in personal airlocks for some plants.
Detailed discussion regarding leak path and orientation and design of critical penetrations is necessary to assess the protection available in the GESSAR design against severe accident.
Response
-[,
~4c
/
Therefore the responses to similar questions (e.g. 720.85)
'would be applicable to this question.
7 4
i KWH: rf/G01031*-32 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.173 j
Identify any new equipment and modification to specific plant features that if incorporated have a high potential for risk reduction relating to seismic events, fire, and internal floods.
Response
The high level of protection afforded by the GESSAR II design for seismic, fire and flood initiated events is evidenced by the low contribution to core damage frequency from such initiations.
Therefore, no additions or modifications to the GESSAR II design are needed to lower the already insignificant risk and none have been identified.
ts (t.:r p jr KWH:rf/G01031*-33 2/28/84
G3d3Mn3LR3L5VtJUlenSMuk PROPRIETARY INFORMATION Question 720.174 Indicate what interface performance specifications are envisioned to compensate for the lack of design detail in the BOP features.
Response
The interface performance specifications are delineated in the GESSAR II Safety Evaluation Report (NUREG-0979) Section 1.10.
puo '
~
e KWH:rf/G01031*-34 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.175 Provide basis for not considering variations in the construction material of the containment basemat or variations in the ultimate strength of containment.
Response
Section 3.2.3.1.2 Strength Margin, F, of the Seismic Event Analysis does include a discussion of the effects of variations in the strength properties of the construction materials included in the subject analyses.
It is further noted that the Codes end Standards which control the construction material properties are all based on establishing minimum requirements.
The practical outcome of this practice is to effectively cut off the low end of any property distribution curve.
Actually most such curves based on materials used-in nuclear power plants will be very narrow based.
(4 f" 9
1 KWH:rf/G01031*-35 2/28/84
i GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.177 The ultimate capacity calculations by GE for the steel containment based on limit analysis using the ultimate strength of SA 516 grade 70 steel may be questionable.
Please provide justification for such an approach.
Response
Thehontainmentcapabilitycalculationsbasedontheultimate strength of SA 516 grade 79 steel wrere made to provide one index point _for..a fai. lure.probabi.1_ity. analysis.
b, '
'I L<
KWH: rf/G01031*-37 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.177 The ultimate capacity calculations by GE for the steel containment based on limit analysis using the ultimate strength of SA 516
, grade 70 steel may be questionable.
Please provide justification for such an approach.
Response
The containment capability calculations based on the ultimate strength of SA 516 grade 79 steel were made to provide one index point _for..a fai. lure probability.. analysis.
2' L<
KWH: rf/G01031*-37 2/28/Q4
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.178 The discussion on 'racture, especially in weldments is not presented f
in a manner that could be evaluated.
The contention that cracks will develop only when stresses are between yield and ultimate strengths should be justified.
Response
The materials which are used for the plate elements and the welding materials for steel containment structures are tough, ductile materials.
The procedural controls for welding include, among other things, appropriate preheating requirements; the objective of course, is to produce weldments which have material properties close to those of'the base plate materials.
In addition the structure is neither restrained nor acted upon by external straining devices which produce high local strains which might lead to local cracking.
It is our belief that our assumptions are reasonable, conservative, and well within demonstrated engineering practice.
,i l't F
I KWH:rf/G01031*-38 2/28/89
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.179 The use of Equation G.8.25 for calculation of buckling of the knuckle region geometry should be justified.
It is also not clear how thermal effects are factored into the calculation.
Response
1:-
2.-
KWH:rf/G01031*-39 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.180 In reviewing the pressure-time curves for hydrogen detonation (i.e.,
Fig. G.10-1) with the BNL accident analysis group, we were told that the figure may not represent the. time pressure phenomenon accurately.
Any impact due to the inaccuracy of the time pressure phenomenon on the containment capacity should be further assessed.
Response
t'.'
The actual shape of the pressure versus time curve is described in 9.h several classic documents such as " Combustion, Flames and Explosions of Gases" by Lewis and VanElbe, Academic Press, 1961.
s KWH:rf/G01031*-40 2/28/84
m GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.181 The use of dynamic load factors to represent the dynamic effects of H detonation in the analysis of complex structural systems undergoing 7large plastic deformations is questionable.
Provide a discussion to justify such a usage.
Response
The accuracy of a detailed analysis depends on the accuracy of the inputs (e.g. time history of H detonation) therefore a more detailed 7
analysis is not appropriate when inputs are inaccurate.
I u
/ 0#
l KWH:rf/G01031*-41 2/28/84
GENERAL ELECTRIC' COMPANY PROPRIETARY INFORMATION Question 720.182 PP.597:
The loss of integrity has been assumed to occur when either the ultimate tensile strength in the high stress region is reached, or cracks develop.
