ML20210G100
| ML20210G100 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 08/31/1986 |
| From: | Heberle G, Huegel D, Osborne M WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML20210G079 | List: |
| References | |
| NUDOCS 8609250335 | |
| Download: ML20210G100 (20) | |
Text
i SAFETY EVALUATION FOR OPERATION OF CATAWBA UNITS 1 & 2 WITH A POSITIVE MODERATOR COEFFICIENT REVISION 1 D. S. Huegel G. H. Heberle August 1986 Mhhh Approved:
M. P. Osborne, Manager Transient Analysis II WESTINGHOUSE ELECTRIC CORPORATION -
Water Reactor Divisions P. O. Box 355 Pittsburgh, Pennsylvania 15230 8609250335 860916 PDR ADOCK 05000413 P PDR 94350:1D/072186 1
CONTENTS
- 1. INTRODUCTION 3 4
- 2. ACCIDENT EVALUATIONS 4
I. Introduction II. Transients Not Affected By a Positive Moderator Coefficient 5 A. RCCA Misoperation 5 B. Startup of an Inactive Reactor Coolant Loop 5 C. Excessive Heat Removal Due to Feedwater -
System Malfunctions 6
- D. Excessive Load Increase 6 E. Spurious Actuation of Safety Injection 6 F. Main Steam Line Depressurization/ Rupture of a Main Steam Pipe 7 G. Loss of Coolant Accident (LOCA) 7 III. Transients Sensitive to a Positive Moderator 7
Coefficient
! A. Baron Dilution 7 B. Control Rod Bank Withdrawal From a Suberitical Condition 8 C. Uncontrolled Control Rod Bank Withdrawal at Power 9 I
D. Loss of Reactor Coolant Flow 10 Locked Rotor 11 E.
F. Turbine Trip 12-Loss of Normal Feedwater/ Loss of Offsite Power 14 G.
H. Rupture of a Main Feedwater Pipe 15 I. Control Rod Ejection 16 J. Accidental Depressurization of the Reactor Coolant System 17 19
- 3. CONCLUSIONS Figure 1 20 21 Table 1 943sO:lo/0721ss 2
- SECTION 1 INTRODUCTION Safety analyses and evaluations have been performed to support the proposed Technical Specification change for Catawba Units 1 & 2 ,
which would allow a positive moderator temperature coefficient to ,
exist during power operation. The results of the analysis, which are presented in the following section, show that the proposed change can be accommodated with margin to applicable FSAR safety limits.
The present Catawba Technical Specifications require the moderator -
temperature coefficient (MTC) to be +0 pcm/*F* at all times while the reactor is critical. The proposed Technical Specification change would allow a +7 pcm/*F MTC below 70 percent of rated power, ramping down to 0 pcm/*F at 100 percent power. This MTC is diagrammed in Figure'1.
A positive coefficient at reduced power levels results in a significant increase in fuel cycle flexibility, but has only a minor effect on the safety analysis of the accident events presented in the FSAR. A power-level dependent MTC was chosen to minimize the effect of the specification on postulated accidents at l high power levels. Moreover, as the power level is raised, the average core water temperature becomes higher as allowed by the f
programmed average temperature for the plant, tending to bring the moderator coefficient more negative. Also, the boron concentration can be reduced as xenon builds into the core. Thus, there is less need to allow a positive coefficient as full power is approached.
As fuel burnup is achieved, boron is further reduced and the moderator coefficient will become negative over the entire operating power range.
-5
- 1 pcm = 10 Ak/k 94350:1o/o721as 3
SECTION 2 ACCIDENT EVALUATIONS I. Introduction The impact of a positive moderator temperature coefficient on the -
accident analyses presented in Chapter 15 of the Catawba Units 1 &
2 Final Safety Analysis Report (FSAR) has been assessed. Those incidents which were found to be sensitive to positive or minimum moderator coefficients were reanalyzed. In general, these incidents are limited to transients which cause reactor coolant temperature to increase. The analyses presented herein were based _
on a +7 pcm/*F moderator temperature coefficient, which was assumed to remain constant for variations in temperature.
