ML20029A617
| ML20029A617 | |
| Person / Time | |
|---|---|
| Site: | Mcguire, Catawba, McGuire  |
| Issue date: | 02/20/1991 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20029A616 | List: |
| References | |
| NUDOCS 9102250353 | |
| Download: ML20029A617 (17) | |
Text
.
g[g>R fitOg 1%,
UNITED STATES p,
NUCLEAR REGULATORY COMMISSION g
wAsmNGTON, D. C. 70$$$
\\,
/
=...+
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATING TO TOPICAL REPORT BAW-10173P, REVISION 2 MARK-BW RELOAD SAFETY ANALYSIS FOR CATAWBA AND MCGUIRE DUKE POWER COMPANY CATAWBA NUCLEAR STATION AND MCGUIRE NUCLEAR STATION DOCKET NOS. 50-413, 50-414, 50-369, 50-370
1.0 INTRODUCTION
By letter dated Wrch 30,1989 (Ref.1), as amended by letters of October 22, and November 28,1990 (Refs. 2 and 3), Duke Power Company submitted Todcal Report BAW-10173P, Revision 2, " Mark-BW Reload Safety Analysis for Cat erba and McGuire," for NRC staff review. The report contains safety analyses of the transients and accidents other than loss-of-coolant accident (LOCA), and is intended to serve as reference for future reload safety evaluations as part of the licensing bc. sis for reload with B&W Fuel Company's (BWFC) Mark-BW fuel in the Catawba and ficGuire Nuclear Plants.
The scope of events considered is consistent with that addressed in the final safety analysis reports (FSAR) for Cataw'sa t nd McGuire, and covers all the transients and accidents delineated in sections 15.1 through 15.6 of Regulatory Guide 1.70.
The sequence of events and bounding analysis results of each of the accidents and tre.sients in the reference FSARs are examined with respect to continued applicability to reload cycles with B&W fuel. For those transients determined to be pote:ntially affected by the reload fuel design or operation, reanalyses are made. Other transients are evaluated, identifying the televant core-related parameters and bounding values to be confirmed for consistency with reference safety analyses.
In Section 6 of the report, a list of Key parameters and their boun: ling values is provided as a basis for reference in future reload safety evaluation to assure that the analyses in this report and the reference safety analyses remain valid. A new analysis would be required if the cycle-specific values are not bounded by these values.
The staff evaluation follows.
2.0 STAFF EVALUATION 2.1 Feel Design Compatibility The Mark-BW fuel assedly is a 17x17, standard lattic. design similar to iiestinghouse 1/X17 standard (STD) and optimized fiiel assedites (OFA). A comparison of the Mark-BW, STD and 0FA fuel designs is summarized in Table 2.1 9102250353 910220 PDR ADOCK 05000369 P
4 of the report. The Mark-BW fuel has the same fuel rod diameter as the STO fuel which is slightly larger than the OFA fuel. Similar to the OF/ fuel, Mark-BW contains spacer grid made of Zircaloy instead of Inconel to recice parasitic neutron absorption.
Five spacer grid assemblies employing flow mixing vanes on the downstream edges are used in the upper portion of the fuel assembly to improve thermal performance of the fuel assembly by enhancing coolant turbulence. Mixing vanes are not used on the lowest intere.ediate spacer grid since the thermal enhancement is not needed in this cooler region of the fuel as sembly.
The report indicates that this mixing vane design and pattern have been verified by CHF testing and is based on proven mixing vane designs.
During the transition to the Mark-BW fuel at the McGuire and Catawba Nuclear Stations, the resident fuel will be of the Westinghouse OFA design.
To minimize the hydraulic differences between the two fuel designs to reduce transition core effects, the basic Mark-BW design incorporates the following key features to be compatible with the OFA fuel: leaf type holddown springs, dashpot region in guide thimbles for control rod deceleration, mixing vanes on spacer grids, and guide thimbles used for intennediate spacer restraint, To verify hydraulic compatibility, the Mark-BW and the OFA assemblies underwent flow tests at Catawba using BWFC's transporteble flow test rig (TFTR).
Pressure drop testing was performed on individual fuel assemblies to obtain pressure drop characteristics for both the Mark-BW End the OFA fuel designs,
~
inc fing a (*ect comparison of the pressure drop distributions for the two designs.
The FTR test data were acquired up to a Reynolds number of approximately 100,000, and the results demonstrated the total pressure drop of the Mark-BW to be slightly less than that of the 0FA.
The Mark-BW assemblies have a higher loading of uranium and a lower hydraulic to uranium ratio (H/U) than the OFA design because of larger fuel rod diameter.
Even though the Doppler and moderatur coefficients of the Mark-BW design are essentially the same as that of the STD fuel, the Doppler coefficient is slightly different and the moderator coefficient is more negative than that of the 0FA design due to lower H/U ratio. The licensee indicated that the Mark-BW fuel had neutronic behavior dmilar to the STD fuel and only slightly different than the OFA fuel. This neutronic similarity was demonstrated in a res]onse to a staff question (Ref. 4;, which provided the comparisons among the Marr.-BW, STD and 0FA fuel of the K-infinity values versus burnup, moderator temperature, and fuel temperature, respectively. The licensee concluded that the Mark-BW design represented a small change from the Westinghouse designs, and che use of the Mark-BW assemblies in the We:tinghouse core would not adversely affect plant operation or neutronic parameters.
2.2 Non-LOLA Transient Analysis Methodology The analyse: of non-LOCA accidents and transients are performed with the safety analysis methods described in topical report BAW-10169P-A (Ref 5) plant In general, a system transient analysis code, RELAP5/M002-B&W (Ref. 6), is used to model and calculate the reactor coolant system responses for each transient.