Reasons for not using other criteria, e.g.,
maximum strain in the steel containment, are not given.
Please discuss the basis for not using other failure criteria.
Response
The structural capability study which was conducted as a part of the much larger PRA study.
As such, the containment analysis portion was not intended to be an exhaustive study of all possible structural aspects of containment behavior.
Rather, reasonable and understandable assumptions were made to produce a rational analysis.
It is important to note, that other assumptions could have been made.
Also, use of maximum strain criteria for failure criteria will probably give answers very close to those resulting from this study.
pu, KWH:rf/G01031*-42 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.183 PP.600:
Details of determining the crack size in concrete (1/4 in, width) are not provided.
Provide the basis for the crack size determination.
Response
9 i*-
4 W
6 KWH:rf/G01031*-43 2/28/84
GENERAL-ELECTRIC COMPANY PROPRIETARY INFORMATION
- Question 720.184 PP. 607:
Temperature range that was considered in the analysis has not been identified.
Please identify the range and the basis thereof.
Response
~~
L-.-.....-
-~ -~
~-
~
joc!.
6
.'p &
A S C,1 4
KWH: rf/G01031*-44 2/28/84
GENERAL ELFCTRIC COMPANY PROPRIETARY INFORMATION Question 720.185 PP.609:
Details of stress calculations in the knuckle region (Table 6.2-1) are not given.
Provide details of the stress calculation.
Response
3, p
.The model is shown in Figure G2-2.
+
KWH: rf/G01031*-45 2/28/84
PROPRIETARY INFORMATION Question 720.186 PP. 609:
Form'ula for a in the column for P should be pr/2h instead of pr/h.
In the same column a should be rIplaced by o.
6
Response
The corrections have been noted.
fr'l,
'C' 4
KWH:rf/G01031*-46
'2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.188 PP. 611:
The ring-stiffeners and the crane girder used in the finite element model (Figure G.2-2) are not described in the report.
-Please describe:these items in sufficient detail for use in analysis.
Response
The details in question have been provided by a previous submittal data 9-9-83.
KWH: rf/G01031*-47 2/28/84 s-
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.189 PP. 616:
The calculations for P in Region 4, and for Pg+Pb I"
Regions 1, 3 and 4 need to be c1Irified.
Response
See the response to question 720.185.
KWH: rf/G01031*-48 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.190 PP. 619:
It is not clear what the compatibility condition (Eq.
G.4-3) represents.
Provide a discussion.
Response
Eq. G-4-3 represents compatibility of radial displacements of the inner shell and the outer shell.
KWH:rf/G01031*-49 2/28/84
LdKLt!KlREVUTCT0HPLSR7 PROPRIETARY INFORMATION Question 720.191 PP. 620:
Details of the pressure carrying capability of the ECCS lines are not given.
Please provide a more detailed discussion.
Response
The external pressure-carrying capability of 24 inch diameter schedule HC piping is well in excess of 300 psig.
I i
l 4
l t
i KWH:rf/G01031*-50 l
2/28/84
[
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION
. Question 720.192
'PP. 655:
The details of the calculation for p = 1.056 x'10 3 should be provided.
Response
Details are provided in G.8.1 of Appendix 15D.3.
KWH:rf/G01031*-51 2/28/84
catuis usesvtenwens PROPRIETARY INFORMATION Questien 720.193 PP. 655 & 656:
The notation used in Eq. G.9-2 is not consistent with Eq. G.9-3.
In Eq. G.9-2, (F (s) should be replaced by f (s) g 7
and x should be replaced by s:
l
Response
The corrections have been noted.
KWH: rf/G01031*-52 2/28/84
,-.,..--------re-,--
- ---- - - - -- - ++- + + * -**~*+*-"* - - - - '-' --* '
- ' ~ ' ^ ~ - * ~ ~ - ' " ' ~ ~ ~ ' '
LNLdMKlMEIMEl% MAX /
PROPRIETARY INFORMATION
-Question 720.194 PP. 656:
How was the expression for the ultimate moment [i.e., 1.5 Mo + (S
/Sy - 1) M ] derived?
ult
Response
Given the following cross-section and stress distribution oe 6,g, F'
~1 n
h 48.Y U
r' H ml l
b Sm M
= 1.5 x 5 bhz.
bh2 ult
)
Letting M, = b h
then 6
lt M
= 1.5 M, + (
- 1) M, ult
- KWH: rf/G01031*-53 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.195 PP. 656:
What do the cases A, B and C represent?
Please explain.
Response
Please~ replace " cases A, B and C" by " cases with gradual pressurization".
J e
KWH:rf/G01031*-54 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.196 PP. 659:
What is the basis for X = 51 psig shown in Fig. G.9-1?
