ML20082M885
| ML20082M885 | |
| Person / Time | |
|---|---|
| Site: | Surry, North Anna, 05000000 |
| Issue date: | 10/31/1983 |
| From: | Cross R, Erb J, John Miller VIRGINIA POWER (VIRGINIA ELECTRIC & POWER CO.) |
| To: | |
| Shared Package | |
| ML18141A254 | List: |
| References | |
| VEP-NFE-2-(NP), NUDOCS 8312060318 | |
| Download: ML20082M885 (162) | |
Text
,
VEP - NFE - 2 OCTOBER,1983 Vepco VEPCO EVALUATION OF THE CONTROL ROD EJECTION TRANSIENT iT, W$
m eggagggggeggggggo o
PDR p
j POWER STATION ENGINEERING DEPARTMENT NUCLEAR FUEL ENGINEERING Virignia Electric and Power Company
. VEPC0 EVALUATION OF THE CONTROL ROD EJECTION TRANSIENT BY J. G. MILLER AND J. O. ERB NUCLEAR. FUEL ENGINEERING GROUP POWER STATION ENGI!1EERING DEPARTMENT VIRGINIA ELECTRIC AND POWER C0tiPANY RICHMOND, VIRGINI A OCTOBER 1983 RECOMMENUED FOR APPROVAL:
R. W. Cross Supervisor, Nuclear Fuel Engineering APPROVED:
km W-:m-__
R. M. Berryman Director, Nuclear Fuel Engineering 4
O
i pAGE 2
CLASSIFICATION / DISCLAIMER Tho
- data, information, analytical techniques, and conclusions in this ropcrt have been prepared solely for use by the Virginia Electric and pcuer Company (the Company), and they may not be appropriate for use in cituations other than those for which they were specifically prepared.
Tho company therefore maker no claim or warranty whatsoever, express or
- icplied, as to their
- accuracy, usefulness, or applicability.
In porticular, THE COMPANY MAKES NO WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, NOR SHALL ANY WARRANTY BE DEEMED TO ARISE FROM COURSE OF DEALING OR USAGE OF TRADE, with respect to this report or any of the data, information, analytical techniques, or conclusions in it.
By making this report available, the company does not authori=e its use by others, and any such use is expressly forbidden except with the prior written approval of the Company. Any such written approval shall itself ba deemed to incorporate the disclaimers of liability and disclaimers of uerranties provided herein.
In no event shall the Company be liable, under ar.y legal theory whatsoever (whether contract, tort, warranty, or otrict or absolute liability),
for any property damage, mental or Physical injury or
- death, loss of use of property, or other damage roculting from or arising out of the use, authori=ed or uncuthori=ed, of this report or the
- data, information, and analytical techniques, or conclusions in it.
PAGE 3
a ABSTRACT This report describes the methods developed by the Virginia Electric and
(
I pcuer Company (Vapco) for the analysis of the postulated control rod ojoction transient.
The principle calculational tool used in this davelopment has been the RETRAM transient thermal-hydraulics code; a point kinetics core physics option is used for predicting the system response of a
plant undergoing a rod ejection event. The local fuel ta=perature response is calculated using a separate RETRAM model of the core hot' spot. comparison with the results from vendor methodologies are included to demonstrate the acceptability of the Vepco methodology for uso in the reload core safety analysis and licensing process.
______________m_.
PAGE I4 ACKMOWLEDGEMENTS Tho authors would like to express their thanks to Messrs.
M.
A.
- Smith, M.
p.
- Wolfhope, R.
W.
Cross c i- +
S.
M.
Bowman for their technical cooistance in the development and preparation of this report. The authors would also like to express their appreciation to a number of poople who reviewed and provided comments on this report.