The reactor core power during each trcnsient is calculateo by the point kinetics neutronic model in RELAP5/M002-B&W with physics parameters such as reactivity coefficients and power peaking facters obtained from independent
3-core physics codes FLAME 3, N0ODLE and PDQ07 (Refs. 7, 8, and 9). The subchannel core thermal hydraulic code LYNXT (Ref. 10) is then used to calculate the hot fuel rod temperature and departure from nucleate boiling ratio (DNBR) using as boundary conditions the RELAPS/ MOD 2-B&W results, such as power level, core flow rate, temperature and pressure as functions of tine.
The critical heat flux (CHF) correlation BWCMV (Ref. 11) is used in LYNXT for the calcLGhtion of CHF and DNBR.
Two plant noding models are used in conjunction with RELAP5/H002-B&W for safety analyses, i.e., the low-power model for analysis of a steamline break at low poer, and the f ull-power model for analyses of other transients at full power.
These two models are developed to provide the no6ing arrangements with sufficient details to describe in.portant transient phenomena with sufficient accuracy of calculation, yet simple enough to minimize computation time. These models, described in BAW-10169P-A, have been reviewed and accepted by the NRC.
2.3 Core Thermal-Hydraulic Analysis The core tnermal-hydraulic analysis is performed for a transient to determine if the hot channel minit m DNBR is above the DNBR limit such that there is 95 percent probtbility with a 95 percent confidence N el that no fuel rod wiil experience a departure from nucleate boiling. The analysis uses a one-rass LYHXT model and the statistical core design (SCD) technique described o BAW-10170P-A (Ref. 12).
The SCD analysis treets the uncertainties of the ore state and bundle parameters statistically, and a design DNBR limit, tne statistical design limit (SDL), is established incorporating the DNBR uncertainties. All variables treated in the development of the SDL are then input into the thermal-hydraulic analysis computer codes at their nominal values.
For the Catawba and McGuire reactors with Mark-BW fuel, the SDL of 1.35 has been determined with the BWCMV correlation. However, for conservatism in the safety analysis, additional margin is adaed to the SDL to define a thermal design limit (TDL) of 1.50, which is used as the limit DNBR value.
Except for steam line break, all transients are analyzed with the SCD method so that the initial reactor power, pressure, and RCS temperature in the DNBR analysis are assumed to be at their nominal values with the uncertainties included in the DNBR limit.
For the transition cycles in which the resident OFA fuel is being displaced by Mark-BW fuel, the core thermal hydraulic analyses are based on the configuration of the final full-core Mark-BW fuel. Because there are differences in the fuel diameters and axial pressure drop profiles between the Mark-BW and 0FA fuel designs, diversion cross flow between the assemblies of two different designs is affected by various degree at various elevations depending on the relative resistances of the two fuel designs. The Mark-BW fuel assembly may gain flow at some elevations, while lose flow at the other elev. ions compared to a homogeneos s core. The same is true for the 0FA fuel-Therefore, both the 0FA and the Mark-BW assemblies must be examined for potential transi'. ion core penalties, which may be a7 plied either by increasing the thermal dusign limit or by r W ocing the calculated DNBR in the DNB
.A
e.z Ne 4--
i s
analysis.- In response to a _ staff questOn regarding applicability of the.
analysis ofja full-core Mark-BW. fuel to the transition cycles, the licensee-performed'a mixed core at,alysis to quantify the transition' cycle penalty that
- will be applied to either the-resident or the Mark-BW fuel designs.
- In determining a generic transition core penalty, the licensee used two bounding transition core models for the evaluation, i.e., one Mark-BW fuel assembly in;a core of the OFA fuel assentlies, and a single OFA in a Mark-BW
- core. The analyses -were perfortaed at the limiting core safety limit and loss of' reactor. coolant flow statepoints, using the design radial power distribution
-and the single-pass _LYNXT model. L*ihe DNBR results of the limiting case were compared to the results calculated _at identical conditions with the full-core Mirk-BW model, and showed a,4 percent DNBR penalty for the OFA and no penalty for the. Mark-BW fuel.
.The licensee indicated that; in order to accommcdate the OFA tranriMon mixed u
m penalty,'the OFA peaking limits will be set at a level equa? :o 95 percent of the Mark-BW peaking-limits. - This' maintains the OFA limits at a h vel that -
ensures that OFA DNBR values will-be higher than those credicted in the Mark-BW
~_ safety analyses and safety limits..Since the W cores' for both McGJire and Catawba:are currently _ limited to a' maximum entholpy rise Dctor of 1.49, while the Mark-BW core analyses for the core operating limit presented in BAW-10163P-A-(Ref.13) assumed a enthalpy rise factor of 1.55, the transition penalty is bounded by the imposed peaking difference, and no additional transition core penalty need be assessed against the thermal design DNBR limit.
-However, if a-reload ' design is such that the enthalpy rise factor of. the OFA 1~
fuel is higher.than'96 percent of that of the Mark-BW. fuel,-the mixed core
- penalty should be assessed.to~'the OFA assembly.
2.4 Transient _ And Accident Events Evaluation 2. 4 '.1 Increase In' Secondary Heat Removal Transients An, increase'~1n secondary side heat removal causes a reduction in reactor
- coolant' temperature and pressure, and produces a positive reactivity insertion 4nd power increase that could chal bnge the fuel thermal limits. The 1 transients in this category inchde-excessive increase in-secondary steam flow,.
inadvertent: opening of a steam _ generator relief or safety valve, a steam pipe
- tenverature or increase. in feedwater flow. The most severe event of this
' category is a ' steam system piping failure, which 'is an ANS-Condition IV event.
7he licensee performed an analysis of the steam'line break transient using thei RELAPS/ MOD 2-B&W low-power plant' system model. This model has cross-flow paths Linithe-core to allow for flow mixing between the faulted-loop and the combined intact loop. lSince the~ reactivity feedback' is affected by the coolant density, cross-flow mixing is an important parameter in determining the reactivity and U
ersv tv y
--uJWarse-r we.
w e w
aw-i am-
's-
~mwg w.
w T
e w
w
'e v.
wy e
+-- =
+z y-ae sww e=
dpteF-tg-twg--p
+--w m
4
=*'t-e
. return to power.