Please provide a justifiqation.
' Response
.I
(,(0f, go' i 4 KWH: r.f/G01031*-56 2/28/84 GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.198 How does the Appendix G fit into PRA analysis? How are the results of Appendix G utilized in PRA? What specific end products of Appendix G are required by PRA?
Response
Appendix G presents a deterministic analysis of the pressure-carrying capability of the containment and drywell structures for static and dynamic loads.
The results of Appendix G are combined with the pressure loads produced by phenomena described in Appendices F and I.
Appendix G provides the value at which containment failure occurs which is utilized to determine the time of containment failure for each accident sequence.
The failure location and site determined in Appendix G is used in the assessment of fission product release pathways.
The specific end products are containment and drywell failure pressure (for static and dynamic loads), failure location (s), and crack size.
(, 'l '
I KWH: rf/G01031*-57 2/28/84
PROPRIETARY INFORMATION Qu7stien 720.199 Provide bases and calculations that backed up the results of pressure capabilities listed in Tables G.1-1.and G.1-2. _Also explain how the conc _lusion that_'
is irrived at:
% :__ O,
Response
I 'i Table G.1.1 is the summary of results presented in section G.2, G.3, G.4 and G.S.
Table G.I.2 is the summary of results presented in Section G.6.
KWH:rf/G01031*-58 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.200 Provide background information of the ASHSD computer program (Ref.
G.2-1).
Can this program, which is based on axisymmetrical finite element shell model, be used for non-axisymmetrical loading?
If yes, please explain how it is achieved.
Response
ASHSD is a computer program for structural analysis of elastic, axisymmetric structures using the finite element method.
It can deal with both axisymmetric and non-axisymmetric loads and the loadings can be static or time-varying forces or ground acceleration.
By approximating a three-dimensional deformation (due to non-axisymmetric loading) with Fourier series of sine and cosine functions of the circumferential coordinate, the program reduces the, three-dimensional problem to an equivalent set of two-dimensional solutions and combines them to yield the three-dimensional solution.
i r ' ",
g 91
> f!'1 KWH:rf/G01031*-59 2/28/84
-GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.201 Indicate the significance and location where various stresses listed in Table G.2-1 occurred, such as P,,
P, P, Q, etc.
j b
Response
Pm is the membrane stress through the region.
P, P and Q occur at the locations of discontinuities.
p;.. I '
KWH: rf/G01031*-60 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.202 At the lower portion of containment shell the 8' concrete shall has been treated as." thin" shell (Equations G.4-2 and G 4-3).
What magnitude of erron has been introduced by this assumption? Compare the calculated capability pressure of 74.9 psi (p. 15.D.3-619) with the calculation of the stress in the steel shell when the concrete wall is treated as thick shell.
Response
j 38
(
4 0 i
KWH:rf/G01031*-61 2/28/84
GENERAL ELECTRIC COMPNNY T
PROPRIETARY INFORMATION Question 720.204 NE 3133 of ASME Section III code consists of design formulas and procedures for shells.under external pressure.
It is not clear what is meant by "the buckling criteria given in NE 3133."
(p. 15.D.3-638).
Response
ASME NE 3222 " Buckling stress values" refers to NE 3133.
- p',l
J KWH:rf/G01031*-62 2/28/84
~.
t%LuK'1MLAsMM1%MAM PROPRIETARY INFORMATION Question 720.205 It is not clear how the failure mode of maximum shear be considered.
It is mentioned in Section G.2, but not in G.8.
Response
The maximum shear theory assumes that yielding begins when the maximum shear stress exists at the yield point in a simple test, since the maximum shear stress in the material is equal to one half the difference between the maximum and minimum principal stresses.
The conclusion given in G.8 is consistent with this theory.
The plastic yield is developed in the knuckle region where the hoop stress o is in compression and the meridonal stress o is in tension. gThey were mentionedinSectionG.2(page15.D.3-808)aswellasinSectionG.8 (page 15.0.3-621).
,i. '
l
.(I'(
'l KWH:rf/G01031*-63 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.206 Define X in Eq. G.8.10 and P in Eq. G.8.25.
Is Q in Eq. G.8.4 transverfeshearorinplaneEEear? There are mixed-ups of notations in App. G:o has been used for stress and standard deviations; P for pressure, load, and -probability; and stress can be o, s, or f.
Response
X,in equation G.8.10 should be replaced by $,
P in equation G.8.25 is defined as the buckling pressure of the t8risphericalshell.
Q in equation G.8.4 is the transverse shear force per unit length.
The o's (o and o ) used in G.9 (pages 15.D.3-654 to 15.D.3-658) are standafd deviltions.
P{...} and P {...} used in G.9 are i
probabilities.
