ML20205H390
| ML20205H390 | |
| Person / Time | |
|---|---|
| Site: | Mcguire, McGuire  |
| Issue date: | 10/29/1985 |
| From: | WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| To: | |
| Shared Package | |
| ML19344C184 | List: |
| References | |
| NUDOCS 8511150123 | |
| Download: ML20205H390 (32) | |
Text
.
p.
/ WESTINGHOUSE PROPRIETARY CLASS 3 i
RTO BYPASS ELIMINATION LICENSING REPORT FOR McGUIRE UNITS 1 & 2 9
Westinghouse Electric Corporation Pittsburgh, PA B511130123 851029 PDR ADOCK 05000369 P pop
a
'I WESTINGHOUSE PROPRIETARY CLASS 3 4
TA8LE OF CONTENTS Section Page 1.0 Introduction 2 1.1 Historical 8ackground 2 1.2 Mechanical Modifications 3 1.3 Electrical Modifications 8 2.0 Testing 10 2.1 Response Time Test 10 2.2 Streaming Test 11 3.0 Uncertainty Considerations 12 3.1 Calorimetric Flow Measurement Uncertainty 12
. 3.2 Hot Leg Temperature Streaming Uncertainty 14 4.0 Safety Evaluation 17 4.1 Response Time 17 4.2 RTO Uncertainty 18 4.3 Instrumentation and Control Safety Evaluation 23 l
4.4 Mechanical Safety Evaluation 27 l 5.0 Control System Evaluation 29 6.0 Conclusions 30 3522e:1d/092385 1
I WESTINGHOUSE PROPRIETARY CLASS 3 l
1.0 INTRODUCTION
Westinghouse Electric Corporation has been contracted by Duke Power to remove the existing RT9 (Resistance Temperature Detector) Bypass System and replace the hot leg and cold leg temperature measurement with fast-response RTOs installed in the reactor coolant loop piping. This report is submitted in support of continued operation of the McGuire Units with the new RTD System installed.
1.1 HISTORICAL BACKGROUND Prior to 1968, PWR designs had been based on the assumption that the hot leg temperature was uniform across the pipe. Therefore, placement of the temperature instruments was not considered to be a factor affecting the accuracy of the measurement. The hot leg temperature was unasured with direct-immersion RTDs extending a short distance into the pipe at one location. By the late 1960s, as a result of accumulated operating experience at several plants, the following problems associated with direct immersion RTO's were identified.
o Temperature streaming conditions; the incomplete mixing of the coolant leaving regions of the reactor core at different temperatures produces significant temperature gradients within the pipe.
r o cooling and draining of the loops before the RTD's could be replaced. >
The RTO bypass system was designed to resolve these problems; however, operating plant experience has now shown that operation with the RTD bypass loops has caused some new problems:
o Plant shutdowns caused by excessive primary leakage through valves, flanges, etc., or by interruptions of bypass flow due to valve stem failure.
3522e:1d/091885 2 1
'I WESTINGHOUSE PROPRIETARY CLASS 3 o Increased radiation exposure due to maintenance on the bypass line and to crud traps which increase radiation exposure throughout the loop compartments.
The proposed temperature measurement modification has been developed in tesponse to both sets of problems encountered in the past. Specifically:
o Removal of the bypass lines eliminates the components which have been a major source of plant outages as well as Occupational Radiation Exposure (ORE).
o Three thermowell-mounted hot leg RTDs provide an average measurement to account for the temperature streaming phenomenon.
o Use of.thermowells permits RTO replacement without plant draindown.
Following is a detailed description of the effort required to perform this modification.
1.2 MECHANICAL MODIFICATIONS 4
I The individual loop temperature signals required for input to the Reactor j Control and Protection System will be obtained using RTDs installed in each reactor coolant loop.
1 1.2.1 Hot Leo a) The hot leg temperature nwasurement on each loop will Le accomplished with 1 three fast response narrow range RTDs mounted in thermowells. To accomplish the sampling function of the RTD bypass manifold system and eliminate the need for additional hot leg piping penetrations, the thennowells will be located within the three existing RTO bypass manifold
- scoops. A hole will be drilled through the end of each scoop so that water will flow in through the existing holes in the leading edge of the i
3522e:ld/091885 3
WESTINGHOUSE PROPRIETARY CLASS 3 scoop, past the RTD, and out throagh the new hole [ ]+"'*. These three RTDs will measure the hot leg temperature which is used to calculate the reactor coolant loop differential temperature (AT) and average i
temperature (T,yg).
I b) This modification will not affact the single wide range RTD currently installed near the entrance of each steam generator. This RTO will continue to provide the hot leg temperature used to monitor reactor coolant temperature during startup, shutdown, and post accident conditions.
