A06867, Forwards Addl Info Re NUSCO-151, Haddam Neck Plant Reanalysis of Non-LOCA Design Basis Accidents, Per NRC 871015 Request.Responses to Questions from 870630 Meeting & Revised Pages for NUSCO-151 Also Encl
| ML20236R183 | |
| Person / Time | |
|---|---|
| Site: | Haddam Neck File:Connecticut Yankee Atomic Power Co icon.png |
| Issue date: | 11/19/1987 |
| From: | Mroczka E CONNECTICUT YANKEE ATOMIC POWER CO. |
| To: | NRC OFFICE OF ADMINISTRATION & RESOURCES MANAGEMENT (ARM) |
| References | |
| A06867, A6867, NUDOCS 8711230142 | |
| Download: ML20236R183 (51) | |
Text
- - - - m _. _ _
7 l
.a na.h ;
S. -
1 1
CONNECTICUT YANKEE ATO MIC POWER COMPANY j l
B E R L I N, CONNECTICUT P.O. box 270 e HARTFORD, CONNECTICUT 06141-0270 f
j l
i
['[". "' November 19,1987 {
i Docket No. 50-213 A06867 Re: 10CFR50.34(b) l U. S. Nuclear Regulatory Commission i Attn: Document Control Desk Washington, D. C. 20555
References:
(1) F. M. Akstulewicz letter to E. 3. Mroczka, " Request for Additional Information Concerning Reanalysis of Non-LOCA Design Basis Transients," dated October 15,1987.
(2) E. '3. Mroczka letter to C. I. Grimes, "Haddam Neck Plant Reanalysis of Non-LOCA Design Basis Accidents, NUSCO-151," dated June 30,1986.
(3) E. 3. Mroczka letter to U. S. Nuclear Regulatory Commission, " Revisions to Non-LOCA Reanalysis," dated May 8,1987.
Gentlemen:
Haddam Neck Plant Additional Information Reanalysis of Non-LOCA Design Basis Accidents Reference (1) requested additional information on Topical Report NUSCO-151 non-LOCA transient analysis. Northeast Utilities Service Company (NUSCO), on behalf of the Connecticut Yankee Atomic Power Company (CYAPCO), is hereby providing, as Attachment 1, responses to your requests for additional information. In addition, during a meeting held on June 30,1987 between the NRC and CYAPCO representatives, informal questions were raised about the calculation of the 95/95 DNBR limit. Responses to these questions are provided l in Attachment 2. Finally, revised pages for NUSCO-151 (provided by Reference (2) and revised by Reference (3)) are provided in Attachment 3. These revisions are being made for the following reasons:
o Steam-Line Break - A revised analysis has been performed to take into account a split core model.
o Excess Feedwater - A correction is being made to the text (page 4.37).
o Loss of Load - An incorrect reference statement to a table has been j deleted (page 4.106).
1 o Steam Generator Tube Rupture - Reference to Figure 15.6-3 'has been changed to Figure 4.10-8 which has been added (page 4.92). A sentence '
further discussing the figure has been added.
8711230142 87'1119 PDR P
ADOCK 05000213 m.
l 1 n00 l PDR ..
o --
.. e.
U.S'. Nuclear Regulatory Commission A06867/Page 2 ,
November 19,1987 '
All changes have been Indicated with change bars.
We trust the NRC Staff will find these responses adequate to resolve all ;
remaining concerns regarding NUSCO-151. Should you have any further questions, feel free to contact us directly.
Very truly yours, CONNECTICUT YANKEE ATOMIC POWER COMPANY j E. J. Mriipfka (/ i Senior Vf' e President cc: W. T. Russell, Region I Administrator A. Wang, NRC Project Manager, Haddam Neck Plant J. T. Shediosky, Resident inspector, Haddam Neck Plant l
l 1
- , '. ~
Docket No. 50-213 A06867 l
1 I
l l
Attachment 1 Response to Requests for Additional Information on the Reanalysis of Non-LOCA Design Basis Accidents 1
i l
I November,1987 l
l
___.._._ _ __ _ _ J
1 d; e, _l 1.0 General-A. Explain why NUSCO did not analyze transients in the category of Increase in Reactor Coolant Inventory, which constitute-the fifth -
category of SRP Chapter 15 transients.
B. Provide and justify the nodalization diagram and the conservatism used in code model selection and in the input assumptions for each transient.
C. Explain and justify how the submitted analyses in NUSCO 151 bound those which vere not submitted.
i D. Identify for each transient analyzed the core burnup selected to a yield the most limiting' combination of physics parameters.
f !
E. Explain how the scenario selected for each transient with the I identified initiator is a moderate-frequency transient and most j limiting. j i
Response
A. TheNUSCOgpicalReportNUSCO-151,HaddamNeckNon-LOCATransient Analysis , was submitted to replace the current design basis l analysis with a new analysis based upon the RETRAN/VIFRE methodology. Thus, only those accidents that constitute the 3
j Haddam Neck design basis, were included in NUSCO-151, with one j exception. The Reactor Coolant Pump Rotor Seizure and Shaft Break !
vas included in NUSCO-151, since this was an open item of the Systematic Evaluation Program (SEP). Since there is no analysis: .) i in the design basis for Haddam Neck for Increase in Reactor .l Coolant Inventory, no analysis was provided in NUSCO-151. 'I B. The nodalization diagrams used for all the analyses, except idled j loop startup and the revised steam line break.' analyses are shown -
on Figures 1 and 2. For these two transients, a different vessel nodalization was used. The vessel nodalization'is shown on Figure .
- 3. JgificationofthenodalizationwasprovidedinNUSCO !
14p'1 and the additional information supplied on September 2' . In addition, NUSCO 140-1 also provided justification and ,
conservatism in code model selection.
l '*' Letter from J. F. Opeka to Office of Nuclear Reactor Regulation,
) "Haddam Neck Plant Reanalysis of Non-LOCA Design Basis Accidents",
June 30, 1986 ;
Letter from E. J. Hroczka to U.S. Nuclear Regulatory Commission, .
"Haddam Neck Plant Additional Information - Reanalysis of Non-LOCA Design Basis Accidents", September 2, 1987.
W. G. Counsil letter to J. R. Miller and D. M. Crutchfield, dated July 30, 1984 and enclosures entitled "NUSCO Thermal Hydraulic.-
Model Qualification volume 1 (RETRAN)", NUSCO 140-1, dated August 1, 1984 and "NUSCO Thermal Hydraulic Model Qualification Volume II (VIPRE)", NUSCO 140-2, dated August-1,.1984. -;
__-_____-______.___.____.._____-m__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . _ _ _ _ . _ . . - _ _ _ -..._.. __ . u
a s System input parameters were given in Table 4.0-1 of NUSCO 151.
For the parameters where two values were given, the attached Table 1 provides which value was used for each transient. Sensitivity studies were performed whenever the conservative direction of a i parameter was not obvious.
C. As discussed in A above, only those accidents that constitute the Haddam Neck design basis were reanalyzed in NUSCO-151. Since the Facility Description and Safety Analysis Report (equivalent to FSAR) does not provide discussions of accidents not analyzed, none was provided in NUSCO-151.