The control rod ejection and rod withdrawal from suberitical analyses were based on a coefficient which was at least +7 pcm/*F at zero power nominal average temperature, and which became less positive for higher temperatures. This was necessary since the TWINKLE computer code, on which the analyses are based, is a diffusion-theory code rather than a point-kinetics approximation i
and the moderator temperature feedback cannot be artificially held constant with temperature. For all accidents which were reanalyzed, the assumption of a positive moderator temperature coefficient existing at full power is conservative since as diagranned in Figure 1, the proposed Technical Specification requires that the coefficient be linearly ramped to zero above 70 percent power.
In general, reanalysis was based on the identical analysis methods, computer codes, and assumptions employed in the FSAR; any exceptions are noted in the discussion of each incident. Accidents not reanalyzed included those resulting in excessive heat removal ,
from the reactor coolant system for which a large negative moderator coefficient is conservative. Table 1 gives a list of 94350.1o/072186 4
accidents presented in the Catawba Units 1 & 2 FSAR, and denotes those events reanalyzed for a positive coefficient. The following sections provide discussions for each of the FSAR events.
II. Transients Not Affected By a Positive Moderator Coefficient The following transients were not reanalyzed since they either
- result in a reduction in reactor coolant system temperature, and are therefore sensitive to a negative moderator temperature coefficient, or are otherwise not affected by a positive moderator temperature coefficient.
A. RCCA Misoperation Only the RCCA drop case presented in Section 15.4.3 of the FSAR is potentially affected by a positive moderator temperature coefficient. Use of a positive coefficient in the analysis would result in a larger reduction in core power level following the RCCA drop, thereby increasing the probability of a reactor trip. For the return to power automatic rod control case, a positive coefficient would result in a small increase in the power overshoot. Since the limiting conditions for this accident are at or near 100 percent power where the moderator temperature coefficient must be close to zero or negative, this accident is unaffected by the proposed Technical Specification and thus the analysis was not repeated.
B. Startup of an Inactive Reactor Coolant Loop An inadvertent startup of an idle reactor coolant loop at an incorrect temperature results in a decrease in core average temperature. As the most negative values of moderator reactivity-coefficient produce the greatest reactivity addition, the most limiting case is represanted by'the analysis reported in the FSAR, Section 15.4.4.
94350:1D/072186 5 t._ . _ _ _ . _ _ . _ _ . _ . . _ _ _ _ . _ _ . _ _ _ _ . _ _ _ _ . _ _ _ _ _ _ _ _ __ _ _ _ _ .
C. Excessive Heat Removal Due to Feedwater System Malfunctions The addition of excessive feedwater or the reduction of feedwater temperature are excessive heat removal incidents, and are consequently most sensitive to a negative moderator temperature coefficient. Results presented in Section 15.'1.1 .
and 15.1.2 of the FSAR, based on a negative coefficient, represent the limiting case. Therefore, this incident was not reanalyzed.
D. Excessive Load Increase An excessive load increase event, in which the steam load i
exceeds the core power, results in a decrease in reactor coolant system temperature. With the reactor in manual control, the analysis presented in Section 15.1.3 of the FSAR shows that the limiting case is with a large negative moderator coefficient. If the reactor is in automatic control, the control rods are withdrawn to increase power and restore the average temperature to the programmed value. The analysis of this case in the FSAR show that the minimum DNBR is not sensitive to moderator temperature coefficient. Therefore, the results presented in the FSAR are still applicable to this
- incident.
E. Spurious Actuation of Safety injection Analysis of a spurious actuation of safety injection at power is presented in Section 15.5.1 of the FSAR. This transient results in a decrease in average coolant temperature and core power and the results are not sensitive to moderator 1 temperature coefficient. Therefore,ihisincidentwasnot l
l reanalyzed with a positive moderator coefficient. .
l l
l l
9435o 10/072186 6
F. Main Steam Line Depressurization/ Rupture of a Main Steam Pipe l
Since the steam line depressurization or rupture of a main steam pipe is a temperature reduction transient, minimum core shutdown margin is associated with a strong negative moderator temperature coefficient. The worst conditions for a steamline -
break are therefore those analyzed in the FSAR (Section 15.1.4 and 15.1.5).