1.2.2 Cold Lea a) One fast response, narrow range, RTO will be located in each cold leg at the discharge of the reactor coolant pump (as replacements for the cold leg RTD's located in the bypass manifold). Temperature streaming in the cold leg is not a concern due to the mixing action of the RCP. For this reason, only one RTD is required. This RTD will measure the cold leg temperature which is used to calculate reactor coolant loop AT and T,g,g . The existing cold leg RTD bypass penetration nozzle will be
! modified [ ]**** to accept the RTD thermowell.
b) This modification will not affeet the single wide range RTD in each cold leg currently installed at the discharge of the reactor coolant pump.
t This RTD will continue to provide the cold leg temperature used to monitor i
reactor coolant temperature during startup, shutdown, and post accident i
conditions.
c) A new penetration will also be made to each cold leg to accept an additional well mounted narrow range RTD, for use as an installed spare.
This will give the new modification a tolerance for RTD failures
- equivalent to the bypass loops. A new cold leg boss will be added t
[ ]** to accept the RTD thernowell.
l.
3522e:1d/091885 4
WESTINGHOUSE PROPRIETARY CLASS 3 1 --
.JuL THIS FIGURE IS CONSIDE8ED P90PRIETARY IN ITS ENTIPITY Figure 1. Not Leg RTO Scoop Modification for Fast-Response RTO Installation
WESTINGHOUSE PROPRIETARY CLASS 3
- J,6 THIS FIGURE IS CONSIDE8ED P'nPRIETARY IN ITS ENTIPITY
. Figure 2. Cold Leg Pipe Nozzle Modification Fast-Response RTD Installation
to WESTINGHOUSE PROPRIETARY CLASS 3
~ JsL THIS FIGURE IS CONSinEDED PonPRIETARY IN ITS ENTIPITY Figure 3. Additional toss for Cold Leg Fast-Rtsponse RTO Installation 7
4 WESTINGHOUSE PROPRIETARY CLASS 3 1.2.3 Crossover Lea j The RTD bypass manifold return line will be capped at the nozzle on the crossover leg.
1.3 ELECTRICAL MODIFICATIONS 1.3.1 Function
[ ]+8'C shows a block diagram of the modified electronics. The hot leg RTD measurements (three per loop) will be electronically averaged in the process protection system. The averaged T hot signal will then be input to the appropriate protection function. This will be accomplished by additions to the existing 7300 equipment.
1.3.2 Qualification Equipment seismic and environmental qualification will be to IEEE standards 344-1975 and 323-1974, respectively, as described in WCAP-8587, Rev. 5
" Methodology for Qualifying Westinghouse WRD Supplied NSSS Safety Related Electrical Equipment". ,
1.3.3 RTD Failures Existing control board AT and T,yg indicators and alarms will provide the means of identifying RTD failures. The spare cold leg RTD provides sufficient spare capacity to accommodate a single cold leg RTD failure per loop. Failure of a hot leg RTD will require manual action to defeat the failed signal, and a manual rescaling of the electronics to average the remaining signals
( )+a c, 3522e:1d/092385 8
WESTINGHOUSE PROPRIETARY CLASS 3 J,b c THIS FIGURE IS CONSIMEDED P90P#IETARY IN ITS ENTIPITY l .
i Figure 4. RTO Averaging Block Diagram, Typical for Each of 4 Channels 9
WESTINGHOUSE PROPRIETARY CLASS 3 2.0 TESTING
~
There are two specific tests which have been performed to support the installation of the fast-response RTDs in the reactor coolant piping: a response time test and a hot leg temperature streaming test.
2.1 RESPONSE TIME TEST Westinghouse has performed an RTD Response Time Test at its Forest Hills Test Facility. This test placed a fast response RTD, manufactured by RdF Corporation, inside a scoop, within a thermowell, which modelled actual in-plant installation. The flow conditions were adjusted to equal the high velocity Reactor Coolant System flows of approximately [ ]+a . The RTD's response time is determined based on a comparison of the RTD with previously calibrated and response time characterized thernmcouples. l Sixty-five test runs were made at various flow rates while gathering data on 2 RTDs. The test results demonstrated a mean response time for the RTD, thermowell and scoop of [ ]+a,b,c seconds. Table 2.1 provides a comparison of the present RTD Bypass System response time and how it would differ with the new system in place.
TABLE 2.1 RESPONSE TIME PARAMETERS FOR RCS TEMPERATURE MEASUREMENT RTD Bvoass System Proposed System RTD Bypass Piping and Thernal Lag (sec) +a,c +a,c RTD Response Time (sec)
RTD Filter Time Constant (sec)
Electronics Delay (sec)
Total Response Time (sec) 10.0 10.0 3522e:Id/092385 10
l WESTINGHOUSE PROPRIETARY CLASS 3
, l Based upon the response time parameters in Table 2.1, it becomes evident that the McGuire Units can accommodate the new response time with no_ further plant testing required.