1 D. Sensitivity studies were performed to determine whether minimum or ?
maximum values were most limiting for feedback parameters such as j the Doppler coefficient and the moderator coefficient. In j general, the most limiting value was used independent of time in j life. This can lead to inconsistent but conservative sets of input such as an End of Life Moderator coefficient and Beginning of Life Doppler coefficient. The limiting values for Doppler coefficient and moderator coefficient are given in Table 4.0-1 of )
NUSCO-151 and the conservative direction for each accident is summarized in Table 1.
E. The design basis analysis for Haddam Neck is not divided into probability categories. Thus, the analysis provided in NUSCO-151 is also not divided into probability categories. The scenarios that were analyzed in NUSCO-151 are identical to the scenarios analyzed in the current design basis. However, sensitivity I
studies on parameters, such as power level, feedback coefficients and initial RCS and secondary conditions, have been performed to identify the limiting initial conditions for each scenario.
2.0 Startup of an Isolated or an Idled Lqog Explain and justify the statement " Effects of imperfect thermal mixing are included in the analysis by using bounding assumptions to conservatively model this effect." What single active failures were examined to enable NUSCO to conclude that no such single active failure could impact the pre-trip event consequences and how was this conclusion reached?
l Response The RETRAN02 model used to perform the Idled Loop Startup analysis employed a " split core" model. The model contained two downcomer l lover plena, core, and upper plena' nodes. No junctions between the two downcomer or core volumes were modeled, thus no mixing between these volumes was allowed. Mixing in the lower and upper plena was allowed in order to initialize the system with flov through the
j l
1 quarter core node. Once the transient started, flow reversed in the cross-plena junctions and no flow from the " hot" lover plenum entered the " cold" lover plenum. This treatment of mixing is conservative since none of the " hot" side cold leg flow was mixed with the " cold" side of the core, thereby reducing the reactivity transient. ]
The idled loop startup event is over in a very short period of time (5 seconds). The event consists of opening the loop isolation valves, j starting a reactor coolant pump, and generation of a reactor trip on ,
high core power (neutron flux). No other systems are credited in mitigating the consequences of the event. The starting of the RCP is the event initiator. Single failures which would stop the RCP are not considered since this would also stop the reactivity addition of the cold loop. A single failure of the RPS which would prevent rod 1 J
insertion is not credible. Thus, there are no single active failures vhich can exacerbate the reactivity addition or the system response.
3.0 Excessive Feedvater A. Justify the different times assumed for operator actions (37 seconds for 4-loop operation and 45 seconds for 3-loop operation) j to trip the main feedvater pumps for the cases with potential for l SG overfill. Does use of a greater operator reaction time lead to l steam generator overfill? !
B. Justify not using the split-core model.
I
Response
l A. The difference in the time available for operator actions between l the 4 loop and 3 loop case is mainly due to the difference in the !
initial SG vater mass assumed between the two cases. The i conservative maximum initial SG water mass for 4 loop operation is )
larger than that for 3 loop operation (57,000 lbm. vs 53,000 lbm.). The affected SG vill not be overfilled if the operator actions are performed within the calculated times of 37 seconds q for four loop operation and 45 seconds for three loop operation. ]
These calculated times are conservative; however, if these times are significantly exceeded, steam generator overfill vould occur. 1 B. The core flow model used for the analysis assumes complete mixing [
in the lover plenum. The excess feedvater is not a limiting i overcooling event and it was shown that the minimum DNBR for the !
event is well bounded by the Steam Line Break Event. During the transient, the decrease in RCS cold leg temperature is minimal
(< 5 F) for the affected loop and thus, the effect of incomplete ;
mixing due to asymmetry in cold leg temperature is considered to j be negligible. ;
1 1
l l
i
)
~ .A
4.0 Steam Line Break A. Explain the nodalization used in the thermal-hydraulic analysis and justify the conservatism when compared with a split-core model in the thermal-hydraulic calculations. j B. Justify the Boron transport model used in the analysis.
Response: l A. The steami g break analysis results given in Revision 2 of )
NUSCO-151 , vere based upon the RETRAN predictions using a j single core node. The asymmetric cooldown affects were taken into ;
account by inputting the cold leg temperatures, prior to mixing in j the core node, into the 3-D reactor physics program. The ]
unaffected and affected cold leg temperature, the RCS pressure and
- the boron concentration were used in the static 3-D reactor physics program to predict the magnitude of the return to er.
This prediction was then compared with the RETRAN results.
Iterations were performed by adjusting the feedback in RETRAN l until the RETRAN power prediction matched the 3-D physics !
prediction. The RETRAN system prediction along with the 3-D physics prediction of power distribution was then used in VIPRE to calculate DNBR and peak center-line fuel temperature. ,
1 In response to the concern raised about the use of a one core node I model in RETRAN, the steamline break analyses for 4 loop full power and hot zero power were repeated using the split core model shown in Figure 3. In this model, the reactor is modeled with two ,
1 core nodes and two inlet and outlet plena, split 3/4 and 1/4 by volume. The cold side temperature was used for the reactivity calculation in RETRAN. The same iteration process was performed with the 3-D reactor physics program. The results with the split core model shows a 4% higher return to power. Since the split core results are more limiting, the steam line break analysis in NUSCO 151 has been revised and the results are given in Attachment
- 3. As requested, additional information has been provided in Attachment 3 about the crossflow modeled in the reactor vessel.
B. A sensitivity study on the boron modeling has been performed to determine the effect of the boron modeling on the steam line break results. Three different models were evaluated in this study.
The first model is the model used in NUSCO 151 for the steam line break analysis. This model is a slug flow representation for the transport of boron. This model uses the RETRAN control blocks to calculate the core boron concentration. The calculational process can be outlined as follows:
I
Letter from E. J. Mroczka to U.S. Nuclear Regulatory Commission, "Haddam Neck Plant Revisions to Reanalysis of Non-LOCA Design Basis Accidents", May 8, 1987
~1 a L.
. * - First the cold leg injection point boron concentration is' calculated using the safety injection boron concentration-and the flow rate.
- - Time dependent loop transport times are then calculated, using the nodal' volumes and volumetric flow rates as. , ,
calculated by RETRAN.
- - The transport-times are then used to march the slug of'the borated water to.the core,1the steam generators and finally back to the' starting injection point, where'the.
process.is repeated. .l l
We believe that this is the best. representation for calculating-boron transport since it correctly account for' delays.in' transport > <
around the loop and it does notLartificially mix boron into. i volumes.that are ahead of the' slug. -i j
The second model uses the.RETRAN control block logic.to solve the j time dependent conservation of mass equation.for.every node. .This 1 model in the limit, as'the number of nodes are increased, vill-approach the first model. . ]
Finally, the third model.is an overly. conservative'one in which ,
the total RCS volume, including the stagnant' regions such as the -
RV head, is lumped as a single volume and an~ average boron concentration is calculated.
The comparison of the boron concentration for the 4 loop hot'. full power is shown in Figure 4. 'While significant differences in'.the ]i boron concentration were predicted for times far .into'the
~
accident, there is very little difference at the' point of maximum return to power. The slight. changes in peak return'to power have been accounted for in the margin added in the VIPRE calculation.