PAGE 5
TABLE OF COMTENTS Title Page CLASSIFICATI0M/ DISCLAIMER 2
ABSTRACT 3
ACKMOWLEDGEMENTS 4
TABLE OF COMTENTS 5
LIST OF TABLES 7
LIST OF FIGURES 8
SECTION 1 - INTRODUCTION 11 1.1 Purpose and Organization'of the Report 11 1.2 Description of the Tran'sient 14 1.3 Acceptance Criteria 16 SECTION 2 - METHODOLOGY 18 2.1 Calculational Model -- The RETRAM Code 18 2.1.1 RETRAM Single Loop Model 23 2.1.2 RETRAM Hot Spot Mcdel 37 2.2 Calculational Technique 47 2.2.1 Steady State Physics Analysis 48 2.2.2 Core Average Transient Analysis 52 2.2.3 Power Weighting Factor 57 2.2.4 Hot Spot Transient Analysis 63 2.2.5 System Overpressure Analysis 67 2.2.6 Radiological Concerns 68 SECTION 3 - SENSITIVITY STUDIES 69 3.1 Introduction................................
69
o PAGE 6
I 3.2 Sensitivity Study Results 70 3.2.1 Point Kinetics Neutronics Parameters 70 3.2.2 Point Kinetics Model Thermal Hydraulic Parameters 95 3.2.3 Mot Spot Parameters 101 SECTION 4 - VERITICATION COMPARISONS 110 4.1 Introduction................................
110 4.2 Vendor Licensing Methodology 112 4.3 Verification with Licensing Analysis 119 4.3.1 Core Average Power History 121 4.3.2 Hot Spot Analysis 130 4.3.3 Conclusions 139 4.4 Comparison to Three-Dimensional Space-Time Kinetics 140 4.4.1 Three-Dimensional Model 141 4.4.2 Comparison Results 149 SECTION 5 -
SUMMARY
AND CONCLUSIONS 160 SECTION 6 - REFERENCES 161
PAGE 7
LIST OF TABLES Table Title page 2-1 Thermal-Hydraulic Design parameters 31 2-2 Single Loop Model control Volume Description 32 2-3 Single Loop Model Junction Description 33 2-4 Single Loop Model Trip Description............
35 2-5 Hot Spot Model Heat Transfer Correlations 42 2-6 Hot Spot Average Fuel Temperature and Enthalpy 46 2-7 Hot Channel Fuel Melt Fraction Table 66 3-1 point Kinetics Meutronics Sensitivity Study 79 i
l 3-2 point Kinetics Thermal Hydraulic Sensitivity I
Study 97 3-3 Hot Spot Sensitivity Study 106 4-1 Comparison of Vendor /Vapco Licensing Methodologies 116 4-2 Vendor /Vepco Analysis Results Using Vendor Methodologies 118 4-3 Verification Comparison cases 120 4-4 FACTRAM/RETRAN Hot Spot Model Comparisons 138 4-5 3-D Comparison Cases 148 4-6 3-D Hot Spot Model Comparison Results 153 l
i
PAGE 8
LIST OF FIGURES Figure Title Page 2-1 Vepco RETRAN Single Loop Model 30 2-2 Vepco RETRAN Hot Spot Model 41 2-3 RETRAN Control Model for Thom Correlation 44 2-4 RETRAM Control Model for Bishop-Sandberg-Tong Correlation...................................
45 2-5 Normalized Trip Reactivity Curve 56 2-6 Power Weighting Factor 62 3-1 Sensitivity Study -- HZP Doppler Reactivity Feedback (Power History) 83 3-2 Sensitivity Study -- HZP Doppler Reactivity Feedback (Energy Release) 84 3-3 Sensitivity Study -- HFP Doppler Reactivity Feedback 85 3-4 Sensitivity Study -- Moderator Reactivity Feedback 26 3-5 Sensitivity Study -- Delayed Neutron Fraction 37 3-6 Sensitivity Study -- Ejected Rod. Worth 88 l
3-7 Sensitivity Study -- Decreased Time of Ejection 89 3-8 Sensitivity Study -- Increased Time of Ejection 90 i
l 3-9 Sensitivity Study -- Trip Delay Time 91 3-10 Sensitivity Study -- Trip Worth 92 l
l 3-11 Sensitivity Study -- Initial Zero Power Level 93 1
3-12 Sensitivity Study -- Beta Yield Fractions 94 i
1 l
A PAGE 9
3-13 Sensitivity Study -- Gap Heat Transfer Coefficient............................
?'