The "P's" used elsewhere represent pressures.
, J:0
- , y, -
ll" fl0 KWH:rf/G01031*-64 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.207 What is the relation between loss of integrity and failure? What is the difference between " fracture" and " crack"? What is " plastic yiel d."?
It is something different from " yield" or " elastic yield"?
Response
(1).
In this study, failure implies breach of containment, or other structure in question.
Loss of integrity implies the same.
(ii).
The terms " fracture" and " crack" are synonymous.
- Also,
" plastic yield", " yield" and " elastic yield" imply the same in this report.
p
, i: *.
gg??'
I KWH:rf/G01031*-65 2/28/84 m.,
L4sdLL1 DEL 6utMfM3UE PROPRIETARY INFORMATION Question 720.208 What are the bases of making the following assumptions?
a.
Structural capability depends only on geometrical dimension and yield strength of material.
b.
Normal distribution of probability of yield and ultimate strength at testing.
c.
Probability of developing cracks varies linearly between S and S
Y ult
- d.
Only
= 0.1 X and
= 0.15 x are considered.
x x
Response
(a) This assumption is based on the conclusion from the work
. presented in Section G.8.1.
(b) See Ref. G.9.1.
(c) A simple assumption to reflect that the higher the stress the higher the probability of having cracks in the region of consideration.
(d) o = Standard deviation x
I
= Coefficient of variation = 0.1 or 0.15 for sensitivity E
purpose, where x is the mean value of pressure x.
- n:
f.I, f/ 9"# ~
KWH:rf/G01031*-66 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.209 What is the physical significance that DLF is less than l? Should it be required that static load be used when DLF is less.than 1 such as the case indicated in Table G.10-5?
Response
s 5;o !,
'g 4 !,
KWH: rf/G01031*-67 2/28/84 1
PROPRIETARY INFORMATION l
j, Qu stien 720.210-Describe how and who does a local detonation affect the structural response (e.g., location and distribution of the pressure pulse, shock wave propagation and refraction / reflection, thermal effects) and how are the dynamic load factors obtained.
Indicate locations of potential failures in containment or at drywell and the probability corresponding to each failure.
Response
210(i).
See section G.10.3 for detailed explanation on the effects from a local detonation.
Locations and potential failures in the containment or at the drywell are discussed in Sections G.3 thru G.7.
(:rf
3
/
r l
KWH:rf/G01031*-68 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION
=
Question 720.211 In assessing the response to non condensible gas generation or to local and global hydrogen combustion, it is stated that loss of containment integrity would eventually occur in the torispherical dome region (15.D.3-661, 662).
What are the physical failure-boundaries? How would it affect the release of radioactive material to the environment?
Response
The physical failure boundaries are primarily in the knuckle region.
Because of the scrubbing function of the suppression pool is maintained, release of radioactive product is negligible.
, 4.^
a
- f f'f,1 -
,t. 'I '
KWH:rf/G01031*-69 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION'
' Question 720.212 If the pressure-carrying capability in the torispherical region is significantly higher than that predicted by the analysis, what would be the worst impact on steel containment due to a hydrogen detonation?.
Response
P t
' ; a C cI E a r C' fic C l',
KWH: rf/G01031*-70 2/28/84
M PROPRIETARY INFORMATION Question 720.213 The specific GESSAR design seems to have minimized the possibility of a fire causing both an initiating event and disabling the mitigation systems.
The analysis performed by GE indicates that the dominant contribution to failure of mitigation systems will be due to random failure of components rather than the impact of fire.
Hence, for a fire PRA performed under this condition it is necessary to include all the areas for which a fire can cause initiating events into any study of the estimations of proper residue core melt probability, regardless of the impact of fire on the mitigation systems.
Provide an assessment of the impact on the core-damage frequency from such consideration.
Response
A screening process was used to identify the critical locations for the fire analysis.
Some areas were excluded as critical locations for one or more of the following reasons:
~
1) the fire frequency was very small, 2) the location had no safety related equipment.
For example, the Safe Shutdown failure probability with all power available is 100 times smaller than when Division 1 and 4 are unavailable.
If all four Divisions of power are unavailtble, the core damage frequency is a.,sumed to be 1.0, 3) the area has a margional amount of combustible material and any
~ fire would be suppressed before spreading or causing damage to safety related systems.
The fire loading for each room in the GESSAR design is given in Section 9A.4 of GESSAR II.
This section describes the quantity and types of combustible material by location.
, p..; fi ey w 'W '
KWH:rf/G01031*-71 2/28/84
MtMMKTGLCLDKMGdSKl PROPRIETARY INFORMATION Question 720.214 Due to the fact that the GESSAR fire PRA takes credit for rated barriers as a strong fire propagation prevent measure, it is important that the adequacy of barriers and penetrations be systematically analyzed.