In conclusion, we believe that the slug flow model' described above and in the response to.the questions on the RETRAN topical, NUSCO-141 is the best representation of. boron injection into the RCS. However, the SLB results are not sensitive to the: boron modeling, and'no significant changes in results' vere seen even with overly conservative boron modeling. j J
5.0 Steam Generator Tube Rupture A. Explain in detail why tripping the reactor at event initiation-p maximized the radiological consequences. Explain further why1the. ;
E assumption of concurrent' loss'of offsite power is conservative. l 1
l
\
4 t
(
B. Provide Figure 15.6-3 as referenced on page 4.92.
C. Explain what is meant by "blovdown" using steam generator safety !
valves. l D. Demonstrate, by carrying out the calculations further in time, that the assumption of 30 minutes for isolation of affected SG is reasonable and conservative. Do the same for the rest of the transient. It appears that the discussion pertaining to the cooldown phase is based upon assumptions. I E. Explain the procedure which the licensee plans to use the intact steam generators to bring the rest of the plant to the RHRS entry (
conditions. How was the steam release data (730,000 lbm, Table i 4.10-2) from the intact SG obtained.
Response
A. Reactor trip at event initiation and concurrent loss of offsite power were assumed to minimize the steam flow to the condenser. i By minimizing the steam flow to the condenser, minimum credit is )
taken for any iodine concentration build-up in the condenser. The total steam released from the steam generators during the event is essentially independent of whether it is released to the condenser )
or to the atmosphere. Maximizing radiological consequences is I done by maximizing the iodine released to the atmosphere. By minimizirg any credit for iodine concentration build-up in the condenser, the iodine released to the atmosphere is maximized.
B. The figure should have been Figure 10.4-8. This figure is attached as Figure 5 and also as part of Attachment 3 (revision to NUSCO 151).
C. The steam generator safety valves do not reclose at the same pressure that they open. Blowdown is defined, in percentages, as the ratio of the delta between the open and close pressures and the open pressure. For the SGTR analyses 15% blowdown was modeled. The result of this, including allowance for valve setpoint drift, is that the valve opens at 970 psia and closes 15%
lower at 824.5 psia. Nominal safety valve blowdown is approximately 5%. Larger blowdown results in a lover average steam generator pressure and, therefore, larger primary to secondary leakage.
D. The SGTR Emergency Operating Procedures (EOPs) follow the guidance of the Westinghouse Owners Group Emergency Response Guidelines Revision 1. Basically, the E0Ps instruct the operators to identify and isolate on the secondary side the ruptured steam generator. Then the RCS is cooled down and depressurized so that primary to secondary leakage is terminated. Cooldown and' depressurization is required prior to loop isolation valve closure so that the maximum delta pressure across the valve during closure does not exceed the valve's design value.
m o l l
\
l 1
I Following termination of primary to secondary leakage there are four options for cooling down the ruptured steam generator. Two i of these options utilize the steam generator blowdown system, one utilizes reducing RCS pressure so that leakage is from the H secondary to the primary and the fourth utilizes dumping steam from the ruptured steam generator. The steam generator blowdown 1 options vould utilize the primary loop drain system if the loop !
isolation valves are closed or normal primary system cooldown if j they are open. Cooling down the part of the RCS, not isolated by '
loop isolation valve closure, to RHR system entry conditions is only accomplished by dumping steam from the intact steam generators to either the atmosphere or the condenser.- j l
The only time constraint the operators h;.ve following a SGTR is to j terminate primary to secondary leakage prior to the steam i generator filling vater solid. After termination of leakage, an assessment of the plant status would be made prior to further RCS cooldown to RHR entry conditions. l The amount of steam released to the atmosphere from the ruptured steam generator is driven by two factors. One is the amount of ;
primary to secondary leakage. The second is heat transfer from i the primary to secondary. Maximizing these two maximizes the steam releases and, therefore, the radiological consequences. The operator actions of cooldown and depressurization reduce both the leakage and heat transferred. To conservatively maximize the radiological consequences, these actions are not modeled in the analyses. Instead, no actions are modeled for the thirty minute time period which bounds the time required to terminate the ,
primary to secondary leakage.
As discussed above, there is no time constraint to reach RHR entry conditions. Therefore, assumptions consistent with standard Westinghouse analyses were made to determine the radiological consequences. These assumptions were given on page 4.93 of NUSCO 151. Based on these assumptions, RETRAN analyses were performed to determine the steam releases to the atmosphere.
E. The procedure to bring the plant to RHR entry conditions vould be to dump steam from the intact steam generators to the condenser or atmosphere. Steam generator inventory would be maintained using either main or auxiliary feedvater. The steam release data were calculated by utilizing RETRAN. In order to maximize the releases and consequences, it was assumed that the operators restart the reactor coolant pumps and dump steam to the atmosphere. This maximizes the energy that must be removed. Also, since one of the options for cooling the ruptured steam generator is to utilize secondary to primary leakage, the four loop RETRAN model was used. This accounts for the steaming required to cool the ruptured steam generator.
1 ,
6.0 Loss of Normal Feedvater Flow A. Explain what is meant by a " conservative core residual heat generation rate."
B. Demonstrate that the case analyzed is the worst case and explain why loss of offsite power vould represent a more conservative assumption than vould offsite power available.
Response
A. The core residual heat generation rate utilized the fission product decay and actinide decay in RETRAN. The fission product decay is comparable to the 1971 draft ANS 5.1 standard. The heat rate conservatively modeled infinite reactor operation. A multiplier of 1.0 was used.
B. The analyses performed assumed that the condenser was unavailable.
The result of this assumption is that the steam generator pressure vould be the same regardless of the availability of offsite power.
The auxiliary feedvater system is independent of offsite a.c.
power since it utilizes steam-driven pumps and air to close valves. Therefore, assumptions regarding offsite power availability do not impact auxiliary feedvater system availability.
s Based on the above, assumptions regarding offsite power i availability would not impact auxiliary feedvater system !
performance following a total loss of normal feedvater. The )
impact of offsite power availability would be due to reactor l coolant pump (RCP) operation. Section 4.12 of NUSCO-151 discusses i RCP operation following reactor trip and the impact on the loss of normal feedvater transient. )
l The criterion used in determining the vorst case loss of normal !
feedvater transient was peak pressurizer level. Sensitivity studies were run to determine the worst single active failure.
Also, whether with or without loss of offsite power is more l conservative was looked at for four loop operation. The effect of J RCP operation is more pronounced during three loop operation when i combined with the worst single active failure. The results of the j analyses is that with offsite power available is more conservative !
and the vorst single active failure is as discussed in NUSCO-151. {
7.0 Reactor Coolant Pump Rotor Seizure and Shaft Break Reactor Coolant Pump Rotor Seizure A. Why was this analysis performed at 100% power and not higher l initial power than 100% for conservatism? j 1
I l
)
f
[ .. .
B. Which peaking factors and axial power profiles are used for this analysis? How many axial nodes were modeled in the core and at what location are the minimum inner surface heat transfer coefficient and the peak gap conductance which would maximize the predicted cladding inner surface temperature?