3-14 Sensitivity Study -- Point Xinatics Fuel Pin Geometry 100 3-15 Sensitivity Study -- Hot Spot Fuel Pin Geometry 108 3-16 Sensitivity Study -- Hot Spot Pellet Temperature Distribution 109 4-1 Nuclear Power Transient -- S1CS BOL HZP 124 4-2 Muclear Power Transient -- S1C5 BOL HFP 125 4-3 Nuclear Power Transient -- 51C5 EOL HZP 126 4-4 Nuclear Power Transient -- S1C5 EOL HFP 127 4-5 Nuclear Power Transient -- S+MTC BOL HZP 128 4-6 Nuclear Power Transient -- S+MTC BOL HTP 129 4-7 Hot Spot Fuel Centerline Temperature Transients
-- S1C5 BOL Cases 132 4-4 Hot Spot Fuel Centerline Temperature Transients
-- SICS EOL Cases 133
- i-9 Hot Spot Fuel *=nterline Temperature Transients
-- 5+MTC Cases 134 4-10 Hot Spot Fuel Outer Clad Temperature Transients
-- S1C5 BOL Cases 135 4-11 Hot Spot Fuel Outer Clad Tempersture Transients
-- 51C5 EOL Cases 136 4-12 Hot Spot Fuel Outer Clad Temperature Transients
-- S+MTC Cases 137 4-13 Surry Unit 1 Cycle 1 Core Loading Plan 144
PAGE 10 l
l 4-14 Radial Geometry for 3-D TWINKLE 145 4-15 Steady State Radial Power Distributions 146 4-16 Muclear Power Transient -- 3-D Benchmarks 154 4-17 Total Energy Release -- 3-D Benchmarks 155 4-18 Hot Spot Power History -- 3-D Benchmarks 156 4-19 Hot Spot Fuel Centerline Temperature Transients
-- 3-D Benchmarks 157 4-20 Hot Spot Fuel Outer Clad Temperature Transients
-- 3-D Benchmarks 158 6
PAGE 11 SECTION 1 - INTRODUCTION 1.1 purpose and Organization of the Report This report documents the methodology developed by the Virginia Electric cnd Power Company (Vepco) for the anslysis of the postulated control rod ojection transient for the North Anna and Surry Muclear Power Stations.
The intent of this methodology is to provide Vapco with the capability of performing the licensing analysis required for the Condition IV rod ojection transient addressed in the Final safety Analysis Report.
The purpose of this report is to provide an acceptable reference for safety and libeNsing analysis of the rod ejection transient for a reload core.
This analysis is performed in order to dem _, crate that an occurrence of the transient for the core loading in question will noither interfere with the core cooling capability nor result in fuel daange which will lead to an unacceptable radiation release within the guidelines of 10 CFR part 100, " Reactor Site Criteria." The approach cdopted for verification of this transient analysis method is to compare rosults obtained with the Vepco methodology to results obtained by Vepco using an accepted vendor methodology. These results are demonstrated to b3 conservative by comparison to Vepco results based on a
three-dimensional space-kinetics 'model and a detailed hot spot thermal hydraulic model.
The Intreduction Section of this report presents a statement as to the Purpose of the report, a description of the rod ejection transient, and a discussion of the accepta'nce criteria which must be met by an analysis
PAGE 12 to insure the safe operation of the plant in the event such a transient occurs.
Soction 2 provides detailed descriptions of the calculational model used end the methods by which it is employed to analyce the transient in a ocnservative manner. The licensing analysis is performed in two parts:
1.
a point kinetics analysis to calculate the average core nuclear powar history, and 2.
a hot spot thermal-hydraulic calculation to determine the hot spot enthalpy and temperature transients from which the amount of fuel damage and radiological consequences of the accident may be assessed.
Both of the analyses are performed with the RETRAM transient thermal-hydraulic analysis code, but with different code models for each pcrt.
Description of the two RETRAM models, henceforth referred to as the TITRAM Single Loop Model for the average core nuclear power history ociculation, and the RETRAM Hot Spot Model for the hot spot transient cciculation, are provided in Section 2.
The actual techniques employed in the analysis are described in the letter part of Section 2, including the additional concerns of providing core physics parameters for the RETRAM models for the specific core loading to be analyzed, and the investigation of the system overpressure end radiological concerns.
