ML20079A558
| ML20079A558 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 02/24/1995 |
| From: | Denton R BALTIMORE GAS & ELECTRIC CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| TAC-M90897, TAC-M90898, NUDOCS 9503010257 | |
| Download: ML20079A558 (12) | |
Text
_-_ _ - -_ ____
r y
Re:cT E. DmTos Baltimore Gas and Electric Company Vice President Calvert Cliffs Nuclear Power Plant Nuclear Energy 1650 Calvert Cliffs Parkway Lusby, Maryland 20657 f
410 586-2200 Ext.4455 Local 410 260-4455 Baltimore f
February 24,1995 U. S. Nuclear Regulatory Commission Washington,DC 20555 ATTENTION:
Document Control Desk
SUBJECT:
Calvert Cliffs Nuclear Power Plant Unit Nos.1 & 2; Docket Nos. 50-317 & 50-318 Response to Request for Additional Information Regarding the Use of the Convolution Techniaue for Main Steam Line Break Analysis
REFERENCES:
(a)
Letter from R. E. Denton (BGE) to Document Control Desk (NRC), dated November 1,1994, Request for Approval to Use Convolution Technique in Main Steam Line Break Analysis i
(b)
Letter from Daniel G. Mcdonald (NRC) to R. E. Denton (BGE), dated December 1,1994, Request for Additional Information Regardmg the Use of the Convolution Technique for Main Steam Line Break Analysis -
Calvert Cliffs Nuclear Power Plant, Unit Nos. I and 2 (TAC Nos. M90897 and M90898)
Baltimore Gas and Electric Company hereby responds to your request for additional information (Reference b) regarding our submittal requesting use of the convolution technique in Main Steam Line Break Analysis (Reference a). Our response to each of your questions is included in Attachment (1).
In Reference (a), we indicated that failure to apply the convolution technique to the current Unit 2 fuel cycle could result in small power reductions due to approachmg the Technical Specification peakmg limits.
Since that submittal, the measured power peaking on Unit 2 has reached its maximum for the current cycle and is now decreasing. The measured value approached within 1.7% of the limit (1.608 measured as J
compared to the limit of 1.635), but no power reductions were required. However, as we point out in our 2
attached response, the risk of power reductions still exists for future fuel cycles.
01002G l
f \\
9503010257 950224
/
PDR ADOCK 05000317 P
PDR-I
. _,Li' Document Control Desk February 24,1995 Page 2 Should you have questions regarding this matter, we will be pleased to discuss them with >ui.
Very truly yours,
^
b RED /BDM/ dim Attachment cc:
D. A. Brune, Esquire J. E. Silberg, Esquire L. B. Marsh, NRC D. G. Mcdonald, Jr., NRC T. T. Martin, NRC P, R. Wilson, NRC R. I. McLean, DNR J. H. Walter, PSC
t ATTACHMENT (1)
CONVOLUTION QUESTIONS
- 1. Q. Has the statistical convolution method been approved for other postulated events at Calvert Cliffs?
L A. He statistical convolution method has been approved for the determination of fuel failure due to.
Dpwre from Nucleate Boiling (DNB) for the Sheared Shaft / Seized Rotor event. De ~
application of the method of convolution to this event is the example used in App-~83-A of Topical Report CENPD-183-A, " Loss of Flow - CE Methods for Ims of Flow Analysis". His Topical Report was generically approved by the NRC for Seized Rotor Events in a Safety.
Evaluntmn, dated May 12,1982. The current Calvert Cliffs Seized Rotor Event analysis, UFSAR Section 14.16, utilizes the statistical convolution method.
- 2. Q. What are the site boundary doses for 5.6% fuel failure?
4 A. The 5.6% fuel failure results from a Pre-Trip Main Steam Line Break (MSLB) with an outside containment break location. The steam generator primary-to -d y leak rate is limited to :
100 gallons per day, per steam generator, as specified by Technical Specification 3.4.6.2.c. For this leak rate, the two-hour site boundary doses are 35 REM Hyroid and 0.2 REM Whole Body.
Previous calculations of the site boundary doses used the previous Technical Specification limit on primary-to-secondary leak rate of one gallon per minute,. At this leak rate, the two-hour site-boundary doses are 263 REM %yroid and 0.8 REM Whole Body.
Both these values are less than 10 CFR 100 limits for Limiting Fault accident conditions.
i
- 3. Q. What adjustments have been made to the Core Operating Limits Report (COLR) to maintain the MSLB predicted fuel damage fraction to less than 2%?