As stated in BAW-10169P-A, conservative assumption were made to use a 80/20 junction flow area mixing and a 50/50 faulted /unfaulted reactivity feedback weighting scheme.
For core thermal hydraulic and hot channel minimum DNBR calculations, the RELAP5/M002-B&W results of inlet flow, temperature, exit pressure, and core power, are used as boundary conditions of the LYNXT code.
The FLAME 3 code supplied the radial peaking and axial power shapes of the individual channels.
Since the steam line break accident is relatively slow, the transient could be reduced to a steady-state calculation with the bounding state conditions.
The analysis was performed with the following assumptions: (1) a positive moderator density coefficient corresponding to the end-of-life rodded core with the most reactive rod control cluster assembly (RCCA) in the fully withdrawn position, (2) the unavailability of one safety injection train as the worst single failure, and no boric acid solution was injected via the safety injection system so as to maximize neutroa power resulting from the steam line break, (3) limitstion of the break flow to the throat area of 1.4 square feet of the integral flow restrictors installed on the steam generators, and (4) use of power peaking factors corresponding to end-of-core-life with one stuck RCCA and non-uniform core inlet coolant temperatures.
The original analysis was yerformed with offsite power available.
In response to a staff lowever, the licensee submitted Revision 1 to BAW-10173 (Ref. 2) question, to include an analysis without offsite power.
For the case where offsite power is available, full reactor coolant flow is maintained during the entire transient, and the analysis was performed with the single pass 5-channel 1/1-core LYNXT model, which hos been reviewed previously as an acceptable LYNXT core modeling. However, with the loss of cffsite power, a reactor coolant flow coastdown occurs, v.d the more thermal-hydraulic conditions are characterized by natural circulation coolant flow with strong gradient in inlet temperature and power peaking coMith ns. Therefore, a modification was made to the LYNXT code by incorporatirg the implicit pressure-velocity (PV) algorithm to ensure convergent solution. This PV algorithm solves the same governing equations as the i!nplicit numerical scheme presently available in LYllXT, but also allows for an arbitrary flow direction.
The PV algorithm uses a staggered grid formulation and a Newton method to solve the governing equations. In addition, a 9-channel, half-core model was used with the core modeled as three dir, tinct regions, i.e., a cold region to represent the faulted quadrant, a hot region to represent the combined intact quadrants, and a mix region to provide an interface between the hot and cold core areas. A different inlet flow and tempsrature was applied to the channels within each region. A stuck RCCA was assumed to be in the most limiting assembly within the cold quadrant. Even though the approach appears to be acceptable, the licensee should preform a benchmark analysis to verify appropriateness of the LYNXT modifications and modeling for the steamline break analysis wi'h a loss of effsite power (LOOP).
. The results of analysis are shown in Figures 4.1.5-3 through 4.1.5-34.
The LOOP case has lower neutron and thermal powers due to lower reactivity insertion than the power available case. The report also indicated that the minimum DNBRs are 1.57 and 1.88, respectively, for the offsite power available and LOOP cases. The higher DNBR for the LOOP case is partly due to lower power level and an axial power peaking toward the core inlet. The minimum DNBRs for both cases are higher than the DNBR limit for the W-3 correlation used at low pressure conditions.
Therefore, no DNB occurs.
However, the acceptability of the results of the steam line break-LOOP case is subject to a successful validation of the LYNXT inodifications and modeling for its analysis.
Of the ANS Condition II moderate frequency transients associated with the secondary system depressurization, the spurious opening or failure of a steam generator relief, safety, or steam dump valve, is the most severe overcooling event. However, the existing FSAR bounding analysis showed that the acceptance criteria were met.
In addition, this event is less severe than the main steam line failure. Since a steam line break has been shown to have no DNB, there is sufficient assurance that the moderate frequency steam release events will also meet the acceptance criterio.
Other events such as a reduction in feedwater temperature and an increase in fec.dwater flow transients are bounded by the more limiting case of an increase of steam flow. The insertion of B&W fuel will not change the magnitude of the initiating decrease in feedwater temperature, nor an increase in feedwater flow. The licensee indicated that it will confirm the reactivity coefficients for the reload cores to be within the bounds of those in the reference safety analyses so that operation with BWFC fuel will not affect the consequences of these events. The staff finds this to be acceptable.
2.4.2 D_ecrease in Secondary Heat Removal Transients This category includes transient events which result in a loss of steam load or feedwater flow, and feedwater system pipe break. These events result in a reduction of the capacity of the secondary system to remove the core heat and causes increases in primary coolant temperature and pressure that may challenge the allowable system pressure and fuci design limits.
Loss of steam load may be caused by a loss of external load, turt'ine trip, inadvertent closure of main steam isolation valves, or loss of condenser vacuum. Of these events, turbine trip is the limiting transient which causes closure of the turbine stop valves and cutoff of the secondary steam flow, l
resulting in a rise in the secondary and primary coolant pressures and i
temperatures, as well as a reactor trip on high pressurizer pressure and an actuation of the pressurizer safety valves to limit the pressure excursion.
A loss of feedwater flow may be caused by a loss of non-emergency AC poor to the station auxiliaries, feedwator pump failure, valve malfunction or loss of i
l offsite AC power. The loss of non-emergency AC power to the station i
auxiliaries initiates a loss of main feedwater event and is similar to the loss L
of feedwater initiated by other mechanisms, except that a loss of AC power l
. results in unavailability of the reactor coolant pumps and a RCS flow coastdown. However, the loss of feedwater analysis requires consideration of a loss of offsite power and RCS pumps trip, and is therefore the same as a loss of AC power.
The main feedwater line break is a design basis event analyzed to demonstrete overpressure protection of the RCS and continued capability for core cooling.