2.2 STREAMING TEST Past testing at Westinghouse PWRs has established that temperature stratification exists in the hot leg pipe with a temperature gradient from top to bottom of ( ) '. A test program was implemented at McGuire Unit 1 to confirm the temperature streaming magnitude and stability with measurements of the RTD bypass branch line temperatures on two adjacent reactor coolant loops. Specifically, it was intended to determine the g magnitude of the differences between branch line temperatures, confim the short-term and long-term stability of the temperature streaming patterns and evaluate the impact on the indicated temperature if only 2 of the 3 branch line temperatures are used to detemine an average temperature. This plant specific data will be used in conjunction with data taken from other Westinghouse designed plants to determine an appropriate temperature error for use in the safety analysis and calorimetric flow calculations. Section 3 will discuss the specifics of these uncertainty considerations.
, The McGuire Unit 1 test data has been reduced and characterized to answer the three objectives of the test program. First, it is conservative to state that the streaming pattern [ ] #. Steady state data taken at 100% power for a period of four weeks indicates that the streaming pattern [ ] 'C. In other words, the temperature gradient (
)+b,c.e . This is inferred by [
)+b c.e observed between branch lines. Since the [
3+b,c.e into the RTD averaging circuit if a hot leg RTD fails and only 2 RTDs are used to obtain an average hot leg temperature. The operator can review temperatures recorded prior to the RTD failure and determine an (
)+b c.e into the 'two RTD" average to obtain the "three RTD" 3522e:1d/092385 11
WESTINGHOUSE PROPRIETARY CLASS 3 expected reading. This significantly reduces the error introduced by a failed RTD. ,
The McGuire Unit 1 data'also supports previous calculations of streaming errors determined from tests at other Westinghouse plants. The McGuire Unit 1 data is consistent with the upper bound temperature gradients that characterize the previous data. There were no new discoveries, but the data did add a dimension previous tests did not have. The McGuire Unit 1 test sampled temperatures from the pipe interior while all previous tests investigated temperature gradients at the pipe surface. The pipe internal data has greatly strengthened the assumptions and inferences made with previous test data.
The streaming test and response time test have both provided valuable information needed to support the design of the fast-response RTDs installed in the reactor coolant piping.
3.0 UNCERTAINTY CONSIDERATIONS This new method of hot leg temperature measurement has been analyzed to determine if it will have an impact on two uncertainties included in the ,
Safety Analysis: Calorimetric Flow Measurement Uncertainty and Hot Leg Temperature Streaming Uncertainty.
3.1 CALORIMETRIC FLOW MEASUREMENT UNCERTAINTY Reactor coolant flow is verified with a calorimetric measurement performed after the retarn to full power operation following a refueling shutdown. Two of the most important parameters for the calorimetric measurement are the narrow range hot leg and cold leg coolant temperatures. The accuracy of the RTDs has, therefore, a major impact on the accuracy of the flow measurement.
The current flos measurement uncertainty for McGuire for the sum of the four loop flows including elbow tap, is about 1.7 percent flow (not including 0.1 percent for feedwater venturi fouling allowance). However, the uncertainty 3522e:1d/092385 12
, WESTINGHOUSE PROPRIETARY CLASS 3 for the fast response RTD to be used in the proposed modification is
[ ]****, somewhat higher than for the RTO being replaced. However, the ,
impact of the larger uncertainty is reduced by incorporating an improved calorimetric procedure. The new procedure is intended for application to the measurements to be made at McGuire with the proposed temperature measurement system, but the procedure is applicable to and intended for application at all Pressurized Water Reactor nuclear plants.
In the previous calorimetric flow measurement procedure, the enthalpy rise across the reactor vessel is determined by measuring the hot and cold leg temperature, and the uncertainty is based on the errors in measuring these absolute temperatures. A more accurate measurement can be obtained by
[
- j+b c.e, There are three uncertainty components for the measurement performed with this procedure
- (
l
)+a,c ,
As a result, the overall flow measurement uncertainty is about the same as the existing value of 1.7 percent flow (not including 0.1 percent for feedwater venturi fouling allowance).
i 3522e:1d/092385 13 4
3 .
g 9 ,
WESTINGHOUSE PROPRIETARY CLASS 3
. r TherevisedprocedurefortheflowmeasurementatMcGuireisoutlinedbe1N:~
- a. [
/-
j+b c.e ,
- b. [
. N p
)+b,c.e ,
- c. [
)+b.c.e ,
- d. [
j+b.c.e ,
3.2 HOT LEG TEMPERATURE STREAMING UNCERTAINTi Thesafetyanalysesincorporateanuncertaidt[toaccountforthedifference between the actual hot leg temperature and the measured hot leg temperature caused by the incomplete mixing of coolant leaving regions of the reactor core at different temperatures'. This temperature streaming uncertainty is based on an analysis of test data from other Westinghouse plants, and on calculations
~ <.