The use of point reactor kinetics instead of spatial kinetics in the core average pcuer history analysis allows for application of a
w3ighting factor to the Doppler reactivity feedback model used in the
PAGE 13 point kinetics analysis. This weighting factor, henceforth referred to oc the power Weighting
- Factor, provides a
more accurate, although conservative, estimate of the actual Doppler reactivity feedback to be expected during the transient. A discussion of the derivation and use of this weighting factor is also presented in Section 2.
Tha results of a series of sensitivity studies performed with the RETRAM cedels used for the rod ejection study are provided in Section 3.
These consitivity studies are used to quantify the impact of uncertainties in important core parameters and modeling assumptions on the models' prodictions.
Ccaparisons of the results of a vendor methodology for standard FSAR and roioad core licensing analyses with those of the Vepco developed cathodology are providad in Section 4
to further quantify the ccceptability of the Vapco approach. In addition, a comparison of power history calculations between the point kinetics RETRAM model and a three-dimensional space-time kinetics model is* presented to demonstrate tha conservatism of the point kinetics approach.
Tha report's. conclusions and references are provided in Sections 5 and 6 roopectively.
PAGE 14 1.2 Description of the Transient The rod ejection transient is a postulated condition IV event initiated by the mechanical in11ure of a control rod mechanism pressure housing.
Following such a failure, the action of the coolant pressure is assumed to result in a
rapid ejection of a rod cluster control assembly and drive shaft from the core to a fully withdrawn position within a time interval on the order of 0.1 seconds. This in turn leads to a fast roactivity insertion and may cause a
severe asymmetric core power distribution, possibly leading to fuel rod damage.
Roactor protection for the transient is provided by negative reactivity foedback effects and by reactor trips on high neutron flux levels. Trip cotpoints for the Surry and North Anna nuclear units are 118% of full pcuer (including setpoint and instrument errors) for reactor operations at or above 10% full power, and 35% of full power (including setpoint and instrument errors) for reactor operations below 10% full power.
After an appropriate delay, the rod cluster control assemblies assumed cvailable (less the eje'cted rod) are assumed to drop into the core.
Hcuever, before such rod movement is initiated, the large power surge of the core is turned around by the negative Doppler reactivity feedback rosulting from the quick rise in the core's fuel temperature, particularly in the vicinity of the ejected rod. This proves to be the dominant effect in limiting the consequences of the transient.
In general, the core loading for each reload cycle is designed so as to limit the amount of reactivity insertion which would result from the ojection of a
rod cluster control assembly from any of the locations
i PAGE 15 where one would he inserted during normal power operation.
The ocnsequence of such a rod ejection event would be a very rapid increase in core average power level, with an accompanying core pressure surge end a
particularly severe temperature transient in the vicinity of the ojected rod.
This localized temperature transient may be severe enough to cause some of the fuel to experience DNB (Departure from Nucleate Bailing) and localized fuel damage. The degree of potential damage will b3 mainly governed by the worth of the ejected rod.
Additional details on the postulated transient's description, available protection mechanisms, and consequences may be found in the appropriate Updated TSARS (UTSARs) for the Surry and North Anna Nuclear power Stations (Refs. 1 and 2.)
\\
PAGE 16 1.3 Acceptance criteria AD detailed in USNRC Regulatory Guide 1.77 (Ref. 3), the acceptance criteria to be used in evaluating the re s9:lts of a rod ejection transient analysis are 1.
Reactivity encursions will not result in a radial average fuel enthalpy greater than 280 cal /g (504 Btu /lb) at any axial location in any fuel rod.
2.
Maximum remotor pressure during any portion of the assumed transient will be less than the value that will cause stresses to exceed the Emergency condition stress limits as defined in section III of the ASME Boiler and pressure Vessel Code.
3.
offsite dose consequences will be well within the guidelines of 10 CFR part 100, " Reactor Site Criteria."
It should be noted that the 280 cal /g criteria applies to both irradiated and unizradiated fuel in providing a conservative maximum limit to limit fuel damage from the prompt pin burst phenomenon and not inpair significantly the core cooling capabilities.