Provide an assessment of the impact'due to such considerations on the core damage frequency.
Response
The GESSAR design is, in effect, a " pod" design.
That is, the emergency core cooling equipment is grouped in adjacent rooms to form a pod within that area of the plant.
Individual pods may be broken down into subcompartments by walls, floors or ceilings which are rated as fire barriers.
Raceways do.not normally penetrate from a pod (fire area) of one safety division to a pod (fire area) of another safety division.
Where cross-overs between divisions do occur it is because a redundant division must monitor the operating state of the opposite division.
The only time this happens is if there is a sensor from a different division located within a given fire area to determine if the system within that area is functioning properly.
Failure of the intruding division sensor concurrent with failure of the redundant division within the given fire area is assumed and analyzed tu verify that the situation is acceptable.
In addition, the penetrations are rated barriers which have been qualified by type test.
The propagation of a fire across the fire barriers is therefore expected to be highly improbable.
However, without research on the probability of barrier failure, the probability of fire propagation cannot be quantified at this time.
The low frequency to core damage by fire-initiated events can be attributed to these considerations.
// cj 1
e i
I e
KWH:rf/G01031*-72 2/28/84
PROPRIETARY INFORMATION Qu?stien 720.215 Since almost half of the cabling in the GESSAR plant is routed through conduit embedded in concrete, it is important to consider fire propagation to adjacent areas through migration of combustible vapors in conduit.
Prov.ide an assessmen.t of the impact due to such considerations on the core damage frequency.
Response
The statement that half of the cabling in the GESSAR plant is routed through conduit embedded in concrete is not true.
Less than 10% would be a more likely number.
However, where embedded conduit is used, it.
normally starts in an area of a given division and ends in an area of the same division.
For example, the division 3 embedded conduit between the control building and the diesel generator building starts in the division 3 diesel generator building and ends in the division 3 termination cabinet in the control building.
There is no propagation path to any other division.
Also there are intermediate pull boxes along the conduit runs and the conduits are sealed to prevent water flow along them.
This water seal will also prevent propagation of combustion products down the conduit.
Even if the seal is destroyed for a conduit end in the fire area, the other end which is remote from the fire area will remain sealed.
The probability of the propagation of a fire beyond the fire area in which the fire originates is considered to be negligible due to the transmission of combustible vapors.
.( p f j c y r ' '^
KWH: rf/G01031*-73 2/28/84
LeiflMKlM@lWTC~COM3ANY PROPRIETARY INFORMATION Question 720.216 For a plant design such as GESSAR it is strongly recommended that a fire area adequacy matrix be constructed to assess effects of high fuel load areas contiguous to critical areas.
Provide an assessment of the impact due to such considerations on the core damage frequency.
Response
With the exception of the diesel buildings, there are no fire areas which contain a "high fuel load".
The fire barriers which are basically walls, ceilings and floors are designed to have a capability in excess of that required by the relatively low fuel loading within any given area.
The high fuel load in the diesel generator buildings is within the safety area and has therefore been evaluated along with other failures within the area.
Therefore, there are no high fuel areas adjacent to safety areas in the GESSAR design which could impact the considerations on core damage frequency.
,f
, c,;t w !.'
,l KWH:rf/G01031*-74 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.217 Smoke propagation to the control room or the remote shutdown panel should be considered.
Provide an assessment of the impact due to such considerations, including human error probability, on the core damage frequency.
Response
The control room and remote shutdown panels are in two completely different buildings.
Each building has its own ventilation system and smoke venting system and is capable of being operated in the button-up mode where there is no make-up air being taken in from the outside.
The remote shutdown panels are in the auxiliary building and the control room in the control building.
There is a three hour fire rated wall between the two buildings.
The remote shutdown system is divided into two divisions with two separate remote shutdown areas within the auxiliary building.
Each remote shutdown area has its own HVAC system and smoke removal system.
This in actuality provides three independent control areas insofar as smoke propagation is concerned.
For this reason it is considered that there is negligible impact on the probability of core damage from smoke propagation to control areas.
,j
,;, spir '
- 1' '(
KWH:rf/G01031*-75 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.218 How was the screening analysis performed?
Response
See the response to Question 720.213 and section 2.2 of the fire event report.
rr, i -' *I','
KWH: rf/G01031*-76 2/28/84
~~
~ ~
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.219 The probability of loss of offsite power together with the probability of fire in the diesel generator when running has to be considered (not standby mod demand = 7.3x10 g).
The probability of diesel catching fire /per Probability of loss of offsite power (LOS P LOSP x fire in one diesel generator room = 2.1x10 4 ) =
0.3/ year.