With respect to operation with the Mark-BW reload fuel, the licensee evaluated its effect on the existing safety analyses in the McGuire and Catawba FSARs.
The FSAR turbine trip bounding analysis assumed a complete loss of steam load from full power, the availability of the reector trip functions except for the anticipatory direct reactor trip on turbine trip, and the loss of main feedwater at the start of.the event with no credit taken for auxiliary feedwater to mitigate the consequence of the transient. To encompass both the limiting overpressure response and the worst DNB conditions, the analyses bound plant response with four combinations of minimum and maximum reactivity feedback, and availability of pressurizer pressure control functions of relief valve and spray.
For example, the worst DNB case would combine the assumption of minimum reactivity feedback with positive moderator temperature coefficient and active pressurizer controls.
For this case, the minimum DNBR was calculated to be 1.8 compared to minimum DNBR of 1.5 for a more limiting complete loss of RCS flow transient, and is well above the minimum DNBR limit of the WRB-1 CHF correla-tion.
Regarding the loss of feedwater transient, even though the steam generator heat removal capability is reduced below the core heat generation rate, there is sufficient steam generator inventory for adequate heat transfer as the reactor is tripped before thermal limits are approached.
Because of the continued heat transfer capacity, the loss of main feedwater event as analyzed in the FSAR did not approach the more limiting DNBR conditions of other transients such as a complete loss of RCS flow transient. The loss of feedwater and loss of AC power events were analyzed primarily to demonstrate the reactor protection system and the engineered safeguard systems such as an auxiliary feedwater system have sufficient capability for removal of the core decay heat by the steam generator without RCS overpressurization.
In the analyses of RCS pressurization due to decrease in secondary heat removal transients, only the pressurizer safety valves (pSV) and the reactor trips were used to maintain the RCS pressure below the design criteria with no credit taken for the pressure relieving action of the steam dump or pressurizer pressure control systems. For the worst pressurization event of the main feedline break, the analysis assumed the maximum values of moderator temperature and Doppler reactivity coefficients, and maximum delayed neutron fractions to retard the post trip decline in neutron power and maximize the long term heatup of the reactor coolant system. The FSAR analyses showed that the acceptance criteria were met.
. Of the various core parameters, such as the reactivity coefficients, delayed neutrvn fraction and decay heat, etc, that could be affected by the difference in the fuel and Yuel cycle designs, the licensee has determined that the dominating parameter that could significantly affect the bounding pressuriza-tion analysis resu.'s of these transients is the decay heet level following the reactor trip.
In response to a staff question, the licensee provided comparisons of the total decay heat power data at various irradiation levels to show the decay heat of Mark-BW fuel is virtually the same as that of the STD fuel, and is bounded by the OFA fuel under similar operating conditions.
Thus the FSAR bounding decay heat level chosen for the Westinghouse fuel designs would also bound Mark-BW fuel.
Because the system conditions and protection functions represented by the boundin,a FSAR sequences are not affected by operation with the Mark-BW fuel, the core-related parameters that define the bounding cases in the FSAR should renwin applicable. The licensee indicated that-the bounding values for these parameters will be checked on a cycle-specific basis. The key paraneters for reload safety evaluation are sunnariztd in Table A.1 of the topical report.
2.4.3 Decrease In Reactor Coolant System Flow The transients in this category consist of a partial loss and a complete loss of RCS flow, and a reactor coolant pump shif t seizure or locked rotor.
2.4.3.1 Loss Of Forced Reactor Coolant Flow Two cases of loss of RCS flow were considered: a partial loss of forced RCS flow caused by a mechanical or electrical failure in a reactor pump, and a complete loss of RCS flow caused by a simultaneous loss of electric supplies to all RC pumps. A loss of RCS flow when the reactor is at power will result in a rapid increase in the RC temperature and attendant DNB with subsequent fuel damage if the reactor is not promptly tripped.
Protection is provided by the reactor trip on (1) low RCS flow trip against a partial loss of flow accident, and (2) power supply undervoltage or underfrequency and low RCS loop flow trips-against a complete loss of flow.
The RCS response of the loss of RC flow transients are analyzed using the RELAP5/ MOD 2-E,&W code and the full power plant model. The method of analysis and the assumptions regarding initial operating conditions and reactivity coefficients are identical for both partial and complete loss of flow cases, except for the difference in flow coastdown and the reactor trip functions.
For the partial loss of flow transient, the flow coastdown is calculated using a bounding input set of pump speed versus time data for the faulted loop, and the reactor is tripped on low primary coolant flow signal.
For the complete loss of flow transient, four pump flow coastdown is considered and the reactor trip is ectuated by either the RC pump power supply undervoltage or under-f requency. Conservatism in the flow coastdown modeling is confirmed by comparison with the startup test data of the Catawba and McGuire units. A conservativelf large absolute value of the Doppler temperature coefficient is used, equivalent to a total integrated Doppler reactivity from 0 to 100 percent
-g-power of 0.016 Dk/k. A positive moderator temperature coefficient of +7 pcm/*F is assumed since this results in the maximum care power during the initial part of the transient when the mii :Tum DNBR is reached. The analyses conservatively assumes the highest worth F IA stuck in the fully withdrawn position.
The LYNXT code is then used to calculate the heat flux transient based on the neutron power and flows from the RELAPS to determine the cure mininum DNBR.
The results are shown in the Figures 4.1.1-1 through 4.3.1-6 and Figures 4.3.2-1 through 4.3.2-5, respectively for partial and complete loss of flow transients.
In both cases, the minimum DNBRs are above the thermal design limit value, and therefore no DNB or fuel failure is predicted.
2.4.3.2 Reactor Coolant Pump Shaf t Seizure (Locked Rotor)
Reactor coolant (RC) pump shaf t seizure or locked rotor is an ANS Condition IV event. Upon an instantaneous seizure of a RC pump rotor, flow through the affected RC loop is rapidly reduced, leading to the initiation of a reactor trip on a low flow signal. Heat transfer to the secondary side of steam generators is reduced because of (1) the decrease of the tube side film heat transfer coefficient resciting fron the reduced RCS flow, (2) the RC cooldown, and (3) the increase of the shell side temperature due to turbine flow reduction to zero upon plant trip.