Additional in-house design limits for acceptability of the transient cnalysis for a reload have been imposed by Vepco. These limits are:
1.
peak clad temperature less than or equal to 2700
'r, 2.
fuel centerline melting less than or equal to 10% at the hot
- spot, 3.
average hot spot fuel enthalpy less than 225 cal /gm (405 Btu /lb) for unirradiated fuel, and
=
PAGE 17 4.
average hot spot fuel enthalpy less than 200 cal /gm (360 Btu /lb) for irradiated fuel.
Tho peak clad temperature limit of 2700
'T reflects a conservative cotimate of the temperature at which clad embrittlement may occur. These edditional criteria reflect more severe limits than those delineated in Roghlatory Guide 1.77, and are consistent with the limits applied in the dooign of both the Surry and North Anna Nuclear Power Stations.
The predicted results of an analysis which fall within the acceptability limits of these criteria demonstrates that the consequences'of a rod ojoction transient for the specific core reload design will be coceptable to the safety of the general public and will maintain the integrity of the core cooling capability, i
--,,r---
,+m,-e---~
-=v=-=
=---+-*-e+-
e
--"*-e - '
PAGE 18 SECTION 2 - METHODOLOGY 2.1 Calculational Model -- The RETRAN Code The principal calculational tool used by Vapco for the rod ejection transient analysis is the RETRAM computer code. The RETRAM code has been doveloped by Energy Incorporated (EI) under the auspices of the Electric pcuer Research Institute (EpRI) as a generalized and versatile computer code for the analysis of light water reactor systems. The theory and numerical algorithms, programming
- details, user's input manual, and oxamples of its applications to Muclear Steam Supply Systems (NSSS) are thoroughly documented for the RETRAN code in Volumes 1 through 4 of Ref.
4.
The RETRAM computer code is based upon the RELAp4/003 Update 85 code which was released by the United State Nuclear Regulatory Commission as port of the Mater Reactor Evaluation hodel (WREM), (Ref. 5).
The RETp.AN fluid ' differential and state equations used to represent homogeneous equilibrium flow in one dimension are described in Ref.
4.
The representations used in previous RELAp codes for control volumes and l
t junctions are also used in RETRAN and allow the analyst to model a l
system in as much detail as dasized. The modeling flexibility of the code is important and will be described in more detail in Sections 2.1.1 l
~
below.
l cnd 2.1.2 I
The equation
- systems, which describe the flow conditions within the
- channels, are obtained from the local fluid conservation equations of l
- adss, momentum and energy by use of mathematical integral-averaging
PAGE 19 j
tochniques.
Forms of the momentum equation are available for both compressible and incompressible flow.
The heat conduction representation capabilities of RETRAN have been increased over previous RELAp versions. The principal augmentation to RETRAM is the capability to more accurately calculate two-sided heat transfer. The appropriate heat transfer correlation is selected bared on thermodynamic conditions in each of two flow streams, on either side of i
l n
heat conducting solid.
Consequently, representations of the heat transfer processes occurring in the steam generator, for example, are more accurate than previously possible.
Roactor kinetics are represented in RETRAM using a point kinetics model with reactivity feedback. The reactivity feedback can be represented by constant coefficients or in tabular form and accounts for explicit control actions (e.g.,
rod scram) and changes in fuel temperature, moderator temperature and density, and soluble boron concentration.
The system component models utilized in RETRAM include a pump model that describes the interaction between the centrifugal pump and the primary system
- fluid, and valve models that represent either simple valves, check valves or inertial valves.
The flexibility of the valve representation and their configuration is important in allowing a wide variety of options to the user for the modeling of system dynamics.
Soveral representations of heat exhangers can be modeled by the code.
These include the previously discusssed two-sided heat transfer and one-sided heat transfer in conjunction with user specified boundary conditions.
A non-equilibrium pressurizer can be modeled in which the
PAGE 20 thermodynamic state solutions of the liquid and vapor regions of the pressuriger are determined from a distinct mass and energy balance for occh region.
AD in
- RELAp, a variety of trip functions can be modeled in the RETRAM code to represent various reactor azotection system actions.
A refinement of the RETRAM code over the RELAP4 code is the addition of a roactor control system modeling capability. Consequently, the dynamics of linear and non-linear control systems are represented with RETRAM models of the more common analog computer elements. This additional capability is necessary for both best-estimate and licensing analysis, since the response of various control and protection systems may have a cignificant effect on the overall system response.