3522e:1d/092385 14
WESTINGHOUSE PROPRIETARY CLASS 3 to evaluate the impact on temperature measurement accuracy of numerous possible temperature distributions within the hot leg pipe. The test data has shown that the circumferential temperature variation is no more than [
]* ', and that the inferred temperature gradient within the pipe is limited to about
[ ] #. The calculations for numerous temperature distributions have shown that, even with margins applied to the observed temperature gradients, the three-point temperature measurement (scoops or thernowell RTDs) is very effective in determining the average hot leg temperature. The most recent calculations for the thernowell RTD system have established an overall streaming uncertainty of [ ]4 '' for a hot leg measurement. Of this total, [
]+b,c.e . The overall temperature streaming uncertainty applied to the calorimetric flow measurement is only slightly larger than the uncertainty used in previous analyses.
The new method of measuring hot leg temperatures, with the thermowell RTDs located within the three scoops, is at least as effective as the existing RTD bypass system, [
~
]+a,c . Although the new method measures temperature at one point within the thermowell, compared to the five sample points in a 5-inch span of the scoop measurement, the thermowell measurement point is opposite the center hole of the scoop and therefore measures the equivalent of the average scoop sample if a linear radial temperature gradient exists in the pipe. The thermowell measurement may have a small error relative to the scoop measurement if the temperature gradient over the 5-inch scoop span is nonlinear. Assuming that the maximum inferred temperature ,
gradient of [ ]+b,c.e exists from the center to the end of the scoop, the difference between the thermowell and scoop measurement is limited j
] 'C. Since three RTD measurements are averaged, and the to [
nonlinearities at each scoop are random, the effect of this error on the hot i leg temperature measurement is limited to [ ]*'C. On the other 3522e:1d/092385 15 l
l l
WESTINGHOUSE PROPRIETARY CLASS 3 hand, imbalanced scoop flows can introduce temperature measurement uncertainties of up to [ -
, ]**'C . ,
In all cases, the flow imbalance uncertainty will equal or exceed the
[ ]+b,c.e sampling uncertainty for the thermowell RTDs, so the new measurement system tends to be a more accurate measurement with respect to streaming uncertainties.
Temperature streaming measurements from the test at McGuire Unit 1 have been obtained for a period of 4 weeks. The measurements confirm the [
f 3+b c.e ,
Over the 4-week testing period, there have been only minor variations of less than [ ]+b,c.e in the temperature differentials between scoops, and smaller variations in the average value of the temperature differentials. [ ,
)+b,c.e ,
, Provisions were made in the RTD electronics for operation with only two hot leg RTDs in service. The two-RTD measurement will be biased to correct for the difference compared with the three-RTD average. Based on the McGuire Unit 1 test data, the bias would be limited to between [ ]+b,c.e ,
Data comparisons show that the magnitude of this bias varied less than
# over the test period.
[ ]+
t 4
i L
3522e:1d/092385 16 i . - ._ . . . . - . -_ . . - - _ - _ - _ _ _ _ , -. ._ -. .
WESTINGHOUSE PROPRIETARY CLASS 3 '
'\. ,
4.0 SAFETY EVALUATION ,
4.1 RESPONSE TIME The primary impact of the RTD bypass elimination on the safety analyses is the increased response time associated with the fast responsa thermowell RTO system. Originally, the overall response time cf the McGuire RTD bypass system assumed in the accident analyses was approximately [ ]+" # , I before modification of the RTD filter time constant (see Table 4.1). This l compares with the overall response time of approximately 10 seconds for the fast response thermowell RTD combination (including a [
]+a,e time constant), as described in Section 2.1.
Longer RTD response times result in longer delays from the time when the fluid conditions in the RCS require an Overtemperature Delta-T or Overpower Delta-T reactor trip until a trip signal is actually generated. Thus those transients that rely on these protective functions must be evaluated for the longer response time. For McGuire, subsequent analyses were perforned as part of the most recent reload safety evaluations to allow for an increase of the lag time constants in the delta-T and T,yg channels from [ )+a,e, seconds, resulting in an overall response tiae used in the analyses of 10 seconds. ,
Since the revised time constant bounds the overall response time of the proposed new system, the affected analyses (RCCA Bank Withdrawal at Power and Steamline Rupture at-power) remain valid. The response time parameters are l
', summarized in Table 4.1.
i 3522e:1d/092385 17
+ .
! WESTINGHOUSE PROPRIETARY CLASS 3 4
, TABLE 4.1
- RESPONSE TIME PARAMETERS FOR RCS TEMPERATURE MEASUREMENT ,
I i RTD Bypass System Fast Response Thermowell l l
i RTD electronic filter time constant l Electronics delay >
l Total Response Time 6.0 see 10.0 sec 10.0 sec 4.2 RTG UNCERTAINTY f
The proposed fast response thensowell RTD system will make use of RTDs '
I manufactured by the RdF Corporation with a total uncertainty of [ ]**'C assumed for the analyses. Currently, the McGuire units are supplied with Rosemount RTDs, each of which has a smaller uncertainty value. Therefore, the increased uncertainty associated with the RdF RTDs must be accounted for in the safety analyses. Since three RTDs are used to measure hot leg .