Response
The scenario described in the question is not a fire-initiated event, rather a loss of offsite power initiated event which has been.dncluded in the internal event PRA section 450.3.
f' 9.
go I, KWH:rf/G01031*-77 2/28/84
r PROPRIETARY INFORMATION Questirn 720.220 The P factor is not an appropriate parameter to be incorporated since it does not relate the GESSAR plant to the data source.
If used at all the factor should relate the available fuel surface area (not projected area) of the GESSAR geometry to the same quantity in areas from which fire initiating data was taken.
In other words to use the GESSAR plant-specific P-factor, the plant-specific P-factor from which the fire-frequency data was culled should be factored into the area-specific frequency estimates.
Response
The fire frequencies were obtained from Reference 9.
The data did not include a P-factor.
It would be incorrect to assume that every fire initiated in a room would involve the combustible fuel loading in that room.
For example, a fire far from significant fuel sources would._ burn out or be suppressed withant anv impact on plant operation.
?:< b 405 KWH:rf/G01031*-78 2/28/84
^
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.221
~
~
.p a af 4 ".7 a
~
then it is not clear why only thi~2orie-1 Corridor is included inihe analysis and the Zone-2 is excluded.
Response
. --. ~.. -
~
-.,,_.,7.--
4scd
'~
i s
i l
1 i
l l
l r
KWH: rf/G01031*-79 2/28/84 t..
MY PROPRIETARY INFORMATION Question 720.222 It is claimed that the resultant ignition frequencies as defined in this report are more conservative relative to those defined in the Limerick evaluation (P-2-11).
That may not be true in all cases.
For example, in the control equipment room there are 14 power generation control complexes and termination cabinets, hence 14p2.2x10 4 23.0E-03 where GESSAR used
- t d?
Provide a list of cases in which results in GESSAR II are nonconservative and assess its impact.
Response
In general, the GESSAR II fire initiation frequencies are conservative relative to those defined in the Limerick evaluation.
There may be specific values which are lower than those used in Limerick because of design differences, however these values are not necessarily nonconservative.
W 40 i
I KWH:rf/G01031*-80 2/28/84 L
~.
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.223 It is not clear how fire growth time was estimated to be 12 times suppression time for fire propagation through a 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> wall, 50 ft.
of control room and another 3 hour3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> wall.
One possibility is to estimate growth time as 360 min. for the 2 three hour barriers plus about 90 minutes for the separation (30 minutes to each 20 ft.).
This, combined with use of an overall average suppression time of 39 minutes would result in a growth / suppression time ratio of 11.5 minutes.
Was this the reasoning used?
Response
~
m
\\
m
~-
ca
_ = _ _
f.' CI KWH:rf/G01031*-81 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.224 On page 2-13 should d
-divisions be 0.5x10 ghe probabigity for fir 9 amage to redundant
= 2.5x10- or 2.5x10- as stated? Explain?
Response
There is a typographical error.
The number should be 2.5E-6 which is the number used in Figure 2-10 to obtain 4.1E-9 per year.
. y
,y 3. t KWH:rf/G01031*-82 2/28/84 w
--v
,,----,,g r,
m.
"~~h---~"--
- -~--~- - ---
~ - 1
^ ~^
a..
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.225 In control room and control equipment room the fire scenario is not clear.
The suppression probabilities and growth tihies are similar to the Limerick Fire PRA for the situation where a cabinet fire propagates out of the cabinet and ignites cabling at the top.
However, this situation seems not to exist in the above mentioned zones.
A.
Why is propagation to cables considered? Are these exposed cables or are the cables considered internal to the adjacent panel?
B.
What rational was used to conclude that panel propagation has a higher probability then cabinet propagation?
1 C.
Where is Figure A referenced on Page 2-14?
D.
The cited Limerick calculation for transient combustibles (paper) considered the ignition of cable insulation located 10 ft. above the fire.
How does this relate to the cabinets and panels in these areas?
Response
E.
I.
- - U MI+Swi2 A,B,0.
p :;. 4 p h
.,s,Y & !s%fh,6&.
Y s
-Ij,.
. -y,DR t* l M Q fh;.r$..,
s.
Cobiness u~
\\
Tt
,.O -
w-
/
[
Note:
1" Air Space between any fuel material and -
cabinet Bases for assumption that panel propagation is higher probability than cabinet propagation.
KWH:rf/G01031*-83 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.226 In the electrical equipment room the cited COMPBRN evaluation of the Limerick PRA, yielding a 10 minute fire growth and a fire probability to suppress of 0.4, considers a self-ignited cable tray fire propagating to another cable tray 5 ft. above.
However, there are no exposed cables in this room.
Explain the scenario?