G. Loss of Coolant Accident (LOCA)
The loss of coolant accident (Section 15.6.5 of the FSAR)'is -
analyzed to determine the core heatup consequences caused by a rupture of.the reactor coolant. system boundary. The event results in a depressurization of the RCS and a reactor shutdown at the beginning of the transient. This accident was not reanalyzed since the Technical Specification requirement that the temperature coefficient be zero or negative at 100 percent power ensures that the previous analysis basis for this event is not affected.
III. Transients Sensitive to a Positive Moderator Coefficient A. Boron Dilution l
As stated in Section 15.4.6 of the FSAR, a boron dilution incident cannot occur during refueling due to administrative controls which isolate the RCS from potential sources of diluted water. If a boron dilution incident occurs during cold shutdown, hot standby, or startup, the operator must take action to terminate the dilution before the reactor returns critical. Therefore, since a return to criticality is prevented by the operator, the value of the moderator -
coefficient has no effect during a boron dilution incident in these operating modes. The reactivity addition due to a boron 94350:1o/072186 7
dilution at power will cause an increase in power and reactor coolant system temperature. Due to the temperature increase, a
~
positive moderator coefficient would add additional reactivity and increase the severity of the transient. With the reactor in automatic control, however, the rod insertion alarms provide the operator with adequate time to terminate the dilution ,
before shutdown margin is lest. A boron dilution incident with the reactor in manual control is no more severe than a rod withdrawal at power, which is discussed in Section III.C, and therefore this case was not specifically analyzed. Following reactor trip, the amount of time available before shutdown margin is lost is not affected by the moderator coefficient. _
~
B. Control Rod Bank Withdrawal From a Suberitical Condition Introduction A control rod assembly bank withdrawal incident when the reactor is subcritical results in an uncontrolled addition of reactivity leading to a power excursion (Section 15.4.1 of the FSAR). The nuclear power response is characterized by a very fast rise terminated by the reactivity feedback of the negative fuel temperature coefficient. The power excursion causes a heatup of the moderator and fuel. The reactivity addition due to a positive moderator coefficient results in increases in peak heat flux and peak fuel and clad temperatures.
Method of Analysis
! The analysis was performed in the FSAR for a reactivity insertion rate of 75 x 10 -5 ak/sec. This reactivity insertion rate was used in this analysis and is greater than that for the simultaneous withdrawal of the combination of the two sequential control banks having the greatest combined worth at maximum speed (45 inches / minute). The analysis used a 94350.1D/072186 8 t
moderator temperature coefficient more conservative than a +7 pcm/*F for all appropriate temperature values. The computer codes, initial conditions, and other assumptions are noted in l the FSAR page markups attached in the Appendix.
Results and Conclusions .
Reanalysis of this event assuming a 75 pcm/sec insertion rate, coupled with a positive moderator temperature coefficient of +7 pcm/*F, yields a peak heat flux which does not exceed the nominal full power value. In the event of a RCCA bank withdrawal accident from a suberitical condition, the core and -
the RCS are not adversely affected, since the combination of
~
thermal power and the coolant temperature result in a ONBR greater than the limit value and thus, no fuel or clad damage is predicted. Therefore, the conclusions presented in the FSAR remain valid.
C. Uncontrolled Control Rod Bank Withdrawal at Power Introduction 1
An uncontrolled control rod bank withdrawal at power produces a mismatch in steam flow and core power, resulting in an increase in reactor coolant temperature. A positive moderator coefficient would increase the power mismatch resulting in a faster heatup of the reactor coolant. However, this effect is offset by the fact that the faster heatup and reactivity l addition result in an earlier reactor trip from either overtemperature delta-T or high neutron flux. A discussion of
~
this incident is presented in Section 15.4.2 of the FSAR.
l I
94350:1D/0721ss 9 f
i
.~- yr_.. _ _ _ , - _ _ - . . , ,--,,,.y
,,., .-,-__y -_,._._,,xy,,, _ , , . . . _ . . . , .