! . temperature, the impact of the larger RdF RTD error is reduced.
i 4
- The FSAR analyses make explicit allowances for instrumentation errors for some ;
l of the reactor protection system tetpoints. In addition, allowances are made
- for the initial everage reactor coolant system (RCS) temperature, pressure and power as described in FSAR Section 15.0. These allowances are made explicitly to the initial conditions for non-DNB events; for DNB events these allowances are statistically combined into the design limit DNBR value, consistent with the Improved Thermal Design Procedure. ,
1 The following protection and control system parameters are affected by the assuned narrow range RTD accuracy: Overtemperature Delta-T Reactor Trip (OTDT), Overpower Delta-T Reactor Trip (OPDT), Low RCS Flow Reactor Trip, RCS ;
3522e:1d/092385 18 l i i I
- en m,,, - - , ,-- ~ , -
WESTINGHOUSE PROPRIETARY CLASS 3 average temperature measurements used for control board indication and input to the rod control system, and the calculated value of the RCS flow measurement uncertainty. System uncertainty calculations were nerformed for the noted parameters to assess the effects of the increase in the RTD error associated with the change from Rosemount to RdF manufactured RTDs and of the change from use of the bypass system to an averaged value of the three hot leg RTDs.
The results of the system uncertainty calculations verify that sufficient allowance has been made in the reactor protection system setpoints to account for the increased RTD orror. Therefore, the current values of the nominal setpoints noted above as defined by the McGuire Technical Specifications remain valid.
The Chapter 15 non-LOCA safety analyses were perforned assuming that, at steady state full power, the average RCS temperature was equal to the nominal value plus 5.5'F for non-DNB Events. For DNB events, at steady state full power, the average RCS temperature is assumed to be at its nominal value; the uncertainties are convoluted into the design limit DNBR value. The system uncertainty calculations verify that there is sufficient margin in the current error allowances to account for the estimated increase in the rod controller error. This does not affect any of the zero power transients presented in Chapter 15 of the FSAR since rod controller accuracy is not applicable.
The current licensing basis of the McGuire FSAR Chapter 15 safety analyses assumes that the initial RCS flow rate is 382000 gpm for non-DNB transients, and 388880 gpm for the DNB transients. The Loss of Coolant Accident is analyzed at an RCS flow rate of 377000 gpm. The current Technical Specification RCS flow measurement uncertainty of 1.7 percent (not including 0.1 percent for feedwater venturi fouling allowance) continues to be a conservative allowance for the increased RTD error if the revised flow calorimetric procedure outlined in Section 3.1 is employed. Therefore, no changes to the Technical Specifications for RCS flow or flow measurement uncertainty are required.
3522e:1d/092385 19
WESTINGHOUSE PROPRIETARY CLASS 3 9
4 However, conservative analyses have been performed to account for postulated uncertainties. For example, the effects of an increased allowa.nce for RCS flow measurement uncertainty have been conservatively evaluated. For DNB events, a flow measurement uncertainty of 2.2 percent (not including 0.1 percent for feedwater venturi fouling allowance) has been incorporated into the calculation of the design limit DNBR, consistent with use of the Improved Thernal Design Procedure. Margin remains to the safety analysis limit DNBR
} value. Therefore, reanalysis of the DNB transients which employ the ITDP is not required. The non-DNB accidents (or accidents for which DNB is not the
! only concern) have been evaluated or reanalyzed for a conservatively low RCS l flow rate of 377000 gpm, consistent with the LOCA analysis assumption. The results of the analyses and evaluations are as follows:
1 Uncontrolled RCCA Bank Withdrawal from a Subcritical Condition A control rod assembly 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. However, since the power rise is rapid and is followed by an .
. immediate reactor trip, the moderator temperature rise is small. Thus, nuclear power response is primarily a function of the Doppler coefficient.
This event has been reanalyzed with a value of reactor coolant flow consistent with the full flow of 377000 gpm (only two reactor coolant pumps are assumed to be operating). The analysis shows that for a reactivity insertion rate of 75 x 10 -5 delta-K/sec, the peak hot spot heat flux achieved is 191.3 percent of nominal with a resultant peak fuel average temperature of 2364*F, and a peak clad temperature of 729'F. Thus, maximum fuel and clad temperatures are still significantly below fuel melt (4800*F) and zirconium-H O2 reaction (1800*F) limits. A DNB analysis has shown that the DNB design basis is met. >
) Therefore, all safety criteria for this event are satisfied.