Response
As-noted in the-response to Question 720.225, the Limerick analysis was conservatively applied to the GESSAR design, recognizing the differences in the two designs relative to potential fire propagation.
tx' pnyc < k ;
^
KWH:rf/G01031*-84 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.227 It is not clear why in some scenarios barrier failures are considered when in others they are not.
For example, in the control equipment room.the possibility of barrier failure and propagation of fire to the electrical equipment room has not been considered.
What is the criteria for barrier failure?
Response
There is no criteria for barrier failure.
In the example cited, the i
equipment in the eletrical equipment room is the power supplies for the control equpment located in the control equipment room.
Loss of equipment function in either or both rooms as the same effect.
One division of safety equipment becomes nonfunctional.
The wall between these two areas is not required for safety considerations.
It is there to provide a degree of sub-compartmentalization so as to minimize the amount of equipment which might be affected by fire.
Sub-compartmental-ization is a feature which has been utilized throughout the GESSAR design.
A significant portion of the fire barriers within the GESSAR design are between areas of the same division and are not required to meet the safety design criteria for the plant.
They are provided to mimimize the amount of equipment affected by a fire.
The considerations for this
~
were plant availability and cost of recovery.
From a strict safety standpoint, the bre'ach of these barriers would not result in the loss of any additional safety systems.
& pizyrid"'b' KWH:rf/G01031*-85 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 270.228 It is not clear which initiating events are associated with each critical fire area.
Response
'5 f.; %
The frequency of loss of feedwater is $1.97 events per year.
The frequency of loss of feedwater due to fires is a very small subset of this frequency.
Therefore, the internal event PRA covers this case.
)
'KWH:rf/G01031*-86 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.229 In the auxiliary building electrical equipment room and the cable tunnel the necessity of considering transient combustible exposure fires is indicated in Table 2-1.,'but no analyses are presented.
l
Response
l
?
.~. $
4:. c
\\
l I
l i
l i
L l
i l
i l
1 i
l KWH:rf/G01031*-87
-2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.230 For all critical locations considered in the analyses given in Section 2.4.2, explain the following:
Which initiating event (transient) results as a consequence of a.
the fire?
b.
Which transient mitigating systems (if any) are disabled by the fire? This information should be given for the case in which the fire is suppressed and for the case it is not suppressed.
c.
Do the fires in their critical locations disable cables which supply the PCS?
Response
a.
i b.
a E
i IF f
t r
C.
l g
b: S, 4 0 **
KWH:rf/G01031*-88 2/28/84 7,---
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.231 In page 2-14 is is stated that the control room is assumed to be nonfunctional during a fire in the room.
However, in the corresponding event tree (Fig. 2.7) it is assumed that the control room is non-functional only if the fire is not suppressed.
Explain this difference.
Response
On page 2-14, the control room is non-functional only if-the fire is -.
not suppressed.
pp/propne &nf 4
KWH:rf/G01031*-89 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORHATION Question 720.232 In event tree for the control room (Fig. 2.7) and for the control equipment room (Fig. 2.8) it is assumed that for a shutdown from the remote shutdown panel all systems (HPCS, RCIC,'CPCI, RPCS, CIS, ADS, PCA, and RHR) are potentially operable from that panel.
However, in the GESSAR-II SAR the only systems control from the remote shutdown panel are: RCIC, one RHR loop and 3 non-ADS SRVs.
Explain this difference.
j
Response
[a i f-I\\
h,.
Reference:
USNRC, Safety Evaluation Report related to the final design approval of the GESSAR II BWR/6 Nuclear Island Design, Docket No. 50-447, NUREG-0979, April 1983.
KWH: rf/G01031*-90 2/28/84
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.233 In page 2-15 it is stated that the " failure to the remote shutdown panel was taken to be 1x10 gafely shutdown'from from NUREG-1278".
Explain how this value Gas obtained.
Response
3
--...m.
-.......m
..... =
{
3a b t/c b KWH:rf/G01031*-91 2/28/84
DRAFT FOR PUBLIC REVIEW NUREC/CR-1278 SAND 80-0200 HANDBOOK OF HUMAN RELIABILITY ANALYSIS WITH EMPHASIS ON NUCLEAR POWER PLANT APPLICATIONS October 1980 Alan D. Swain Henry E. Cuttmann Statistics, Computing, and Human Factors Division Reliability Analysis Department Sandia National Laboratories Albuquerque, NM 87185 This work was prepared for and supported by Office of Nuclear Regulatory Research United States Nuclear Regulatory Commission Washington, DC 20555 Under Interagency Agreement DOE 40-550-75 NRC FIN No. A-ll88 For infor=ation on obtaining copies, contact Division of Technical Information and Docu=ent Control, U.S. NRC, Wash. D.C. 20555 20-/
20-19 Manual opsratitn of Centrols Tabla 20-13 Manual Operation of Controls (Chapter 12)
Manual operation of controls includes the operation of all kinds of switch'es, connectors, and valves. Table 20-13 applies to controls other
~
For errors of omission, than valves and lists errors of commission only.