C. Explain how the new scram reactivity as defined (most reactive rod assembly stuck out) in the Technical Specifications would ,
impact this transient. i l
Response
A. The analyses of the reactor coolant pump rotor seizure and shaft break events applied a 2% uncertainty to the initial power level. Therefore, the initial power level assumptions j (for full power conditions) were 102% for four loop operation 1 and 67% for three loop operation. ]
B. The analyses of the reactor coolant pump rotor seizure event use top and bottom peaked axial power shapes which are shown l in Figures 4.0-3 and 4.0-4 of NUSCO-151. The top peaked axial power is used in the DNBR analysis and assumes limiting conditions for axial offset and burnup history. The bottom peaked axial power shape is used in the inaximum fuel temperature analysis in order to maximize the linear heat 3 generation rate (i.e. kv/ft.). 1 The limiting radial peaking factors for four and three loop conditions are listed in Table 4.0-1 of NUSCO-151. I The rotor seizure analyses for four and three loop conditions i use a maximum radial power factor as defined in the Technical Specification, including the appropriate uncertainties.
The VIPRE analyses performed for the rotor seizure event uses 34 axial nodes in the thermal hydraulic solution. The axial power profiles are described by 27 points.
A peak clad temperature analysis was performed for the rotor seizure event. In this analysis the limiting conditions of gap conductance and cladding heat transfer coefficient occurred about 22 inches from the core inlet.
C. The revised shutdown margin requirements in the Technical Specifications affect the scram reactivity assumptions for four and three loop hot zero power conditi,ns. The limiting power level for the rotor seizure and sha ' break events is at full power conditions and the results are not affected by these Technical Specification changes.
t
/
g a ,
, i tt 6
3, N
,g 4 , g
,8 y gb dg e
y9 ,@ g
, 8 ,
,b 4 "'
]w
- ' 3
\ k' \ E 4 4 9
e e m
" m lo u Y A " u lo e- 8 v W v S o t
o v t w
g )
o P
r g(6 9" s
1 S
L E
D T
N R
E T
A 6"
O M
I E
M N
E H I y 2 I* *, i @
~-
iiiiiI l 'I i
i 'iogg
'i 1
I l 1
)m TAI E
S M M
Ww 23 N
o l %
1 A A A E A
W wn c
,sR@(lN '4 11 E B lR 2 ts N
H MM A A
,m@_ g
@ 2 0
6 G O i i c
A TE t
% 6#
I F A N M 5 s a4
- r r
@ j!
g i l
R - - - .
i i i i ; ' ,i i T ,i l }i p;l i i,i' , , , ,
J E p h -
R - -
1 R
Y C
gM A L4 A
g '
u s H L
9
- 2" " 9 0
4 e
0
,i e
m u
9 6
" ~*
r ru lo
< o v
v n
s't o e
r o a
' '* f r
j ),
1 1
q y n. p 2,b ee g \o g I. I) N a y ,p~8 g' J=
b y
34
., h 6
.. s8 ,
B
'O gQ;9 g 1 '
1 1
%L -
p e
/
S k s ,, w
\ -
E E v
l -
v t -
A R R V E E
T u ; E T
Y A A
T Y E
F A
S E
D E
E W r E
r A
W D
E E
F s m F x N I x N e A e I
A A M A M
I c g u r 3 n aS N g O' t
4 q ,
,O yru Q
Q ,gf , G, gu
[ ,
a s
\
,,O u h
O[e C -
u s'
g.
\
.O
$kU 3
hg S
4 9
e U -'
4 9
e m m u ' u
- r lo v
o v s g; .y h O lo t
v o b o+a t
u )
R" p
r ag(t gj s l L
j u c U k E
D g T N
E 1
c F,' J 7 i l , i1 il ! : 8 '
O M n I i - - 's 1 i ',!
N ' ; . - , ' l M
e
)
,, _4*+
w E
2 S - s
. _mAION T d,pgg A A C
@o@
.A. . >
E B u ,..
+ 6 R
U 2 G 0 .,_
s N e g eo!i i
+me, F N I
A R
T
'
6 l 8 , 8
~
i 5
FIGURE 3 CY RETRAN02 SPLIT CO2B MODEL l
8 9
=
o O
I I I l
/
O
_g m, 7
O E
DO C m
.s O M %
o n
2 e
- =
r O ?
G O c -
- L O
> to CA i D O O M o
m to r
cc O C
.s -
Z W C Z O
Orr,
\ Z O O o
\
/~
2 UJ m U-)
E v
M a
e L
O Z
o
//
O ~ ~ W W L w 8m2 g 5 e
/ CO 5J Z J
, C O 1 u 1
/,/
// -
m O
U 3
c Q
\
/E/. O e
[
n M
4H -
O CD 1 I i 1 0 *-
I- 2- E- b9
. (t) A11 ail 0038 NOB 08
FIGURE 5 i STEM 1 GENERATOR TUBE RUPTURE FRQ1 HFP i
E' i i i > i i i i
E Charging Flow 5 r q/u,r ;ty.WN?NN:W!!!:a' *:::::.'JN~ ,' .W
- a. l (g . E L
i
= .
n I l l l I t f I l
d 2n 5: - 7;: 3::: 1:r: is:- 17:- 2:'
l I}Mi !$ECO '!) ;
i 1
l i
i
. l l
I 1
l l
l l
I s s
e r
P S
C R
e x x l r nn n n n n n n x n n au ii i i i i i i a a i a i i s mm m m m m m m m m m m m t s i e nr IP r
o t s as ra eM n nn x x x n I x x x n n x er ii a a a i
- a a a i i a Ge t
mm m m m m I m m m m m m ma aV t
e S
rt ei l c t t t t t t t t t t t pi ss s s s t s s s s s s pf aa a a a s a a ) a a a a oe ee e e e o e e 2 e e e e DD ll l l l m l l ( l l l l r e 1 e r E
t e u t
L m a B a rt A r en T a pe P mi ec t t t t t t t t Ti ss t t t t s s s s s s f aa s s s s a a 3 a a a a rf ee o o o o e e 2 e e e e oe ll m m m m l l ( l l l l t o aC r
e d
o M
e t r nu at l a or Cp oe m xx x x x x x x ) x x n x l e aa a a a a a a 2 a a i a aT mm m m m m m m ( m m m m i
t t i e nl y e I n l r I b u r m t o e p r tk e s u e oa d s s r R t Re e a A o a l t e t e v pB a a r r l c b d m c l c ey o a u e ut i op n tl r e T e Pf d t so r I sb t R k F o i I o e um n a r th R rr L t d l e o dv e o l nS ce n a a Cs C eo r t a a d b w ad v o s cl B a d m ld el uo e d L dA r rF r a r on l a sp fl e o en o e e o o o t l w od e e Rl t o Ft n n L N C n oa mm I F v o si n i e e rr oo p i d r ut f a L G f f rr i td rr un s s et l c ol o owo ou s nh ff t a s s pn Ce o m m t n ot r e e po j so a a s sl ci a ci - - ar c c oC 'do E sC e e s o
sF o
a r nW t o x x r o t t eS T U S E E D R L S S L L R .
l
Docket No. 50-213 A06867 Attachment 2 Response to Informal Questions on the 95/95 DNBR Limit
.]