The analysis of the rod ejection transient has been performed by Vepco with the RETRAM-02 version of the code. The second major code version of
- RETRAN, RETRAM-02 was released to allow for additional capability in codeling certain BWR transients, small break LOCAs, balance of plant modeling and anticipated transients without scrams. Volume 1 of Ref. 4 provides a
detailed summary of the changes and new capabilities available with RETRAM-02.
One of these new capabilities is a one-dimensional kinetics model which allows a
spatial kinetics modeling of the core in the axial direction (that is, along the axis of fluid flow through the core) to be used in place of the simpler point kinetics model. Since this is the dimension through which the rod is ejected, the one-dimensional kinetics model cppears to offer the advantags of a mor9 accurate modeling of the rod i
PAGE 21 ojection event in place of the point kinetics model. However, since the primary power redistribution effect which contributes to the negative roctivity feedback in the transient is in the radial
- plane, no
\\
cignificant advantage over using the point kinetics model is obtained by ucing the one-dimensional model. On the other hand, the simplicity of une and the flexibility of the point kinetics model compared to the ene-dimensional kinetics model make the formarithe preferential choice for modeling this transient.
The licensing analysis for the rod ejection event as performed by Vepco is divided into two parts--a point kinetics analysis to calculate a conservative core average power
- history, and a
prediction of the onthalpy and temperature transients in the hot spot of the core based on that power history.
Tha core average power history calculation is Porformed using the point kinetics option of the RETRAi*-01 code onploying the standard Vepco Single Loop Model. This model, as described balou in Section 2.1.1, is further documented in Raf. 6 for performing other non-LOCA transient analyses as well as best estimate analyses for plant operational support.
The core average power history as calculated with the RETRAM Single Loop Mcdel is weighted by the maximum total power peaking factor for the transient to conservatively estimate the power history of the hot spot.
This data is input to a detailed fuel heat transfer model of the hot spot
- location, denoted as the RETRAN Hot Spot Model, to determine the hot spot temperature and enthalpy history as a
function of time following rod ejection. Results from this analysis are used to ascertain
l PAGE 22 the extent of the fuel damage, if any, and the radiological consequences of the transient.
These two RETRAM models will be described in detail belou, followed by a discussion of their application to the actual transient analysis.
9
l l
l pAGE 23 2.1.1 RETRAN Single Loop Model For the rod ejection transient, the system thermal-hydraulic response of 1
)
ell reactor coolant loops is essentially identical. Hence, a single loop ropresentation of the NSSS suffices. Furthermore, the quickness of the core response during the transient is such that the major changes in core parameters have all taken place in the time interval of e. bout 10
- coconds, approximatoly the time for the coolant to make one complete pass through the primkry coolant loops. Therefore, little impact on the MSSS outside of the core is expected during the rod ejection transient.
Essentially, only a
modeling of the reactor core need be performed in order to predict the core average power history with sufficient conservatism for input to the Hot Spot Model.
However, the ready availability and extensive documentation of the Vepco RETRAM Single Loop Model make it an ideal candidate for use in performing the first part of tho transient analysis.
Therefore, this model is used as part of the standard Vepco methodology for the rod ejection transient.
Vapco RETRAN Single Loop Models for either the Surry or North Anna Nuclear power Stations are similarly constructed. Both stations consist of two identical operational nuclear units.
All four units are Wastinghouse designed three coolant loop pressurized water reactors with core thermal ratings of 2441 Mut for the Surry units and 2775 Mut for the North Anna units.
The three similar heat transfer loops are connected in parallel to the reastor vessel with each loop containing a contrifugal
- pump, loop stop valves and a steam generator. The system includes a
pressurizer and the associated control system and
.I
PAGE 24 instrumentation necessary for operational control and protection.
The reactor vessel encloses the reactor core consisting of 157 fuel I
assemblies with each Surry assembly having 204 fuel rods and 21 thimble tubes arranged in a 15 x 15 array while each North Anna assembly has 264 fuel rods and 25 thimble tubes arranged in a 17 x 17 array. The fuel for both stations consists of slightly enriched uranium dioxide fuel pellets contained within a
Zircaloy-4 cladding. General thermal and hydraulic design parameters for the reactor systems are listed in Table 2-1.