Method of Analysis The transient was reanalyzed employing the same digital computer code and assumptions regarding instrumentation and setpoint errors used for the FSAR. This transient was analyzed at 100, 60 and 10 percent power with a positive moderator ,
coefficient. A constant moderator coefficient of +7 pcm/*F was used in the analysis.
Results and Conclusions For each initial power level the full range of reactivity _
insertion rates was reanalyzed. The limiting case for DNB margin remains above the limit DNBR value. These results demonstrate that the conclusions presented in the FSAR are still valid. That is, the core and reactor coolant system are not adversely affected since nuclear flux and overtemperature AT trips prevent the core minimum DNB ratio from falling below the limit valua for this incident.
D. Loss of Reactor Coolant Flow Introduction The loss of flow events presented in FSAR Sections 15.3.1 and 15.3.2 were reanalyzed to determine the effect of a +7 pcm/*F moderator temperature coefficient on the nuclear power transient and the resultant minimum DNBR reached during the incident. The effect on the nuclear power transient would be limited to the initial stages of the incident during which reactor coolant temperature increases; this increase is terminated shortly after reactor trip.
e4ssa_to/oniss 10 l
i
Method of Analysis Analysis methods and assumptions used in the reevaluation were consistent with those employed in the FSAR. The digital computer codes used to calculate the flow coastdown and resulting system transient were the same as those used to _
perform the analysis described in the FSAR. The analysis assumed a constant moderator coefficient of +7 perv'F.
Results and Conclusions Results of the analysis show that the minimum DNBR remains _
above the limit value for these transients.
Therefore, the conclusions of the FSAR analyses remain valid.
E. Locked Rotor Introduction The case presented in the FSAR (Section 15.3.3) for this transient was reanalyzed. Following a locked rotor incident, reactor coolant system temperature rises until shortly after reactor trip. A positive moderator coefficient will not affect the time to DNB since DNB is conservatively assumed to occur at the beginning of the incident. The transient was reanalyzed, however, due to the effect on the nuclear power transient and thus on the peak reactor coolant system pressure and fuel and clad temperatures.
In addition, the Locked Rotor Rods-in-DNB transient was reanalyzed due to the effect of a posi.tive moderator coefficient on the nuclear power transient and therefore margin
, to DNB. -
94350.1D/072186 11 4
l Method of Analysis i The digital computer codes used in the reanalysis to evaluate -
the pressure transient and thermal transient were the same as those used in the FSAR. The analysis employed a constant moderator coefficient of +7 pcm/*F. Other assumptions used -
were consistent with those employed in the FSAR.
Results and Conclusions Analysis of the locked rotor incident with a +7 pcm/*F moderator temperature coefficient shows that the peak reactor _
coolant system pressure remains below that which would cause stresses to exceed the faulted condition stress limits. The peak clad temperature for the hot spot during the worst transient remains much less than 2700*F and the amount of Zirconium - water reaction is small. Therefore, the conclusions presented in the FSAR remain valid.
In addition, analysis of the Locked Rotor Rods-in-DNB shows that the conclusions presented in the FSAR remain valid.
F. Turbine Trip Introduction Two cases, analyzed for both beginning and end of life conditions, are presented in Sections 15.2.3 of the FSAR:
- 1. Full credit is taken for the effect of the pressurizer spray and the pressurizer power operated relief valves.
Safety valves are also available.
94350.1o/072186 12
- 2. No credit is taken for the effect of the pressurizer spray or power operated relief valves. Safety valves are operable.
Although the moderator temperature coefficient will be negative _
at end of life, all cases were repeated. The result of a loss .
of load is a~ core power level which momentarily exceeds the secondary system power extraction causing an increase in core water temperature. The consequences of the reactivity addition due to a positive moderator coefficient are increases in both peak nuclear power and pressurizer pressure.
Method of Analysis A constant moderator temperature coefficient of +7 pcm/*F was assumed. The method of analysis and assumptions used were otherwise in accordance with those presented in the FSAR.