3522e:1d/092385 20
WESTINGHOUSE PROPRIETARY CLASS 3 Boron Dilution The results of the boron dilution transient will remain unchanged for all
, modes of operation with a reduction in reactor coolant flow. The maximum dilution flow rate, RCS active volumes, and RCS boron concentrations are not l
impacted by a reduction in flow. Since these parameters determine the amount of time available to the operator to terminate the dilution event, the results
- presented in the FSAR remain unchanged.
j Loss of Load i The loss of load accident is presented in Section 15.2.2 of the FSAR and can occur from either a loss of external electrical load or a turbine trip. The result of a loss of load is an increase in core power which exceeds the secondary system power extraction, thus causing an increase in core water
. temperature. This event has been reanalyzed with a reduced RCS flow of 377000 gpm. The analysis shows a peak pressurizer pressure of 2572 psia. The pressurizer does not fill, and the maximum pressures are within the design limits. Therefore, operation at reduced flow will not violate safety limits following a loss of load accident.
. Loss of Normal Feedwater/ Station Blackout This transient is analyzed to demonstrate that the peak RCS pressure does not exceed allowable limits and that the core remains covered with water. These
, criteria are assured by applying the more stringent requirement that the pressurizer must not be filled with water and that no water relief occurs through the pressurizer relief or safety valves. The effect of reducing initial core flow results in an initial more rapid heatup of the RCS. The resultant coolant density decrease increases the volume of water in the pressurizer. These transients have been reanalyzed with the reduced flow assumption. The results show margin to filling the pressurizer. Therefore, all safety criteria are met for these events.
i 3522e:1d/092385 21
WESTINGHOUSE PROPRIETARY CLASS 3 Steamline Break The steamline break transient is analyzed at hot zero power, end-of-life conditions for the fol, lowing cases: ,
Inadvertent opening of a steam dump, safety, or relief valve (Section 15.1.4 of the FSAR)
Main steam pipe ruptures with and without offsite power available (Section 15.1.5 of the FSAR)
A steamline break results in a rapid depressurization of the steam generators and primary side cooldown. This causes a large reactivity insertion due to the presence of a negative moderator temperature coefficient. A reduction in reactor coolant flow will result in a reduction in heat transfer from the fuel to the coolant. Therefore, the reactivity insertion and return to power in the double-ended rupture case for reduced flow conditions would be less limiting than the cases presented in the FSAR. For the double-ended rupture case, the time of safety injection actuation is unaffected by reduced coolant flow. This, coupled with a slower return to power, would result in a reduction in peak average power from the FSAR results. The inadvertent
- opening of a steam dump safety or relief valve is bounded by the double-ended rupture. Since the return to power is less severe and an evaluation has confirmed that the DNB design basis is met for an RCS flow of 377000 gpm, the conclusions presented in the previous analysis are still valid.
Rupture of a Main Feedwater Line l
This transient is analyzed to demonstrate that the peak RCS pressure does not exceed allowable limits and that the core remains covered with water. These criteria are assured by applying the more stringent requirement that bulk voiding does not occur at the outlet of the core. The effect of reducing initial core flow results in an initial more rapid heatup of the reactor coolant system (RCS). This transient has been reanalyzed with the reduced l
3522e:1d/092385 22
WESTINGHOUSE PROPRIETARY CLASS 3 flow assumption. The results show margin to hot leg saturation. Therefore, all safety criteria are met for the event. .
Locked Rotor Following a locked rotor, reactor coolant system temperature rises until shortly after reactor trip. A reduction in RCS flow will not affect the time to DNB since DNB is conservatively assumed to occur at the beginning of the
, transient. This event has been reanalyzed to determine the effects of a
, reduced RCS flow of 377000 gpm. The peak RCS pressure calculated in the analysis is 2613 psia, which is still significantly below the pressure at which vessel stress limits are exceeded. The peak clad temperature calculated is 2009'F, well below the limit of 2700*F. Therefore, all safety criteria are met for this event.
Control Rod Ejection The rod ejection transient is analyzed at full power and hot standby for both beginning and end-of-life conditions (Sections 15.4.8 of the FSAR). A reduction in core flow results in a reduction in heat transfer to the coolant, which will increase peak clad and fuel temperatures and peak fuel stored
, energy. The rod ejection cases have been reanalyzed with the reduced RCS flow assumption. Results are summarized on Table 4.2. All safety criteria are met for the event (i.e. Peak Clad Temperature < 2700*F, Fuel Melt < 10%, Fuel Stored Energy < 200 cal /gm).
4.3 INSTRUMENTATION AND CONTROL (I&C) SAFETY EVALUATION The RTD Bypass Elimination modification for McGuire Units 1 and 2 does not functionally change the AT/T,, protection channels. The implementation of the fast response RTDs in the reactor coolant piping will change the inputs into the AT/T,yg Protection Sets I, II, III, and IV as follows:
3522e:1d/092385 23
. _ . - - - - -----------w -
-- - - . , , . - - , _ , ,, -.- . - . . , ,- _. .-, , w - - - -- ----
o WESTINGHOUSE PROPRIETARY CLASS 3 l l
- 1. The Narrow Range (NR) cold leg RTD in the cold leg manifold will be replaced with a fast response NR RTD well mounted in the RCP pump discharge pipe. The signal from this fast response NR RTD will perform the same function as the existing RTD T s gnal.
cold
- 2. The NR hot leg RTD in the bypass manifold will be replaced with 3 fast response NR RTDs well mounted in hot leg scoops that are averaged in an analog-based signal selector. The signal from this average Thot circuit obtained from these 3 NR T will perform the same function as the hot existing RTD T hot 518""1"
- 3. The modification will include means to manually reject failed T hot signals. Identification of failed signals will be by the same means as before the modifications, i.e., existing control board alarms and 4 indications.