If controls use the applicable EEFs in the subsequent tables on valves.
I The effects of tagging are handled as pairs, assume CD between them.
are described in the following section, " Valves."
Table 20-13. Probabilities of Errors of Commission in Operating Manual Controls (from Table 12-1)
HEP Task
.003 (.001 to.01)
Select wrong control in a group of identical controls identified by labels only
.001 (.0005 to.005)
Select wrong controls from a
' /'
functionally grouped set of I
controls
.0005 (.0001 to.001)
Select wrong control from a panel with clearly drawn nimic lines
.0005 (.0001 to.001)
Turn control in wrong direction (no violation of populational stereotypes)
.05 (.01 to.1)
Turn control in wrong direction
~
under normal operating con-ditions (violation of a strong populational stereotype)
.5 (.1 to.9)
Turn control in wrong direction under high stress (violation of a strong populational stereotype)
.001 (.0001 to.1)
Set a multiposition selector switch to an incorrect setting
.01 (.005 to.05)
Improperly mate a connector add) "/f l
-~,I GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.234 In Figures 2.14 and 2.15, the unavailabilities used for the HPCS (2.5E-2) and for the RCIC (8.2E-2) are smaller than the ones used in
. the GESSAR-II internal e' vents PRA.
Explain this difference.
Response
3ob (J C l' 1
4 KWH:rf/G01031*-92 2/28/84
GENERAL ELECTRIC CDMPANY PROPRIETARY INFORMATION Question 720.235 In Figures 2.14 and 2.15, the unavailability of the ADS with loss of one division of power (Division 1 for Fig. 2.14, and Division 2 for Fig. 2.15) is 8.0E-5.
Is the 120V AC.NSPS bus E(F) (Gate EAC120NSE(F) in GESSAR-II EPS FT in Fig. D.2-14) assumed to be lost when one division of power is lost?
If so, explain the value obtained from the ADS unavailability.
If not, explain what is meant by loss of one division of power
Response
s
~
-73 s KWH: rf/G01031*-93 2/28/84
~
GENERAL ELECTRIC COMPANY PROPRIETARY INFORMATION Question 720.236 Provide an assessment in terms of core damage frequency and risk of an inadvertent operation of the fire sprinkler system as a potential initiating event of internal flooding in the GESSAR-II.
design, given the fact that there is present a large number of electrical equipment which is not qualified for adverse environmental conditions in the vicinity of the sprinkler system.
t-
Response
The sprinkler systems UTILIZE CLOSED HEADS.
The piping is charged with water continually.
Actuation of a sprinkler head is by melting of a thermal link in each individual head.
Therefore inadvertent actuation of the sprinkler system requires the failure of an individual head.
This is equivalent to a pipe leak in that area.
The compartmentalization of the GESSAR II design is such that a pipe leak in any given area containing sprinklers can only affect one safety division.
Therefore, the consequences of a spriakler head failing open is the same as a pipe leak.
The probability of a sprinkler head failure is not considered to be higher than that of the failure of a piping system.
In the course of the hkC review of this response, GE was asked if the sprinklers could be manually operated ano if they could thus disable electrical divisions. uAs noted above (and in Appendix 9A of GESSAR),
actuation of sprinklers requires more than simple manual initiation, thus their inadvertent actuation would not be expected.
Furthermore, if the sprinklers in an area were actuateo they would impact at most one division due to the divisional separation inherent in the GESSAR design.
Ar.y adverse impact of the fire suppression system on electrical components has been minimized by design as noted in Appendix 9A of GESSAR (i.e., placement of components on elevated pads, provision for i
floor drainage, etc.).
The GESSAR Fire H6zard Analysis performed on every room or area, on a floor-by-floor basis for each building of the Nuclear Island desing defines design provisions for protecting the functional capability of safety-related systems and associated cabling from the results of inadvertent operation of protection systems.
I
^
, // popm ~y t
i s
KWH:rf/G01031*-94 2/28/84
(
)
r GENERAL ELECTRIC COMPANY l
PROPRIETARY INFORMATION Question 720.237 Provide an assessment in terms of core damage frequency and risk of the flooding of diesel generator and turbine buildings following a severe earthquake and a. loss of offsite power.
)
Response
The only sources of water in the D/G rooms are support systems for the D/G.
Therefore loss of D/G by breakage of the piping in support systems and D/G building flooding would have the same impact.
Since the reactor is isolated from the PCS during a severe earthquake a coincident flood in the turbine building would have no impact.
x /c f ytrf t!! In"(
KWH: rf/G01031*-95 2/28/84
_