1 Results and Conclusions The beginning of life case system transient response to a total loss of load from 100 percent power assuming pressurizer relief and spray valves was calculated. Peak pressurizer pressure reaches 2518 psia following a reactor trip on high pressurizer pressure. A m'inimum DNBR well above the limit value is reached shortly af ter reactor trip.
The transient response to a loss of load assuming no credit for pressure control was also calculated. Peak pressurizer pressure reaches 2563 psia following reactor trip on high pressurizer pressure. The DNBR increases throughout the transient.
94350:1D/072186 13
The analysis of the beginning of life cases demonstrates that the integrity of the core and the reactor cool. ant system pressure boundary during a loss of load or turbine trip transient will not be impacted by a +7 pcm/*F moderator reactivity coefficient since the minimum DNB ratio remains well above the limit value, and the peak reactor coolant pressure is _
less than 110 percent of the design value of 2500 psia.
Therefore, the conclusions presented in the FSAR remain valid.
G. Loss of Normal Feedwater/ Loss of Offsite Power Introduction _
The loss of normal feedwater and loss of offsite power accidents (Sections 15.2.7 and 15.2.6 of the FSAR) are analyzed to demonstrate the ability of the secondary system auxiliary feedwater to remove decay heat from the reactor coolant system. Following initiation of the event the reactor coolant temperature rises prior to reactor trip due to the reduced heat transfer in the steam generators. Thus, the assumption of a positive moderator temperature coefficient results in a i reactivity insertion and resultant increase in core power prior
! to reactor trip. This is turn increases the amount of heat that must be removed following reactor trip, resulting in a more severe transient Method of Analysis l
A constant moderator temperature coefficient of +7 pcm/*F was l
l assumed. A conservative core residual heat generation based on the 1979 version of ANS-5.1 was used. The method of analysis
! and assumptions used were otherwise in accordance with those presented in the FSAR. .
s o so:to/0721ss 14
1 l
Results and Conclusions The transient results show that the capacity of the auxiliary-feedwater system is adequate to provide sufficient heat removal from the RCS. The pressurizer does not fill with water, .
assuring that the integrity of the core is not adversely -
affected. For the case without offsite power, the results 3
verify the natural circulation capacity of the RCS to provide !
sufficient heat removal capability to prevent fuel or clad damage following reactor coolant pump coastdown.
H. Rupture of a Main Feedwater Pipe Introduction The main feedwater pipe rupture accident (Section 15.2.8 of the FSAR) is analyzed to demonstrate the ability of the secondary system auxiliary feedwater to remove decay heat from the reactor coolant system. Following initiation of the event the reactor coolant temperature rises prior to reactor trip due to i the reduced heat transfer in the steam generators. Thus, the assumption of a positive moderator temperature coefficient results in a reactivity insertion and resultant increase in core power prior to reactor trip. This is turn increases the
- amount of heat that must be removed following reactor trip, resulting in a more severe transient.
Method of Analysis A constant moderator temperature coefficient of +7 pcm/*F was i assumed. The method of analysis and assumptians used were otherwise in accordance with those presented in the FSAR.
e 94350 1D/072186 15
Results and Conclusions The transient results show that the capacity of the auxiliary-feedwater system is adequate to provide sufficient heat removal from the RCS to prevent overpressurization of the RCS or core ,
uncovery. The reactor coolant remains subcooled, assuring that ,
the core remains covered with water. For the case without offsite power, the results verify the natural circulation capacity of the RCS to provide sufficient heat removal capability to prevent RCS overpressurization and fuel or clad damage following reactor coolant pump coastdown.
I. Control Rod Ejection Introduction The rod ejection transient is analyzed at full power and hot standby for both beginning and and of life conditions in the FSAR. Since the moderator temperature coefficient is negative at end of life, only the beginning of life cases are affected by a positive MTC. The high nuclear power levels and hot spot fuel temperatures resulting from a rod ejection are increased by a positive moderator coefficient. A discussion of this transient is presented in Section 15.4.8 of the FSAR.