! 4. Signal process and the added circuitry to the Protection Set racks will be accomplished by additions to the 7300 racks using 7300 technology. When one T signal is defeated, the electronics will allow a bias to be hot manually added to a 2-RTD average Thot (as pposed to a 3-RTD average Thot) in order to obtain a value comparable with the 3-RTD average -
T hot prior to the failed RTD.
Other than the above changes, the instrumentation and control will remain the l
same and unchanged from what has previously been reviewed by the Staff. For example, two out of four voting logic continues to be utilized for protection functions, with the 7300 process control bistables continuing to operate on a
( "de-energize to actuate" principle. Non-safety related control signals continue to be derived from protection channels.
The above principles of the 3.cdification, including information presented in this report, and Figure 4 have Deen reviewed to evaluate conformance to the Section 4 requirements of IEEE-279-1971 criteria and associated 10CFR 50
- General Design Criteria (GDC), Regulatory Guide, and other applicable industry i
3522e:1d/092385 24
t WESTINGHOUSE PROPRIETARY CLASS 3 TABLE 4-2 PARAMETERSUSEDINTHEANALYSISOFTHERODCLUSTERCbHTROL
]
ASSEMBLY EJECTION ACCIDENT Beginning Beginning End of End of of Cvele of Cycle Cycle Cycle Power level, Percent 102 0 102 0 Ejected rod worth, Percent AK 0.a 0.75 0.23 0.90 Delayed n$utron fraction, Percent 0.50 0.50 0.44 0.44 Feedback _ reactivity weighting 1.20 2.07 1.30 3.55 Trip reactivity, Percent AK 4.0 2.0 4.0 2.0 F before rod ejection q 2.50 -
2.50 -
F after rod ejection q
4.5 11.0 5.9 20.0 Number of operational pumps 4 2 4 2 Max. fuel pellet average 3902 3563 3961 3490 temperature *F Max. fuel center temperature. *F 4962 4134 4893 3974 Max. clad average temperature *F 2166 2683 2199 2690 Max. fuel stored energy, cal /gm 185 152 172 148 Percent fuel melt <10 0 <10 0 3522e:1d/092385 25
l
~ \
WESTINGHOUSE PROPRIETARY CLASS 3 standards. Section 3 of the IEEE 279-1971 standard requires documentation of a design basis. The information presented in this report, incl,uding Figure 4, provide the documentation for the proposed design change and conform to the Section 4 requirements of IEEE-279-1971 criteria and associated 10CFR 50 General Design Criteria (GDC), Regulatory Guide, and other applicable industry standards. Section 3 of the IEEE 279-1971 standard requires documentation of a design basis. Following is a discussion of conformance to pertinent I&C criteria:
- a. Sinale failure criterion continues to be satisfied by this change because the independence of redundant protection sets is naintained.
- b. Quality components and modules being added is consistent with use in a Nuclear Generating Station Protection System. For the Westinghouse Quality Assurance program, refer to Chapter 17 of the FSAR.
- c. Equipment seismic and environmental qualification will be to IEEE standards 344-1975 and 323-1974, respectively, as described in WC/ P 8587, Rev. 5 " Methodology for Qualifying Westinghouse WRD Supplied NSSS. Safety Related Electrical Equipment".
~
- d. The changes will continue to maintain the capability of the Protection System to initiate a reactor trip during and following natural phenomena credible to the plant site to the same extent as the existing system.
- e. Channel independence and electrical separation is maintained because the Protection Set circuit assignments continue to be Loop 1 circuits input to Protection Set I; Loop 2, to Protection Set II; Loop 3, to Protection Set III; and Loop 4 to Protection Set IV, with appropriate observance of field wiring interface criteria to assure the independence. Output circuits are the same as before except that there will be one T cold and 3 T hot outputs to the computer sent through Class lE isolators in each Protection Set.
4
, 3522e:ld/092385 26
- WESTINGHOUSE PROPRIETARY CLASS 3 I i- i
- f. The Section 4.7 of IEEE 279-1971 and GDC 24 requirements concerning Control and Protection System interaction are satisfied because, even
, though control signals are derived from Protection Sets, the 2/4 voting j coincidence logic of the Protection Sets is maintained.
- Where a single random failure can cause a control system action that results i in a generating station condition requiring protective action and can also prevent proper action of a protection system channel designed to protect
,' against the condition, the remaining three redundant protection channels will
{ be capable of providing the protective action even when degraded by a second random failure.