November,1987 1
_j
i.
Additional Information About the Determination of the DNBR Limit The experiments from the EPRI test set that were used for the determination of the DNBR limit were selected to include all those that closely match the Haddam Neck fuel design. While more experiments could have been modeled, the additional experiments would have been for different fuel types and would not significantly add to the calculation of the DNBR limit for the Haddam Neck model. In addition, the statistical evaluation factors into the limit the number of experiments used.
The experiments used for the determination of the DNBR all include a non-uniform axial power distribution. The power distribution shown in Figure 1 was used for experiments 113, 114, 121, 122, 122.1, 123 and 124.
The axial power distribution shown in Figure 2 was used for experiment 108.
Ve have performed a statistical evaluation to show that the experimental results come from the same distribution and therefore can be combined together in determining the liinit.
Several methods have been used to determine if the eight data sets belong to the same distribution. First, the F test was performed and the results indicate that the data do in fact come from the same distribution.
However, implicit in applying the F test, it is assumed that each data set comes from a normal distribution. To provide further verification, a distribution independent test was also used.
The Kruskal-Vallis test, a distribution independent test, was performed to determine if the eight data sets belong to the same distribution. The results indicate that the data do not come from the same distribution. To investigate a possible cause of this result, the initial conditions for each test were examined.
The test cases chosen to determine the 95/95 DNBR limit have the same number of rods, rod diameter, rod pitch, heated length, and grid design.
The only geometric difference between the tests is the grid spacing. All of the tests were run at the same facility. Thus, the grid spacing was examined using the Kruskal-Wallis test. The results indicate that the data sets for the two different grid spacing are from the same distribution. 4 Thus, the grid spacing is not the reason for the negative determination for I all 8 sets.
By taking different combinations of data sets, the Kruskal-Wallis test shows that the data can be divided into two roughly equal divisions. '
I l Using the limiting one of the two divisions the 95/95 limit is 1.2.
In looking at the parameters of the test, we believe that the data do come I from the same population as the F Test demonstrates. The Kruskal-Vallis test does not support this conclusion. However, the Kruskal-Vallis appears to be very sensitive to differences in means of the sets and is thus giving a negative determination. The only physical difference in the tests is the grid spacing and the Kruskal-Vallis tests indicates that this is not affecting the distribution. Thus, while the Kruskal-Wallis test results indicate a 95/95 DNBR limit of 1.2, we believe that the previously calculated limit of 1.16 is justified.
. ,,c In calculat.ing the limit, we have used the ratio of DNB flux predicted /DNB flux measured to determine the statistical limit. This was done to simplify the arithmetic in calculating the limit. The limit was calculated by combinin the 95. percent confidence levels for the mean and standard deviation 'g' . is calculation is comparable to that recommended in the Sandia' Report Using the Sandia Report, the 95/95 limit is given as x
+ 1.846S. Where x = 0.977, sample mean and S - 0.095, sample deviation.
The 95/95 limit based upon the Sandia report is 1.153. This,j slightly lower than the limit?of 1.16 reported in the Topical Report l
1 1
i l
l l
l l
- 1. Cener f t7m E. J. Mroczka to U.S. Nuclear Regulatory Commission "Haddam Neck Plant Revisions to Reanalysis of Non-LOCA Design Basis Accidents",
May 8, 1987 l
- 2. D. B. Owens " Factors.for One-Sided Tolerance Limit and for Variables Sampling Plans", Sandia Corporation Monograph, SCR-607, TIP-4500 (19th Edition), March 1963 I
- --_ __A
.-- 1 e *, 1 i
FIGURE 1 OUANTIFICATION OF DNBR LIMIT AXlAL POWER DISTRIBUTION #1 100-N l
90-N 00-N.
70- 1 I
i 60- j 0 s E
E 50 7
4 40-
/
/
l ,e.
/
20 0
10-
's , , . -r , ,
O.0 0.5 1.0 1.5 2.0 PERnt
r..
1 k
FIGURE 2 OUANTIFICATION OF DNBR LIMIT AXIAL POWER DISTRIBUTION #2 1001 f
90 x
ED- \
\
\
70-
'N
-'NA 60-0
- l E
50-7 4
n /
/
33-
/
29
/
10- '
/
/
0, -,, , _ , , . , -, ,
0.0 0.5 1.0 1.5 2.0 AXi4L FEAK
Docket No. 50-213 A06867 j i
1 i
i J
l i
Attachment 3
)
Revised Pages, NUSCO-151 l
J i
November,1987 l
t . .
operation, the feedwater flow increases to 183 percent of its full load value. Following the increase in feedwater flow, the core coolant inlet temperature decreases and core power increases slightly.
The net result is a negligible change in DNBR. The power excursion predicted in the excessive load increase analysis (Section 4.5) l I
bounds that of the increase in feedwater flow event. Therefore, the l
DNBR transient of Section 4.5 will bound that of the increase in l i
feedwater flow event.
Because feedwater flow is greater than steam flow, the SG 1evel increases. When the SG 1evel exceeds the high-high SG 1evel setpoint, an alarm is actuated in the control room. The operator can take manual control of the feedwater regulating valve and close it. This analysis assumes that the failure which caused the ,
regulating valve to open also prevents the operator from closing it. ;
l l Therefore, the operator is assumed to trip the reactor and the main feedwater pumps based on receiving the high-high level alarm.
Following reactor trip and feedwater pump trip the SG 1evel 1
decreases due to the steaming through steam bypass to condenser.
l l
l Figures 4.4-13 through 4.4-24 show the results for the excess feedwater flow transient from 40 percent power. Figure 4.4-13 through 4.4-18 show the results for 4-loop operation and Figure 4.4-19 through 4.4-24 show the results for 3-loop operation.
.l I
4.37
4.9 STEAM LINE BREAK EVENT Event Description A steam line break is defined as a rupture of a main steam line which results in a rapid increase in main steam flow and an increase in heat removal. The increased energy removal from the RCS causes a reduction of coolant temperature and pressure. In the presence of a negative moderator temperature coefficient, the RCS cooldown results l in the insertion of positive reactivity. A reactor trip will occur on high power or low RCS pressure. However, the continued cooldown will result in a reduction in shutdown margin. If the positive reactivity inserted exceeds the shutdown margin, it is possible that the core will become critical and return to power. If this were to l occur, the core would be ultimately shut down by the decrease in l
steam flow and ecoldown as the steam generator (SG) inventory is l
l depleted and by the boron addition delivered by the safety injection 1
system.
l 1
1 The analysis of a main steam line rupture is performed to demonstrate that the following criterion is satisfied:
- 1) Assuming the most reactive control rod stuck in its fully withdrawn position, with or without offsite power, and assuming the most limiting single failure, the radiological consequences do not exceed the requirements of 10CFR100.
4.80
_ _ _ _ _ _ _ _ _ _ 1
I l
The following functions provide the protection for a steam line rupture.
1
- 1. Safety injection system actuation from either of the following: I a, low pressurizer pressure
- b. high containment' pressure
- 2. Reactor trip may occur due to the over poa , reactor trip or on receipt of the Safety Injection Signal (SIS).