Continued heat transfer from the fuel rods stored energy causes the coolant to expand resulting in an insurge into the pressurizer and a pressure increase throughout the RC system.
This insurge compresses the steam volume in the pressurizer, actuetes the automatic spray system, anc' opens the power-operated relief valves (PORV) and PSVs.
The analysis was performed for one RC pump rotor locked with four loops in operation.
Both the RELAP5/M002-B&W and LYNXT codes are used for the reactor system transient and cure thermal-hydraulic calculations, respectively.
After pump seizure, the neutron flux is rapidly reduced by control rod insertion, which was assumed to begin one second after the flow in tha affected loop reached 86.5 percent of nomine' flow. The analysis assumed that offsite pesec is lost at the start of the transient and, therefore, the unaffected RC pumps coast down during the transient. The reactivity coefficients were a zero moderatur coefficient, which maximizes the neutron power befort reactor trip, and a large negative power coefficient of -12.6 pcm, which maximizes the neutron power after reactor trip.
The PSys are open fully and their capacity for steam relief is obtained from the Catawba and McGuiPe safety analysis reports.
For the evaluation of pressure transient, the pressure reducing effect of PORVs and spray, steam dunp, and controlled feedwater flow af ter reactor trip are not included in the analysis. These assumptions are conservative for the evaluation of over-pressure concern, but are nonconservative for the W3R evaluation. While the analysis showed a ped RCS pressure of 2550 psia, well below the over-pressure acceptance criter1a, the licensee in response to a staff question performed an additional LYNXT DNBR analysis in which the core exit pressure is l
)
l l
s held constant at the initial value, which is a conseryctive assumption since a locked rotor is a pressurization event. This analysis resulted in the number of pins in DNB of 3.3 percent compared to the original analysis of 1.8 percent described in the report. The licensee did not perfonn a radiological consequence analysis to demonstrate compliance with 10 CFR 100, and therefore such an analysis is necessary to shcw acceptable dose release.
2.4.4 Reactivity And Power Distribution Anumalies Reactivity and power distribution anomolies can be caused by uncontrolled RCCA withdrawal and inserti(n, startup of an inactive RC pump, boron dilution, inadvertent luading of a fuel assembly into an improper position, or a RCCA ejection.
Both McGuire and Catawba Technical Specifications prohibit operation in Modes 1 and 2 with less than 4 active loops.
Therefore, the stortup of an inactive RC pump during power operation is not pussible. The startup of an idle loop from the non-power modes would not result in significant reactivity insertion because of similar coolant temperature in the idle and active loops.
This conclusion is not affected by the reload of the Mark-BW fuel.
Regarding the inadvertent loading of a fuel assembly into an improper position, the licensee has fuel loading controls and procedures to prevent misloading of fuel rods and fuel assemblies.
Af ter refueling is completed, startup physics testing is performed at zero and low power levels prior to escalation to full power. The incore system of moveable flux detectors is used to verify that the core power distributions are consistent with the design.
These procedures are not affected by the reloading of the Maak-BW fuel.
Other reactivity and power cistribution anomalies are evaluated as follows:
2.4.4.1 Subcritical Uncontr iied RCCA Bank Withdrawal An accidental withdrawal of a RCCA bank during startup, caused by a malfunction of the rod control system, results in an uncontrolled reactivity insertion and rapid increase of neutron puwer. However, the power increase is terminated by the reactivity teedback effect of the negative Doppler coefficient which limits the power to a tolerable level during the delay tir? for protective action of reactor trip.
Eventually, the reactor is tripped by either the high neutron flux trip (Source range, intermediate range or power range) or high neutron flux rate trip.
The analysis is performed using RELAP5/ MOD 2-B&W and the low power plant model for the calculation of the core average nuclear power. The RELAP5 kinetics calculation is done from hot zero power conditions with two pumps in operation and zero steam geno ator heat demand. The analysis assumed the least negative k
Doppler ano most positive moderator reactivity coefficients to produce the highest core average power.
Reactor trip is initiated by the power range high neutron flux with a 10 percent increase to account for the most adverse combination of instrument and setpoint errors, as wall as delay for trip signal actuation and RCCA release. The reactor trip insertion characteristic is based 1
11 on the minimum zero power scram worth, consistent with the highest worth RCCA stuck in the fully withdrawal position. The maximum positive reactivity inst.rtion rate assumed is greater than that for the simultaneous withdrawal of the combination of two sequential control banks having the greatest combined worth at maximum speed.
The heat flux and temperature transients are determined by the LYNXT fuel rod heat transfer calculation. The initial power level is assumed to be below the
>ower level expected for any shutdown condition. The combination of the lighest reactivity insertion rate and lowest initial power produces the highest e
p' peak heat flux.
Bonding limits for the axial and radial power shapes are used in the hot channel transient DNBR calculation. The results provided in the response to a staff question show that the minimum DNBR at all times remains above the limiting value.
2.4.4.2 Uncontrolled RCCA Bank Withdrawal at Power Uncontrolled RCCA bank withdrawal at puwer results in an increase in the core heat flux. Since the heat extraction from the steam generator lags behind the core power generation until the steam generator pressure reaches the relief or safety valve setpoint, there is a not increase on the RC temperature. The reactor is tripped on high neutron flux, overtemperature delta T (OTDT),
overpower delto T (0PDT), high pressurizer pressure, or high pressurizer level.
The RCS transient is analyzed using RELAP5/M002-B&W and the full power plant model. The resulting system variables such as flow, temperature, pressures, and power level are input to LYNXT for core thermal hydraulic and DNBR a nalysis. The reactor was assumed to trip on high neutron flux at 118 percent of nominal full power.