I t
f Method of Analysis The digital computer codes for analysis of the nuclear power
' transient and hot spot heat transfer are the same as those used in the FSAR. The ejected rod worths and transient peaking factors assumed are conservative with respect to the actual i
calculated values for current fuel cycles. The analysis used a 3
94350:10/072186 16
_, _ _ , _ _ _ _ _ _ _ _ . _ . _ _ _ , _.__.______m._ _ _ _ _ _ _ _ _ . _ - - __
moderator temperature coefficient more conservative than a +7 pcm/*F for all appropriate temperature values and power j levels. This is a conservative assumption since the moderator coefficient actually decreases to zero from 70 percent to 100 percent power. .
Results and Conclusions A peak clad average temperature of 2696'F was reached in the beginning of life hot zero power case. Maximum fuel temperatures were associated with the full power case.
j Although the peak hot spot fuel centerline temperature for this _
transient exceeded the melting point, melting was restricted to less than the innermost 10 percent of the pellet.
As fuel and clad temperatures do not exceed the fuel and clad limits specified in the FSAR, there is no danger of sudden fuel dispersal into the coolant, or consequential damage to the primary coolant loop.
J. Accidental Depressurization of the Reactor Coolant System Introduction An accidental 'depressurization of the reactor coolant system results from an inadvertent opening of a pressurizer relief or safety valve (FSAR Section 15.6.1). Since a safety valve is l
sized to relieve at a much greater flow rate than a relief valve and will therefore allow a much more rapid depressuriza-i tion, the case of a safety valve opening is analyzed. This i
situation initially results in a rapidly decreasing reactor coolant system pressure until the hot leg saturation pressure l
i H350.1D/072386 17
is reached. With a negative moderator density coefficient (positive MTC), the decrease in pressure results in an increase in core reactivity because the coolant density decreases as the pressure decreases. The most limiting case therefore assumes the reactor is in manual control, such that the increase in core reactivity causes nuclear power and average coolant .
temperature to increase until the reactor trips. Therefore, the consequence of the reactivity addition due to the +7 pcm/*F moderator coefficient is an increase in peak nuclear power.
Method of Analysis The method of analysis and assumptions used were the same as those presented in the FSAR except for the following:
- 1. A constant moderator temperature coefficient of +7 pcm/*F was assumed.
- 2. The reactor was assumed to operate in the manual mode of i operation to prevent rod insertion prior to reactor trip.
- 3. A least negative Doppler-only power coefficient of reactivity was assumed to augment any power increase due to '
moderator reactivity.
Results and Conclusions I
The system transient response to the inadvertent opening of a pressurizer safety valve with the reactor in manual rod control was calculated. The reactor trips on low pressurizer pressure and the minimum DNBR occurs shortly after control rods begin to drop into the core.
94350.1D/o72186 18
/
. o The analysis demonstrates that the integrity of the core during a reactor coolant system depressurization transient is not adversely affected by a positive moderator reactivity coefficient since the minimum DNB ratio remains above the limit value. Therefore, the conclusions presented in the FSAR remain valid. .
}
e 6
9 i
O 94350.1D/0721ss 19
. o SECTION 3 CONCLUSIONS To assess the effect on accident analysis of operstion of Catawba Units l 1 & 2 with a positive moderator temperature coefficient of +7 pcm/*F safety analyses of transients sensitive to a minimum or positive moderator coefficient were performed. These transients included control rod assembly withdrawal from suberitical, control rod assembly withdrawal at power, loss of reactor coolant flow, locked rotor, turbine trip, loss of normal feedwater, rupture of a main feedwater pipe, control rod ejection, and RCS depressurization. This study indicates '
that a +7 pcm/*F moderator coefficient does not result in the violation of safety limits for any of the transients analyzed.
Except as noted, the analyses employed a constant moderator coefficient of +7 pcm/*F, independent of power level. The results of this study are conservative for the accidents investigated at full power, since the proposed Technical Specification diagrammed in Figure 1 requires that the coefficient linearly decrease from +7 pcm/*F to O pcm/*F from 70 percent to 100 percent of rated power.
9435o 1D/o72186 20 l _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ,_ _ _ _ _ _ _ _ _