1
- 3. Isolation of the main feedwater lines will occur on safety injection actuation, via closure of the redundant main feedwater isolation valves. In addition, the feedwater control system will isolate feedwater by closing the feedwater regulating valves as the RCS average temperature'(Tavg) decreases to 535 F.
- 4. Trip of the fast acting steam line trip valves (designed to close in less than 10 seconds) on high steam flow in 2 out of 4 steam lines.
Additional protection is provided by the nonreturn (check) valves in the steam lines. The nonreturn valves in the main steam ~ lines prevent steam flow from the inta'ct steam generators 4.81
- - 4 v !
.. . 1
} l
)
from discharging through a break located upstream of a nonreturn i valve.
l l
Method of Analysis l
I The steam line break transient is analyzed in three stages. First, the RETRAN02 code is used to perform a detailed plant transient.
analysis. The transient calculation with RETRAN02 includes the modeling of the reactor coolant system, the steam generators, the main steam and feedwater system, and core kinetics including fuel and moderator temperature feedbacks. For the analysis of the four-loop cases, a split vessel RETRAN model was employed to account for nonuniform temperature distribution at the core inlet. This model separately represents the one quarter cold and three quarter hot quadrants of the reactor vessel and uses the coolant density of the cold quadrant of the core for the calculation of the moderator reactivity feedback. For the three loop cases, the limiting i
assumption is that only the RC pump in the affected loop remains {
)
running during the event. The other two pumps are assumed to trip, i l
1 since in the Haddam Neck design, only two pumps fast transfer to '
offsite power upon turbine trip. With only the RC pump of the affected loop running, the reactor inlet temperature distribution will be uniform. Thus, a split vessel model is not necessary for. I the limiting three loop cases. The results calculated by .
RETRAN02 include core inlet flow and temperature for'the affected and unaffected loops, RCS pressure and various components of the core reactivity. These parameters are then used to perform a 3-D reactor physics state point evaluation assuming the worst rod stuck 4,82 1
[
is out and all remaining rods inserted. The reactor physics methodology is described in Reference 7. The state point evaluation takes into account the nonuniform coolant inlet temperature distribution and is performed to verify the RETRAN02 point kinetics core reactivity balance and power level.
The resulting power distribution and power level, as calculated by the 3-D physics code, along with the RETRAN02 calculated system parameters are then used in VIPRE01 to calculate the fuel rod heat
)
flux and temperature distribution and the minimum DNBR for the worst state point in the transient. The MacBeth critical heat flux correlation is used to calculate DNBR when the thermodynamic conditions are outside the range of the W-3L correlation.
1 1
Analyses were performed for both 3-loop and 4-loop operation at full j power and hot zero power. Cases with and without offsite power available were evaluated.
The assumptions made in the analysis are as follows:
- 1. The steam line break analysis is performed with end of life shutdown margin, zero ppm boron concentration, and the most reactive control rod stuck in its fully withdrawn position.
- 2. For full power cases, the plant is initially at 102 percent power for 4-loop operation and 67 percent power for 3-loop 4.83
operation. For the hot zero power cases, the plant is assumed to be initially at 1 percent power. !
- 3. The most negative moderator temperature deficit is used to maximize the positive reactivity insertion due to the cooldown.
The moderator density feedback coefficient is varied with temperature and pressure.
l
- 4. The appropriate rodded Doppler defect is used to properly model the feedback during the return to power phase of the transient. ;
I
- 5. The break is assumed to be located outside the containment l l
upstream of the nonreturn valve. This break location delays. [
l the initiation of SIS, which is generated by low pressurizer pressure. If the break location were to be postulated inside the containment, the SIS would occur earlier on high containment pressure. Break flow is maximized by ignoring line losses. A i
double ended break is postulated. '
- 7. The single failure assumed is the failure of the affected SG feedwater regulating valve to close when the RCS average temperature (Tavg) drops to 535 F following turbine trip.
Feedwater is isolated by the feedwater isola'tion valves, which
'1 receive a closure signal on actuation of the SIS. Since Tavg j 4.84 .
_ __ _ A
T . ,
% l i
a drops to 535 F well before the SIS is generated, this assumption ]
I maximizes the feedwater supply and thus maximizes the RCS cooldown.
- 8. Perfect moisture separation in the steam generator is assumed.
The assumption leads to conservative results since, in fact, a considerable amount of water would be discharged. Water 1
carryover would reduce the magnitude of the temperature decrease 1 1
1 l
in the core.
Results i
The HZP cases with offsite power available provide the most li .tinc cooldown. This is because'at HZP, the primary side has the least 1
i stored energy and the SG secondary side contains the maximum inventory. '
Also, at HZP the feedwater temperature is at a minimum. All these factors maximize the cooldown, which maximizes the return to power.
However, the presence of decay heat in the HFP cases produces a slightly higher overall power level (i.e. , decay plus prompt power).
Therefore, the HFP cases will be used for DNBR and fuel centerline evaluations.
1 The loss of offsite power cases are less severe than the cases with l
l offsite power available. A loss of offsite power affects the transient in three ways. First, the main feedwater pumps trip on loss of offsite power, reducing the amount of wat'r e available for overcooling. Second, the coastdown of the reactor coolant pumps 4.85
1
., .. )
I k
1 (RCPs) reduces the RCS flow rate, which in turn, reduces heat transfer both in the steam generator and in the core. Third, the delay in starting the safety injection pumps is increased by 10 seconds due to the startup of the emergency diesel generators. The transient results show that termination of main feedwater flow to the faulted I steam generator and the reduction in heat transfer due to RCP trip l l
more than compensate for the delay in safety injection.
{
1 t
i l A sensitivity study was performed to determine the limiting assumption with respect to the time of reactor trip. The time of reactor trip i affects the transient in two ways. An earlier trip minimizes the heat input to the RCS, which maximizes the cooldown. However, a later trip time slows down the depressurization and thus delays high pressure injection of borated water. For the HFP cases, reactor trip on high power yielded the highest return to power.
As discussed previously, two reactor coolant pumps (RCPs) will l coastdown upon reactor trip. This will result in two RCPs remaining running in the 4-loop simulation, and one RCP in the 3-loop simulation.
To maximize the cooldown, the RCP in the faulted loop is assumed to l
be running throughout the transient. No credit is taken for operator action to trip the RCP in the faulted steam generator.
Because of the reduced pressure in the faulted steam generator, all l of the feedwater flow was conservatively assumed to be diverted to 1
the faulted generator. On actuation of the SIS, the-feedwater isolation valves will begin to close. The valves will ramp closed 4.86
. . . y
'. )
1
)
i in 70 seconds. No credit is taken for the reduction in feedwater flow until 65 seconds have. elapsed. After main feedwater flow is terminated, the SG 1evel begins to decrease since the auxiliary feedwater (AFW) flow is unable to keep up with loss of inventory through the break. The operator is assumed to terminate.AFW flow to the faulted generator at 10 minutes. !