The RCCA tri) insertion characteristic is based on the assumption that the highest worts assembly is stuck in its fully withdrawn ptsition. A conservatively smali Doppler coefficient end a least positive moderator density coef ficient are assumed corresponding to the beginning of core life. The report indicated that the reactivity coefficients assumed are tonsistent with the limiting DNPR result for the maximum rod withdrawal rate at (ull power. The maximum positive reactivity insertion rate of 75 pcm/sec, which is greater than that for the simultaneous withdrawal of the combinations of the two control banks having.the maximum worth at maximum speed, was assumed to be the worst case.
The staff SER on BAW-10169P-A indicated that the analysis should consider ydrious reactivity insertion rates from very low to maximum possible for the control system and the fuel and moderator feedback reactivity coefficients covering the range expected throughout the cycle.
In response to a staff question, the licensee performed a reanalysis with various reacthity insertion rates and feedback required by the SER.
A least negative MTC of +7 pcm/*F and the least negative fuel coefficient of
-6.0 pcm/% power is used in t'ne minimum feedback analysis.
For the maximum feedback analysis, a maximum negative MTC of -41 pcm/*F and the mon negative fuel coefficient of -12 pcm/% power is used.
- 4 12 figure 13.1 (Ref. 4) shows the analysis results of minimum DNBR as a function of rod withdrawal rate for both the maximum and minimum feedback cases.
The minimum rod withdrawal rate is shown to be the limiting condition with the lowest DNBR. All the cases analyzed have the reactor tripped on high neutron flux. This is because the analysis used a high value (1.411) of the preset manuelly adjustable bias K in the OTDT trip function, resulting in a higher OTDTtripsetpointoftemphraturedifferenceacrossthecore. Use of.ligher K value is conservative because using a smaller value of K may have a earlier rhactor trip by the OTDT function resulting in a higher DNBR for the transient.
3 The reanalysis results show that, for various reactivity insertion rates from very low to the maximum possible for the control system, and maximum and minimum feedback, the minimum DNBR is still above the minimum DNBR limit and is acceptable.
2.4.4.3.
EAHisoperation RCCA misoperation events, which can be caused by either system malfunction or operator e'ror, include one or more dropped RCCAs within the same grou), a dropped RCCA bank, a statically misaligned RCCA, and a single RCCA witidrawal.
Dropping one or more RCCAs inserts negative reactivity which may reduce power sufficiently to trip the reactor on low pressurizer pressure, or power may be reestablished either by reactivity feedback or control bank withdrawal.
If a dropped RCCA event occurs during the automatic rod control mode, the rod control system detects the drop in power and initiates control bank withdrawal which may result in power overshoot.
Rather than performing a detailed transient analysis, the event of dropped RCCAs in automatic mode is analyzed with the bounding system state points detennined based on the sequence of events that occurs during the transient.
The ap) roach is to model the limiting state point during the event with static core p vsics and thermal hydraulic methods and demonstrate that the margin to the design limits exists for each reload cycle.
Therefore, it is important to determine the maximum power overshoot and peaking factors.
The maximum power achieved for the rod drop events is limited by (1) the high flux or OPDT reactor trip function, and (2) the reactivity balance between the dropped RCCA worth, the control bank inserted worth, and the power coefficient.
For the analysis, the Bank D rods are conservatively assumed to be fully withdrawn by the rod controller af ter the rod drop to obtain the maximum positive reactivity addition. To provide the highest power-overshoot, it is also assumed th6t the core average moderator temperature and pressure remain as the initial nominal values. This is because the control system will insert the com ni bank to restore nominal average value when the average temperature is abe the nominal value, and the nressure controller could maintain the prwssure at nominal pressure as the pres:;ure increases due to power overshoot.
. _ - ~
0
' The reactivity from 3-D FLAME 3 or N0ODLE is used to determine the maximum power level and the peaking for eoch of the various combinations of dropped RCCAs.
The peak pin to assembly average powers are calculated with PDQ07. LYNXT is used to calculate the amount of DNBR peaking margin that is available at various puwer levels with the nominal average temperature and pressure.
This is done by determining the combination of radial and axial peaking conditions at the dropped RCCA state points that correspond to the thermal design DNBR limit, defined as the maximum allowable peaking (MAP) limits. These MAP limits are then used to calculate the loss of DNBR peaking margin for the peaking changes resulting from the dropped RCCAs.
For the limiting cycle-specific peaking conditions and power shapes, LYNXT is used to verify the MAP-based peaking margins. This method was used for a Mark-BW equilibrium cycle.
The amount of DNBR peaking margin lost for each of the dropped RCCA combinations versus power level is compared with the amount of DNBR margin available. The results shown in Figure 14-1 in response to a staff question show that no dropped RCCA combination has the DNBR peaking margin loss exceeding the available DNBR peaking margin.
The must severe RCCA misalignment situation with respect to DN8R at significant power levels is the case where one RCCA is fully inserted, or where Bank D is fully inserted and one RCCA is fully withdrawn. The licensee has performed the analyses of both fully inserted and fully withdrawn RCCA cases for a Mark-BW equilibrium cycle with the peaking factors determined by 3-D FLAME 3 and PDQ07, and the loss in DNBR peakirg margin calculated by LYNXT. The licensee indicated that the maximum peaking increase due to asymmetric RCCAs was less then 8 percent which is well within the 20 percent DNBR peaking margin at 100 percent power. However, the li:ensee indicated that analysis will be done to show the DNBR peaking margin loss is less than the available DNBR peaking margin at full power on a cycle by cycle basis.