A. It was possible to mamtain fuel failure to less than 2% by restncting the allowed radial peaks in the COLR. He BASSS setpoints and the Departure from Nucleate Boiling Ratio (DNBR) excore l
Limiting Condition for Operation (LCO) tent for the operating cycles were based upon a radial peaking limit of 1.7. Imvering the radial peaking limit while not adjusting the DNBR LCOs has the effect of preserving the original thermal margin in the accident analysis, thereby limiting the fuel failure fraction.
To provide the additional thermal margin, the COLR radial peaking limits (F/ and F?) have been lowered from 1.70 to 1.66 for Unit 1 Cycle 12 and 1.635 for Unit 2 Cycle 10.
He affect of the COLR changes on the Pre-Trip MSLB cvent may be seen in Figures 3-1 in units of thermal margm to the Specified Acceptable Fuct Design Limits (SAFDL) and in Figure 3-2 in units of DNBR. At time 0 in the " Base Sdnaia*=" case (without the COLR change), the margm betww the most limiting pin in the core and the SAFDL is determmed by the Required Owim.a Margin. He Pre-Trip MSLB event steadily erodes this initial margin until the SAFDL is reached at about 8.5 seconds. De thermal margin continues to degrade, now in violation of the SAFDL, until the point of minimum DNBR occurs at about 11 seconds. %c thermal margin then begins to improve as the heat flux decreases following a reactor trip.
I
- u ATTACHMENT (1)
CONVOLUTION QUESTIONS lhe COLR change sets the radial peakmg limit to less than the value used to determine the Required Overpower Margm. This has the effect of creatmg additional margin between the DNBR LCO and the actual core condition. The upper line in Figure 3-1, labeled " Base Setpoints with Reduced Radial Peak," illustrates the nummum thermal margin in the core with the lower radial peakmg limit, while the lower line is the thermal margin at which the monitoring system assumes the peak pin is operatmg It is seen that the peak pin is actually several percent of thermal margin further from the SAFDL than the conditions assumed by the monitoring system.
'Ihe amount of fuel predicted to fail (fuel damage fraction) is dependent upon the extent of violahan of the SAFDL of the peak pin (and of the other pins in relation to the peak pin). It is seen that with the lowered radial peak, the extent of the SAFDL violation is reduced.
i Figure 3-2 presents similar information in DNBR units, rather than in thermal margin units.
This request for approval for the use of convolution has been made because of a concern that these lowered radial peak limits are close to the actual (measured) radial peak behavior of the core, and measured peaks may exceed these reduced limits in future cycles. As illustrated in Figure 3-3 for
]
Unit 2 Cycle 10, the lowered COLR peaking limit comes quite close to the predicted best estimate radial peakmg at around the 3/4 point in bumup.
Should measured peakmg exceed the lowered peaking limits in future cycles, a reduction in power.
will be necessary. 'Ihe closer the predicted radial peak is to the COLR limit, the greater the likelihood that a power reduction will be necessary.
- 4. Q. Why is the Low Steam Generator Pressure (LSGP) trip setpoint reached so much sooner in the new MSLB analysis (15.3 sec., Updated Final Safety Analysis Report [UFSAR] Revision 17) than in the previous analysis (33.9 sec., UFSAR, Revision 16)?
A. 'Ihe worst case for a Pre-Trip MSLB is found by the consideration of many competing dynamic effects, including the relative timing of the Low Steam Generator Pressure trip and the Power Level-High trip.
The effect of an increase in break size at a given moderator temperature coefficient (MTC) is a faster power increase and quicker depressurization of the affected steam generator.
For a given break size (and hence rate of heat extraction) the more negative the MTC, the greater the addition of positive reactivity and the greater rate of power increase. The power at which a j
Power level - High trip is reached becomes lower as the MTC W=en more negative since the rate of power increase overwhelms the trip delaying effects of temperature shadowing *.
1 l
l
- ' Temperature Shadowing" is the effect on excore neutron detectors due to moderator density changes l
2
~
ATTACHMENT (1)
CONVOLUTION QUESTIONS For a given break size, the peak power before a LSGP trip is seen to increase as the MTC becomes more negative. His is because the time to trip (depressunzation of the steam p. eor) is somewhat decoupled from core power. Herefore, for a given time until trip a higher power peak occurs with a more negative MTC.
An analysis is performed which exammes the effect of varymg the break sizes and MTC values.