T,he RETRAM thermal hydraulic model is formulated by representing i
individual portions of the hydraulic system as nedes or control volumes.
Control volumes are specified by the thermodynamic state of the fluid within the volume and basic geometric data such as volume, flow area,
'oguivalent diameter and elevation. The flow paths connecting the volumes or boundary conditions associated with a
volume are designated as junctions.
Junctions are described by specifying the flou, flow area, olevation, effective geometric inertia, form loss coefficient and flow equation specification for that particular flow path.
Thermal interactions with system metal in the MSSS are modeled with heat conductors.
Heat conductors may represent heat transfer from passive sources such as the metal of the reactor coolant system piping or the steam generator tubes. In addition, the internal generation of heat in the core may be represented by active heat conductors designated as powered conductors. Heat conductors are primarily specified by providing the heat transfer area, volume, hydraulic diameter, heated equivalent diameter and channel longth of the particular part of the system being
PAGE 25 l
sodeled.
Temperature dependent material properties (specific
- heat, thermal conductivity and linear thermal expansion coefficient) are also input.
In
- general, the basic NSSS model is formulated with the code capabilities discussed above. An extensive research effort was conducted to determine the appropriate input required for the models of the Surry cnd Morth Anna units. Information was obtained from plant drawings, the Final Safety Analysis
- Reports, Vepco internal operating documents, cquipment technical manuals and specific information requested from the MSSS vendor.
Specific control capabilities and constitutive models of system components will be discussed in the following paragraphs.
The Vepco RETRAM Single Loop Models of the North Anna and Surry nuclsur units represent the three actual reactor coolant loops as one loop. The rosulting geometry is provided in Figure 2-1 and consists of 19 volumes, 29 junctions and 7 heat conductors. While the specific model input for
-the Surry and North Anna plants is different, the basic model doscription is the same for the single loop models of both plants. The reactor vessel includas representation of the downcomer, upper and lower
- plenums, core bypass and reactor core.
The steam generator is represented by four volumes on the primary side, one volume or. the socondary side and four heat conductors representing the tubes. Single volumes represent the hot leg piping, steam generator inlet plenum, pump suction piping, reactor coolant pump, cold leg piping, pressuri=er, and pressurizer surge line. Primary system boundary conditions are specified with junctions representing the pressuri=er relief and safety valves.
Junctions representir.g the feedwater inlet, steam outlet, atmospheric s te am*.
relief and steam line safety valves provide secondary system 4.
e
PAGE 26 boundary conditions, Specific aspects of the basic model will be discussed below.
All control volumes in the model are homogeneous with the exceptions of volumes 17 and 19, the pressurizer and secondary side of the steam gcnerator, which ocatnin two-phase mixtures. Volumes modeling the loop Piping use the RETRAM ' temperature transport delay" option to represent fluid temperature change movements in the loop as a front, (that is, fluid entering a
pipe does not mix with the fluid present but instead displaces it.)
All junctions specify single-stream, compressible flou oxcept for junction 21, the pressuri=er surge line connection to the
- loop, for which incompressible flow with no momentum flux is specified.
Extended Henzf (subcooled) and Moody (saturated) choking is assigned for 011 junctions.
All junctions use Baroczy two phase multipliers with Fenning friction to define the wall friction except for the four junctions on the secondary side of the steam generator (volume 19) which use a
homogenous two-phase multiplier with Fanning friction. All heat conductors use the Dougall Rohsenow heat transfer correlation to doscribe post-DNB heat transfer. Tables 2-2 and 2-3 summari=e the volume and junction descriptions.
l The RETRAM code contains several system component models which are used in the Vepco Single Loop Model. These include pump models which describe the interaction between the centrifugal pump nnd the primary system fluid.
These models calculate pump behavior through the use of i
onpirically developed pump characteristic curves which uniquely define l
the head and torque response of the pump as functions of volumetric flou I
1 l
i
PAGE 27 cnd pump speed.