Following a double-ended rupture of the main steam line, only one steam generator blows down completely. Thus, the remaining steam generators in operation are still-available for removal of decay heat after the initial blowdown is over. j Table 4.9-1 provides the sequence of events for the steam line break events initiated from 4-loop and 3-loop operation at HFP. The system response during 4-loop operation is shown in Figures 4.9-1 through 4.9-13. The system response during 3-loop operation is shown in Figures 4.9-14 through 4.9-25.
)
A DNB analysis is performed utilizing the power distribution
]
calculated by the 3-D physics code for the worst state point during the return to power portion of the transient. A large margin to DNB is predicted. The maximum fuel centerline temperature is less than the fuel melt limit of 4,780 F.
l 1
4.87
., l 1
- i, l
i Conclusion !
Results show that for the limiting steam line break case, no fuel !
rods exceed the DNB limit or experience fuel melt. The analyses show that no DNB or cladding perforation occurs for a steam line l I
rupture. Therefore, the radiological consequences of this event are !
i within the requirements of 10CFR100. l l
l l
I I
i 1
I I
f 4.87A
. . j
., e.
j
> l l
Table 4.9-1 i Sequence of Events for the Steam Line Break from Hot Full Power With 1
Offsite Power Available i
1 l
Event Time (Secs.) l 4-loop 3-loop i l
l Operation Operation.
i I
l Break Initiated 2.0 0.1 j
'l 8.0 Reactor Trip Initiated 7.42 2 RCPs Tripped 7.77 8.35 SIS Actuation 58.70 36.9 HPSI Pump Starts Injection 81.70 59.9 ,
l Pressurizer Empties 106.00 50.0 AFW Flow Initiated 128.7 95.1 Main Feedwater Flow Terminated 128.7 109.0 Time of Peak Core Power 106.0 105.0 i
AFW Flow Terminated (Operator Action) 600.0 600 l
l l
1' 1
s 1
4.88
1
)
'. l, i
STEAM LINE BREAK FROM HFP 4-LOOP OPERATION I
FIGURE 4. 9- 1 !
1 1
i i i i i i i i i
g a -
1 Ml I "I J
~
C@ '
p- : j l \
$ li E84 c.
l l
~
g 3
\
i I I i i _ __ L 1 .
c:C 60 lE; 2,; 320 CD M6; SEC E'.
TIMEISECONJ)
FIGURE 4.9-2 l i i i i ci i i i l
Moderator
{
3 e '
i Power Doppler E
7
>. n C
f Total
- w. ;
Boron 9
~
Total-Moderator n -l. -.L . L n i n s T 80 ISO 290 320 400 960 550 6*C TIME (SECONDI
STEAM 1,INF BRE AK FROM HFP 4 -1,00P OPER ATION FIGURE 4.9- 3 i
l I I I l l l l lc -
E :
m !
R
~
u \
g ..
I t:
W - -
c S.
R c
5 i I
r u I n i u ,
l d e: 16, .,. 3:0 c ,es se: e,:
TIMElSECCNDI l
l FIGURE 4.9 - 4 l
I g i i i i i i i eI j
o _x .
w l
[ t n' o -
2: 5 I o
gg w'
7 3/4 Hot _
cr 8
' 1/4 cold i i _t i e i. L k 00 160 240 320 900 960 550 E*O TIMEtSECOND) i
1
. 1 STEAM LINE BREAK FROM HFP 4 -LOOP OPERATION l l
FIGURE 4.9- 5
. B l y1 < > > i i i i !
Ts s
en d5 -
E a
E e
$kb
'~
u e
\_ a g f 3/4 llot 5 'T -
Y.. .
i f1/4 Cold i i i i i __2 _
cf 60 if; 2,0 320 900 93; 5f 6*:
TIME (SECOND)
FIGURE 4.9- 6 gl 3 i i i i i
~
~
i?
S
)
Ws- -
"l e
N
$8
~
v>
Ln Y
c g .
1 1 i i c0 80 150 240 320 900 9B0 550 E9; TIMEISECOND)
e .
l l
STEAM LINE BREAK FROM HFP 4-LOOP OPERATION FIGURE 4.9- 7 l
i 4 i i i l
U~
k=' t l'(
m 1
>~
o I c' !
i (E I ~
WF cr e l
u I O l H I e _
g_
i l i l i l I cc es is: 24: sza 9;; *ec 55: e,:
TIME (SECONJ)
FIGURE 4.9- 8 i i i
, g i . . .
g_
U s
C - -
.JR -
3 o : : : : : :
u.
- s -
'E x
o C
i i i i i i
_1 c0 80 160 290 320 900 980 5E0 6*D TIME (SECOND)
STEN 4 LINE BREAK FROM 'HFP 4-LOOP . OPERATION l
l FIGURE 4.9- 9 C
I I I I I I I c_
-o !
DE ..
- c. -
w l
[ j f E I !
mg* ~
- u. e-e i b I
\ \
1 i
- l C l 5: e ..L<.
i
- ?
O z !
a \ .
r . I E. c .
- r. s m
t a n t n I u ,
cc e: les {
1 4: 3:: 900 wec se: e4:
TIMEISECOND) i l
FIGURE 4.9- 10
' I I I i y }
( i 1
I IS ^ ~
i 1
e Ng
._I 1 w !
wJ (, l
~
e O
C I
~
Na w- ~
o r
c w-D ce -
l l
l
\
I f
_t__
i i i i c0 63 150 240 320 900 460 550 6<
T]ME (SEC) 1
i
. e j
STEAM LINE BREAK FROM HFP 4-LOOP OPEPsATION FIGURE 4.9- 11 lF i i i i i i i f-e o ,
t ,
in
\ E cz e -
d i~ i u L
.c r a
' Eld !
w ha. i
~
cc e i i
8 a: '
i I
_ _ L___ _1. r i __t ,_ i ,
cc 6: at 24: 3:: cc we: ss: e,; I TIME (SECONJ) l FIGURE 4.9 12 a I
i yl i i i i i i i s
J l-o, -
I v !
cr w
cc c r co
- l l
~
v, l
l l
' t i e L i i i cD 80 150 240 320 400 480 560 690 TIMElSECOND)
[ ._--___ -
1 n
1 STEAM LINE BREAK FROM HFP 4-LOOP OPERATION FIGURE 4.9-13 l i i e i i i i ;
I I' i w i'
((h! J S'
u.
! t
$g'
.c Lo (cold to hot) l l P' 'wer Plenum .
G f ~
~
l {
Q.
1 I
ja ,
, kUpperPlenum (hot to cold) i l E i -
55~
~
l
' I l l
8 d.__ >____..___.t. i i i e i _,
l
~ 5: 1c: r*: 3:: s:: *s: se: e,: >
TIME (SECC C 1 1
?
i l
l l
o -- 1
,i 4
STEAh! 1,1N1; tilt):5K ' Fit 0M HFP 1,00P OPEllAT10N
.1:1(;llitl; 4 . 9 - 14 I 3 4 1 1 6- I:
g -
I ca:
.p_a u
- q 1
l' ce y .)
{
gg -
g -
7
~
l 'I I i i i c0 80 160 Zw; 320 430 980 $50 0,0 TIME!SECOND) l l;
FIGURE 4.9-15 e i i i i e i i g)loderator
~
e Power Doppler t - --
C '"
5 7 _ _ . _ _ _ _ _ _
8 _
- ' Total . . .