This is because even though 1/4 channels failed without partially tripping, only 2 of the remaining 3 channels are necessary for a plant trip.
On the basis of the foregoing evaluation, it is concluded that these I&C modifications required for RTD bypass removal for the McGuire units will meet IEEE 279-1971, applicable GDC's, and industry standards and regulatory guides.
4.4 MECHANICAL SAFETY EVALUATION
, The presently installed RTD bypass system is to be replaced with fast acting narrow range RTD thermowells. This change requires modifications to the hot leg scoops, the crossover leg bypass return nozzle, the cold leg piping and the cold leg bypass manifold connection. All welding and NDE will be performed per ASME Code Section XI requirements. Each of these modifications is evaluated below.
The original bypass piping which connects each hot leg to a separate bypass manifold, and the bypass manifold itself must be removed and the scoops, which are left intact inside the RCS piping, are modified to accept three fast response RTD thermowells. [
].***C A thermowell design will be used such that the tip of the thermowell [
).+a c The thermowell will be fabricated in accordance 3522e:1d/092385 27
WESTINGHOUSE PROPRIETARY CLASS 3 with [ ]+8 of the ASME code (Class 1). The installation of the thermowell into the scoop will be performed [ ,
].+a,c The root and final weld passes will ,be examined by
[ ]+8'". Prior to welding, the surface of the scoop onto which welding will be performed will also be [ ).+a,c The cold leg RTO bypass nozzle must also be modified to accept a fast response thernowell and the bypass line removed. The nozzle must be cut and then the thermovell welded into place. Additionally, a spare fast response thermowell will be added to the cold leg in the length between the reactor coolant punp discharge and the accumulator nozzle. This necessitates the creation of a new penetration into the piping. The boss for the new connection will be [
].+a,c The thernowells will extend [ ]+8'C into the flow stream from the ID of the pipe. This depth has been justified based on [
]+8'Canalysis.
'The root weld joining the thermowells to the nozzles will be [
),+a,c ,
The three thermowells in the hot leg and the two thermowells in the cold leg provide a total of 20. Thennowells will be utilized at each of the four-loop McGuire unit!. and they will perform the same function as the original bypass T
hot and T cold signals. '...
The cross-over leg bypass return piping connection must be removed and the nozzles capped. The cap design, including materials, will meet [
t i
f- ).** Machining of the bypass return piping, as well as any machining performed during modification of the 3522e:1d/092385 28
- - . - - .- .. -= . _ . .
WESTINGHOUSE PROPRIETARY CLASS 3 thennowells on the hot and cold legs, shall be performed such as to minimize debris escaping into the reactor coolant system. -
1 In accordance with [ ]+a c of the ASME Code, a hydrostatic test of new pressure boundary welds is required when the 4
connection to the pressure boundary is larger [ ]+" in diameter. Since the cap for the crossover leg bypass return pipe is I
[ ]** inches and the cold leg RTD connec'. ions are [ ]**'*, a system hydrostatic test is required after bypass elimination at McGuire. Paragraph
[ ]+"'" defines this test pressure to be [ ]**'Ctimes the norwal operating pressure at a temperature of [ ]+C .
The integrity of the reactor coolant piping as a pressure boundary component, is maintained by adhering to the applicable ASME Code sections and Nuclear Regulatory Commission General Design Criteria 1,14,15, 30, 31 and 32. The
, pressure retaining capability and fracture prevention characteristics of the piping is not compromised by these modifications. Tnerefore, no unresolved safety issue is involved as defined in 10 CFR 50.59.
I 1
5.0 CONTROL SYSTEM EVALUATION
. A prime input signal to the various NSSS control systems is the RCS average temperature (T,yg). This is calculated electronically as the average of the p neasured hot leg and cold leg temperatures in each loop.
The major control systems affected are [
.]+*** The effect of the new RTD is to potentially change the time response of the T,yg channels in the various loops. However, as noted in Section 2.1 Table 1, the naw RTD system will have a time response close to that of the present system. There will therefore be no significant effect on the T,yg channel response, and no apparent need to revise any of the control i
system setpoints from those presently installed in the plant. The need to modify control system setpoints will be determined during the plant startup
- following the installation of the new RTD system by observing control system behavior.
3522e:1d/092385 29
WESTINGHOUSE PROPRIETARY CLASS 3
6.0 CONCLUSION
S The fast-response RTDs installed in the reactor coolant loop piping has undergone extensive analyses, evaluation and testing as described in this report. The incorporation of this system into the McGuire design meets all Safety, Licensing and Control requirements necessary for continued licensed operation of the McGuire station. The analytical evaluation has been supplemented with in-plant and laboratory testing to further verify system performance. .The fast-response RTDs installed in the reactor coolant loop 4
piping adequately replaces the present hot and cold leg temperature measurement system and enhances ALARA efforts and improved plant reliability.
i f
0 4
9 f
i 3522e:ld/092385 30 r.
-- - , ,. ,. , . , , - - - - _ .