Figures 4-1 and 4-2 illustrate a representative Pre-Trip MSLB response to two different break sizes, "A" and "B." For each break size, a more negative MTC results in an improved response of the Power Level - High trip seen as a lower peak power. De LSGP trip, however, occurs at higher and higher power levels because the time to depressurize the steam A.4ur and initate a LSGP trip is nearly constant for a given break size but the peak power increases as the MTC is more negative. %erefore, for a given break size, the linutmg condition is the highest of these peak i
powers that occur before the action of the first of the co...ye-g Reactor Protective System trips.
The limiting case for the transient as a whole is found by takmg the worst predicted response for i
cach ofthe individual break sizes.
As illustrated in Figures 4-1 and 4-2, the limiting power at time of trip vanes only slightly for various worst case combinations of break size and MTC. Since the intersection of the worst case combinations for break size and MTC occurs in a region of the curves that is relatively flat, cycle-by cycle differences in core parameters can significantly change the limiting break size /MTC point (although these differences have only a minor effect on the limiting power at time of trip).
He outside containment break scenario wluch previously resulted in the most adverse fuel failure 2
and was included in Revision 16 of the UFSAR was found to be 0.65 Ft and resulted in the LSGP value of 640 PSIA being reached in 33.9 Seconds. De current analysis demonstrates that a 1.0 2
Ft break size is more limiting for current reloads and results in a more rapid depressurizatimt of the steam generators. Thus the LSGP is reached in only 15.3 seconds in recent analyses.
Although the trip times for these two break sizes differed by 15 seconds, the peak power reached was very similar.
- 5. Q. Recent experiments by the French on reactivity insertion transients indicate that fuel failures occur in high burnup fuel at very low levels of reactivity insertion. What are the peak CAUGM levels reached when DNB occurs during Pre-Trip and Return-to-Power? What is the peak CAUGM level reached in high burnup fuel (over 40 GWD/MTU exposure)? What are the transient amplitudes (CAUGM)?
A. This request for approval of the use of convolution is for application solely in the Pre-trip MSLB analysis. Convolution is not employed to predict the number of pins in DNB in the Return-to-Power MSLB scenario. For the Return-to-Power MSLB analysis, DNB is assumed when the pin experiences a MacBeth DNBR less than 1.30. The conservative assumption used in the evalcation of fuel failure for this event is that all pins which violate this SAFDL experience DNB and are predicted to fail.
3 i
m-
- m.
-m
---u
m m
t.
N ATTACHMENT (1)
CONVOLUTION QUESTIONS Since this application for approval is limited to the Pre-trip MSLB event,'only this event is addressed below.
The fuel temperatures and deposited energy values presented in Table 5-1 are based upon long term operation at the power levels indicated. For the Pre-Trip MSLB, the DNBR crisis is of a few seconds duration being brought on by the coast down of the reactor coolant pumps following reactor trip and the assumed loss of all A.C. power. As reactor power is actually decreasing during the time of minimum DNBR, the steady state assumption is in all likelihood conservative even though the brief reduction in fuel to moderator heat transfer coefficient associated with DNB in the hot pin was ignored. However, the high bumup fuel ofinterest does not expericece DNB, so there is no reduction in the heat transfer coefficient for this fuel.
He Pre-Trip MSLB results in an increase of core power from an initial value of 100% to a peak ofjust under 140%. Changes in power, fuel temperature and fuel enthalpy are given in Table 5-1 for a high bumup rod (defined here as a thrice burned fuel rod with a rod average burnup of over 40 GWD/MTU, operating at a radial peak of unity) and for the hot rod. De use of a unity radial peak for the high burnup rod is conservative relative to the anticipated rod powers at end of cycle for both Unit 1 Cycle 12 and Unit 2 Cycle 10.
TABLE 5-1 Initial Conditions At Time of Peak Power Relative Fuel Peak Fuel Radial Enthalpy Relative Enthalpy Temp'F Temp'F Peak (Cal /Gm)
Power (Cal /Gm) l.0 924 31.8 1.4 1093 38.7 Rod Peak uel 1.7 1243 44.9 2.38 1650 62.2 As illustrated in this table, high bumup fuel normally operates with an enthalpy of about 31.8 cal /gm. The enthalpy of the high burnup fuel slowly increases during the Pre-Trip MSLB event by about 7 cal /gm to a value of 38.7 cal /gm. Note that the high bumup fuel is not predicted to experience DNB.