Y f i l
' Boron I
~ ^
I I
. Total ' bloderator -
t i I f I
- ib ' SD 160 Zwo -. 320 400 M60 ' 56C 640.
TIME (SECOND) l
_ ______-_____u-_-___-___-
11 i
!ij 4 a n
MHrD .
'1 &
Tp t 'M W M> 7 O4 Z*1U T
- tJ ' Poo'O o"O Tt t 3 OZ m 9
- bCM= h . e7 w & e _.
o o
s - _ -
2 i 0 _
0 T nn~
g i2 _
S _
P _
( .
E Rn _
UoT _
Si s g
S E _
R .
P _
R 0mI 7r w i Ii R _
U .
S S
E R
P050 i
e p, En N 7, xg 4"
Y,, nn f
uNU tr pO",9 6 % *,
ywgCym h,Da My n
o , - _
s - _ _ _
o n
s f
)
F
( _
P _
M Eno :
T u '
E G
n R
E V0 s R03 '
E _
R _
O _
C 0 s 0 '
7 0 - - - -
8 Eu Ng uNO xg g { $
n j)v a
- 3r tr t
9O29
~ - _
ll ,l ,
f.
STEAM LINE lil:LAK Fit 0M liFP 3-LOOP OPERATION l'IGtlitt 4. 9 18 h i i ,
i i i i 1
I k-K -
O' w
(
m$ i - I
> 1 C
l
~
W*
C ;
O 1 i
l ic
! I I f t t t cf 80 160 2< 32; w0; ie t $5; Et,*
TIME;SECC'O)
FIGURE 4.9- 19 g i i i i i i i R ~
~
b W, -
w ._ -
J 2
w U
$o w-
~
~
m W
m l
l l
I l I i f i r cc B0 160 2*0 320 400 980 559 E40 TIMEISECOND)
. -t STEAM LINE BREAK FRO.'1 HFP 3-LOOP OPERATION FIGURL 4.9-20 l 4 i i i i i i g-U~ -
N s
cc c'[ r E~
d Cg a
C b
u.
g -
7 _
\
i i i , i , ,
cf e: 15: 2,- *--
w:: e: 55: c,;-
TIMEISECONO)
FIGURE 4.9-21
[r-- . > > i i i i 3
U Ms *
$_R~ _
~
O Y
Es -
x D
C
)
i i i i i i i c0 B0 300 290 320 400 3 4BD $50 6*0 TIMEISECOND)
(,
c STEAM' LINE BREAK FROM HFP 3-LOOP OPERAT1um FIGURE 4.9- 2 2 i
$ > > < i i e i E
~
O_$
E a
W eg -
gz _
c cr E I ES w v T -
z si d l r
5G~
p r, _
! ! I e 6 80 15 2*: 32: wa: in: 55: e,:
TIMEISECCh0!
FIGURE 4.9- 2 3 el s i i , , .) . ,
Em 1
I w _
tc O ..N w
~
- 2:
g in ~
r c
d p i v> -
to -
i
' I t i , ,
d ,
BD 160 2*C 320 900 weg MO 6*D TIME!SECOND)
! , t I
e
(
delay times. This allows more time for heat to be stored in the primary system prior to reactor trip.
l
- 8) No credit is taken for pressurizer sprays or PORVs to reduce or ]
i limit pressurizer pressure during the peak pressure transient.
l In order to minimize DNBR, the PORVs are assumed to be operable.
Except as discussed above, normal reactor control systems and engineered safety systems are not required'to function to show l l
acceptable results. No single active failure will prevent operation I of any system whose operation is required to show acceptable results. 3 i
i Analysis Results l
The system response to a total loss of steoa load from full power l
l 4-loop operation and full power 3-loop operation are provided. The i
sequence of events and plots of important parameters are provided in 1 p Table 4.11-1 and Figures 4.11-1 through-6.11-20, p i
a i
Figures 4.11-1 through 4.11-10 show the transient responses for the 1 4
1 4-loop case. No credit is taken for steam dump to condenser. The i reactor trir signal is generated on high pressurizer pressure at i'
6.8 seconds. Due to the use of the most positive moderator temperature coefficient (MTC) and Doppler feedbacks, the nuclear !
power remains f airly constant until the reactor is tripped. The minimum DNBR remains well above 1.3. The pressurizer safety valves are actuated, and maintain primary system pressure below 110 percent i
s 4.106 u__ _ _ _
, _ _ _ . . .u -- _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .- - _ _ . _ _ _ _ _ _ _ _ .
. . _ . . _ _ _ _ . _j
< 1 1
i i
In estimating the mass transfer from the RCS through the broken ]
j Ys tube, the following assumptions are made: 1
- 1. Reactor trip occurs manually at event initiation in order to maximize the radiological consequences. Loss of offsite power is assumed to occur coincident with reactor trip.
- 2. Following the reactor trip, two charging pumps are actuated and are assumed in the analyses to continue to deliver flow until ;
l 30 minutes after accident initiation. ]
I
- 3. After reactor trip, the break flow approaches equilibrium where incoming charging injection flow is balanced by outgoing break flow as shown on Figure 4.10-8. )a Charging flow is slightly higher because pressurizer level is increasing and the flow balance is on a volumetric not mass basis.
Break flow is assumed to persist for 30 minutes beyond initiation of the accident.
- 4. Steam generator pressure is controlled by the' steam g'enerator safety valves during the initial 30 minutes of the transient.
Safety valve blowdown is assumed to maximize the steam releases.
The effect of the blowdown is seen in the oscillatory system response.
- 5. The operator is assumed to terminate the auxiliary feedwater flow to the faulted steam generator at ten minutes.
4.92
t d
STEAM LINE BREAK FROM HFP 3-LOOP OPERATION FIGURE 4.9 24
! ' ' i . . t i I
{
l q
l -
l l
D u
l w ,
lE -
9 )
1 C
a L.?
g -
e c:
g -
1 i
i i i , , .- ,
Cf 2 16; 2i 320 w;; 460 550 64 TIMEISECCN01 l
l
\
l FIGURE 4.9- 25 l e
dl ' ' I i i 3 l
~
9 I
l D%
- O w
1 1 ~
l We m*
i e,
O
' I I I f j g d 83 160 240 320 NDD 980 563 690 TIMEtSECOND)
i STEN 4 GENERATOR TUBE RUPTURE FROM HFP FIGURE 4.10-7
' ' i i i , ,
gl, 8,
c a -
Intact SGs LA \ _
ce C .i e 4 Cr !
t>
z -
)
g - !
r C
! Wg -
- ! \ aulted F SG _
I
' ' i , , , ,
d 250 50g 750 1000 1250 1500 1750 2000 T]ME ISECONDS) i l
l l FIGURE 4.10-8 l
E i i i , , i ,
h E
l g Charging Flow
(
B e.:
c me l,,a.< ;lN :NJ::N.:N:s!::::::::: :JJ4;;. , ;.f:. -
l E Break Flow /
d l
i -)
L I i I I I f I c0 253 5% 75: 1 *? );n 19: 175: 2:'
TIME 15EC005)
_ _ _