The single RCCA withdrawal is classified as an ANS Condition III event. Even though the system response of a single RCCA withdrawal is similar to that of an uncontrolled RCCA bank withdrawal at power, the increased local power peaking in the area of the withdrawal RCCA results in lower minimum DNBRs than the withdrawn bank event. Depending on the initial Bank D insertion and location of the withdrawn RCCA, reactor trip may not occur fast enough to prevent the minimum DNBR value around the peak fuel rods from falling below the limiting criteria. The licensee in a response to a staff question (Question 15, Ref. 4) indicated that a fuel pin census of the bounding initial peak pin power l
distribution has found 4.9 percent of fuel rods to exceed the radial peaking of 1.55, which is a conservative failure criterion for the RCCA withdrawal event. Therefore, 4.9 percent of fuel rod would have DNBR below the thermal design limit and assumed failed.
Based on this failed fuel amount, the licensee should evaluate the radiological consequence to ensure compliance with a small fraction (10 percent) of the dose release criteria specified in 10 l
CFR 100.
L 1
l l
e 14 2.4.4.4 Borun Dilutiun No analysis is pt.rformed for the boron dilution of the reactor coolant resulting from m41 function of the chemical and volume control system. The report indicated that the licensee will evaluate the boron dilution transient for all applicable modes of uperation. The reference analysis evaluated for future reload cycles will be thusi. 4': FN for the cycles just prior to the insertion of Mark-BW fuel. A comparison of boron concentrations for the subsequent cycles to those of the reference analysis will be performed by the licensee to evaluate this transient.
In the event that the cycle specific values do not meet the assumptions of the reference analysis, the transient will be reanalyzed to verify that the minimum time for operator correction action is preserved.
2.4.4.5 RCCA Ejection Accident RCCA ejection, caused by a mechanical failure of a control rud drive mechanism pressure huusing, results in a rapid positive reactivity insertion and an adverse core power distribution which could lead to localized fuel rod failure.
The system response and hot spot analyses of a rod ejection transient are dependent upon the neutronics characteristics and thermal response of the fuel.
The licensee indicated that the neutronics parameters for Mark-BW fuel are expected to be within the current bounding physics parameters in the FSAR, and has perforraed a calculation of thermal response of Mark-BW fuel to an ejected rod power excursion using a representative cure average nuclear power excursion obtained from FSAR. The results show no discernible differences compared to the previously licensed fuel.
The licensee also indicated that the ejected rod worths, total peaking. factor, Doppler coefficient, moderator temperature coefficient, delay neutron fraction, pin census, and trip reactivity for each cycle will be calculated and shown to be less restrictive than the values presented in FSAR reference analyses and in Table A.2 in Chapter 6 of BAW-10173. These results will be demonstrated for full power and zero power et both BOC and E0C.
If they are not less restrictive, a reanalysis of the rod ejection must be performed.
2.4.5 Incrcase in Reactor Coolant Inventory This category can be caused by either an inadvertent operation of the emergency core cooling) system (ECCS) or a malfunction of the chemical and volume control system (CVCS. An operator error or a false actuation signal could produce spurious operation of the ECCS during full power operation.
Since the plants provice for reactor and turbine trip signals upon a safety injection (SI) actuation, the event is immediately terminated.
In the absence of these trip signals, actuation of the SI system will result in delivery of highly borated water to the RCS, which provides negative reactivity and a decrease in reactor power, RCS temperature and pressure.
Therefore, even if the normal direct reactor and turbine trips are assumed bypassed, the transient is terminated by
/ <t?MI E
- 2k+
IMAGE EVAL.UATION a
'% /7 [ k[/
$b
///p% 0
/&
~j*f4'4,Jg,
//o S TEST TARGET (MT-3) 3 x
% <e,p V &q, g
i a
R l.0 E $2 Ela y jj M1 i,i
[m LM 01 1.25 1.4 1.6 ress 150mm 6"
(#y %,,
/;##,/A
/
g 3,7ppj-s <g<4;;+
7 4//f
<r
___________.__.__________.________NL___________________________1d______________a'C
- [8*49 t$*
B Mk IMAGE EVAL.UATION
////g hE
///, %j/ g [ 4r/ g 4
\\y/o/77' \\ "?g f/
TEST TARGET (MT-3) 3 f
't %,' ?
V
@gy
'4g%
D 4
h B
B E
I.0 E B2 M i
5 9 Hal kE I.I
~
Ua 1.25 1.4
'l_i.6__
4 150mm
)
6" I
4 4
4?g ?,y,,,,,,,_
_ # g %p
<s
/s si, y
y x
-n reactor trip on low pressurizer pressure.
Because the power and coolant temperature decrease throughout the transient, DNBR increases throughout the transient. Therefore, there is no DNB concern.
This is also the case for the CYCS malfuiiction.
2.4.6.
Decrease In Reactor Coolant Inventory The decrease in RC _ inventory events, other then an LOCA, may be caused by an inadvertent opening of a pressurizer safety or reiief valve, or a steam generator tube rupture (SGTR).
Inadvertent opening of a pressurizer safety or relief valve results in a reduction in the RCS pressure and o corresponding reduction in DNBR which could potentially-challenge the core thermal design limits. The reoctor will be tripped by either the low pressLrizer pressure or OTDT trip. While the RCS pressure is decr'.6 sing at a moderate rate throughout the transient, the remaining core conditions, such as power, loop flow and average temperature, remain essentially constant prior to the reactor trip.
The analysis of the event is perfornied by a bounding state point analysis using LYNXT.- Since the transient response of the RCS is such that the DNBR will decrease throughout the pretrip period, the analysis bypassed the OTDT trip and assumed the reactor to be tripped by the low pressurizer pressure setpoint of 1835'psig. Since the minimum DNBR occurs immediately of ter the reactor trip because of a slight lag in thern,al power response, the bounding statepoint for the minimum DNBR calculation is conservatively chosen as a core inlet pressure of.1800 ps16.. The LYNXT calculation shows the minimum DNBR of greater than 1.9, well above the thermal design lin.it of 1.5.