Although we have responded to the NRC's requests regarding the deposited energy values for the MSLB event, it should be emphasized that the Nuclear Energy Institute has established an Issues Task Force to assess the generic applicability and validity of the limited, foreign experiir a.1 data which suggests that the fuel failure criteria for reactivity insertion accidents may be lower for high burnup fuel. At present, we do not have any reason to believe that the high burnup fuel in Calvert Cliffs would fait during an MSLB cvent.
4 1
-r-.
ATTACHMENT (1)
CONVOLUTION QUESTIONS j
Compared to the rapid insertion of reactivity of a Control Element Assembly (CEA) EW.
measured in a fraction of a second, the Pre-Trip MSLB goes from initial power to peak power in a comparatively long 17 seconds 'Ihe failure =-:-'=!= r. present in the recent French @.
- . is not a=~ *~I for the relatively slow power excursions during the Pre-Trip MSLB event.
Specifically, it is our underd== ding that while the failure -h=aism in the foreign test is still uncettain, it is generally beliemi that the speed of the reactivity insertion rate is a key determinant as to whether failure would be predicted. Design Basis Events with the fastest reactivity insertion rates would represent the most limiting case and would be potentially the most adversely affected by the recent French ew.. e ts. In Calvert Cliffs, like other pressurized water reactors, the fastest reactivity insertion event is the CEA ejection accident.
In contrast to the CEA ejection accident where enthalpy additions occur in a fraction of a second, the Pre-Trip MSLB cvent is characterized by a relatively slow power increase over a period of many seconds It is further our under*=ading that the Nuclear Energy Institute Issues Task Force evaluation of the safety significance of the high bumup fuel behavior during the most severe reactivity insertion accidents has been submitted to the NRC. This industry assessment ='-T-:='=:1y i
came to the same conclusions reached by the NRC's preliminary safety assessment. In particular, the safety assessment concludes that this new data is not an i. ; =E- - safety significance and that there will be no detnmental impact to the public health and safety.
l 1
6
N r
4 ;
c.
i t'
,(
s FIGURE 3-1 170 160 -
n Wg 150 -
1 6&
1 140 -
1 x
G i
m I
BASE SETPOINTS WITH I" -
REDUCED RADIAL PEAK j
a B
E 120 -
M 110 -
BASE SETPOINTS w
100 -
y S
E-111.,
0 2
4 6
8 to 12 14 16 18 20 TIME (SECONOS) i l
e w
f';k#
I w-FIGURE 3-2 H
2.1 2-1.9 -
.1.8 -
1.7 -
I*' ~
ee.
BASE SETPOINTS WITH i
j REDUCED RADIAL PEAK
)
1.5 -
i.4 -
BASE SETPOINTS 1.3 -
i 0
2 4
6 8
10 12 14 16 18 20-TIME (SECONDS) 1
__.___.--u___-__..____.-_~___m_-
llI
/
FIGURE 3-3 UNIT 1 CYCLE 10 RADIAL PEAK VS BURNUP
<45 ACTIVE INCORE DETECTORS 12 1
1.7 L a a a a a a a a a : a a a a a a a a ;
g 1.6 -
n.
I a
.5 13 -
1.4 -
IJ i
i e
i i
i i
o 0.2 0.4 0.6 o,8 t
FRACTION OF CYCLE LENGTH l
l o - ORIGINAL UNIT 2 CYCLE 10 FR LIMIT A - REDUCED COLR FR LIMIT
+ - PREDICTED FXY VALUES
-l 0 = PREDICTED FR VALUES J
4 rf n'
FIGURE 4-1 SLB RESPONSE TO BREAK SIZE "A" iso LOL' STEAM GENERATOR
[
PRESURE TRIP RESPONSE e
^
g 140 -
d WORST CASE FOR IM -
BREAK "A" 5
m y
a."
HIGH POWER TRIP E2
RESPONSE
g 120 -
g 5
4 110 -
,o.-
i I
i i
- 0.5
- 0.7
- o.s
- t,1 MTC, x 10-4 Ap/*F l
.. h' -
i 5
l FIGURE 4-2 SLB RESPONSE TO BREAK SIZE "B" 150 LOW STEAM GENERATOR PRESSURE TRIP RESPONSE 8-14o -
I Ei i
W WORST CASE FOR r
33 _
BREAX "B"
--a a:
E HIGH POWER TRIP g
RESPONSE
E
[
120 -
5 i
g 1 to -
n.
too
- 0.5
- 0.7
- o.9
- 1. i l
MTC, x 10-4 Ap/*F i
1 1
- -