The SGTR is a design basis accident. Upon failure of a steam generator tube, the RCS depressurizes, the reactor is tripped, inain feedwater flow isolated, and the SI system is actuated on the low prenurizer pressure signal. The event is effectively tenninated when the SI system makeup flow matches the rate of. coolant loss through the f ailed steam generator tube.
The licensee evalueted the bounding assumptions in the FSAR SGTR analyses aiming at maximizing the concentrations of fission products in the primary and secondary inventories and the mass flow through the faulted steam generator, and concluded that the limiting consequences of the reference safety analyses are not affected by the reload of BtW fuel. Therefore, no cycle-specific evaluation is required for future reloads cycles.-
3.0
SUMMARY
AND CONCLUSION l.
The staff ha; reviewed BAW-10173 submitted by Duke Power Company in support of
-its intended reloads with the BWFC's Mark-BW fuel in the McGuire and Catawba units. Testing and analysis have been perfonned to show sufficient similarity and compatibility between the Mark-BW fuel design and the resident fuel with n
L respect to the thermal-hydraulics and core physics performance.
1 L
l
4 16-The evaluations and analyses have been performed for non-LOCA transients as reference for future reload safety evaluations for operation with B&W Mark-BW fuel a:,semblies. These analyses tad evaluations confirm that operation of the Catowba and McGuire units future cycles with Mark-BW fuel will continue to be within the previously licensed safety limits.
Reanalyses of those transients affected by the fuel reloads demonstrate that the acceptance and design criteria continued to be net, with the exceptions noted below.
The staff finds the report to be acceptable for referencing in support of the operation for the future reload cycles with Mark-BW fuel with the following conditions:
1.
A benchmark analysis should be performed to verify appropriateness of the LYNXT modifications and modeling for the analysis of steam line break with loss of offsite power (SLB-LOOP).
The benchmark analysis should be provided for staff review to confirm the acceptability of the SLB-LOOP analysis results.
2.
Since the transition mixed core DNBR penalty is accommodated by setting the 0FA enthalpy rise peaking limits at a level equal to 96 percent of the Mark-BW peaking limits without increasing the design DNBR limit, the mixed core penalty should be assessed to the 0FA assemblies if a reload design is such that the entholpy rise factor of the 0FA fuel is higher than 96 percent of that of the Mark-BW fuel.
3.
Evaluations should be performed for each reload cycle to confirm that the values of the key reactivity parameters are within the bound of those specified in Tables A.1 and A.2 of BAW-10173.
If any of the cycle-specific value is not bounded, new analyses of those transients affected should be performed to confirm the acceptance criteric are met, or a reload design change should be made to ensure the parameters are bounded by the values in the tables.
In addition, a RCCA misalignment analysis should be made for each reload cycle to verify that the DNBR peaking margin loss is less than the available peaking margin at full power.
l 4.
Since no analysis was provided for the boron dilution accident, an analysis should be provided in the first reload analysit report with the Mark-BW fuel, i
5.
Evaluations of radiological consequences should be made for the locked rotor and the single RCCA withdrawal events to show compliance with their respective dose release criteria.
l l
e 17
4.0 REFERENCES
1.
Letter from H. B. Tucker, Duke Power Company, to V. S. Nuclear Ragulatory Comission, "McGuire huclear Station, Docket Numbers 50-369 and -370, Catawba Nuclear Station, Docket Numbers 50-413 and -414, Topical Report BAW-10173P, ' Mark-BW Reload Saf ety Analysis for Catawba and McGuire',"
March 30, 1989.
2.
Letter from H. B. Tucker, Duke Power Company, to V. S. Nuclear Regulatory Comission, "McGuire Nuclear Station, Docket Numbers 50-369 and -370, Catawbe Nuclear Station, Docket Numbers 50-413 and -414, Topical Report BAW-10173P, ' Mark-BW Reload Safety Analysis for Catawba and McGuire',
(TACS 73769-73772)," October 22, 1990.
3.
Letter from M. S. Tuckman, Duke Power Company, to U. S. Nuclear Regulatory Commission, "McGuire Nuclear Station, Docket Numbers 50-369 and -370, Catawba Nuclear Station, Docket Numbers 50-413 and -414, lopical Report BAW-10173P, ' Mark-BW Reload Safety Analysis for Catawba and McGuire',
(TACS73769-73772)," Novenibcr 28, 1990.
4.
Letter from H. O. Tucker, Duke Power Company, to V. S. Nuclear Regulatory Comission, "McGuire Nuclear Station, Docket Numbers 50-369 and -370, Catawba Nucleat Station, Docket Numbers 50-413 and -414, Response to Request for Ac,ditional Information Regarding BAW-10173P, (TACS 73769/73770/73771/73772)," June 7, 1990.
5.
BAW-10169-A, "RSG Plant Safety Analysis - B&W Safety Analysis Methodology for Recirculating Steam Generator Plants," October 1989.
6.
BAW-10164P, Revision 1, "RELAP5/M002-B&W - An Advanced Computer Program for Light Water Reactor LOCA and Non-LOCA Transient Analysis," January 1990.
7.
BAW-10124A, " FLAME 3 -- A Three-Dimensional Nodal Code for Calculating Core Reactivity and Power Distribution," August 1976.
8.
BAW-10152A, "N0ODLE -- A Multidimensional Two-Group Reactor Simulator,"
June 1985.
S.
BAW-10117A, " Babcock & Wilcox Version of PDQ07 -- User's Manual," January 1977.
- 10. BAW-10156-A, "LYNXT - Core Transient Thermal-Hydraulic Program," February 1986.
- 11. BAW-10159, "BWCMV Correlation of Critical Heat Flux in Mixing Vane Grio Fu?1 Assemblies," May 1986.
- 12. BAW-20170P-A, " Statistical Core Design for Mixing Vane Cores," December 1988.
- 13. BAW-1(A63 P-A, " Core Operating Limit Methodology for Westinghouse -
I Designed PWRs," June 1989.
/
_ _ _ _ _ _ _ _ _ _ _ _