RETRAM includes " built-in" pump characteristics which ore representative of pumps supplied by the major reactor coolant pump canufacturers. These curves may be modified, as appropriate, by the user to more realistically represent a specific pump design. Although the built-in data are not appreciably different from Vepco's plant-specific
- curves, Vepco's Single Loop Models incorporate the specific head versus flow response for first quadrant operation found in the units' UFSARs (Refs. 1 and 2).
The Sing'le Loop Model incorporates the RETRAM pressuri=er model which dofines two separate thermodynamic regions that are not required to be in thermal equilibrium.
A non-equilibrium capability is particularly nocessary when the transient involves a surge of subcooled liquid into the pressurizer.
In
- addition, the Single Loop Model represents the offects of subcooled spray, electrical immersion heaters, liquid droplet reinout and vapor rise in the pressurizer.
The reactor systems trip logic is modeled to the detail required for a specific analysis. RETRAM trip functions are used to model 1) protective functions, such as the overtemperature delta-T trip, which result in roactor
- scram, 2) control system bistable element
- logic, such as coinci'dence trips which model " majority" logic and 3) general problem centrol (e.g.,
problem termination, etc.)
The protective function trips modeled in the standard Single Loop Model include:
1.
High flux 2.
Overtemperature delta-T
pAGE 28 i
l 3.
Overpower delta-T i
l 4.
Lou /high pressurizer pressure l
l 5.
High pressurizar level l
6.
Low coolant flou 7.
Loss of power to the reactor coolant pumps The Single Loop Model also incorporates the RETRAM control system ccpability to model the following NSSS control and protection features:
1.
Overtemperature delta-T setpoint 2.
Overpower delta-T setpoint 3.
pressure controller 4.
Lead / lag compensation of the lou pressure trip signal.
Tchle 2-4 presents a summary of the trips in the standard Single Loop Models.
I l
The core power response is determined by the point kinetics model in conjunction with explicit reactivity forcing functions and thermal foedback effects from moderator and fuel in the three core regions. The l
l point kinetics model specified for the Single Loop Model incorporates t
eno prompt neutron
- group, six delayed neutron groups with decay heat represented by 11 gamma
- emitters, and the important radioactive cctinides U-239 and Np-239.
Explicit reactivity forcing functions represent scram and reactivity insertion due to control rod withdrawal in the Single Loop Model as the particular analysis requires. Constant I
tasperature coefficients or reactivity tables as a
function of tanperature (fuel),
density (moderator) or power represent feedback offects.
Core power is distributed axially among the three core l
t
pAGE 29 conductors approximating a symmetric cosine shape. Three 'aore materials ragions are used to represent uranium dioxide fuel pellets, the helium filled gap and the Zircaloy cladding.
Direct moderator heating is appropriately accounted for in the model. The transient fuel and clad tamperatures are calculated based on temperature-dependent thermal properties, which are input in tabular form. Additional details on the point kinetics model as specifically used to analy=e the rod ejection transient are provided in Section 2.2.2.
6
PAGE 30 FIGURE 2-I VEPCO RETRAN. SINGLE LOOP MODEL t
?
STEAM STEAMLINE RELIEF SAFETY VALVES VALVES STEAM ll b
d 28 SAFETY 26 29 23 VALVES OPERATED JL 27 y, egg 2~4.
- - SeRAY a
- z 25 (]l9 5 \\\\%\\ww s; h wwxw3
~
O,S
@R$
=
.J4 6
4
.Y l
- 2'2 f, HEATERS M
FEEDWATER (p
20 Y
h 7--
@3 d
y a2 2,
l 0)
'f '
,[Ll7O {l6 I
l l
qo) i 8
18 10 5
If h
SPRAY b
INTA KE 3h @gi4 d
i.e 1 55 d
llN'Il2 13 P/
LEGEND wOIRECTION OF FLOW O v0tuMeS E HEAT CONDUCTORS S
PAGE 31 TABLE 2-1 THERMAL-HYDRAULIC DESIGN PARAMETERS SURRY PLANT:
Total core heat output, Mut 2441 Hoat generated in fuel, X 97.4 i
System operating pressure, psi 2250 Total coolant flow rate, gym 265500 Coolant temperatures,
'T (J 100% power)
Nominal inlet 543 Average rise in core 65.3 Average rise in vessel 62.6 Core average 577.0 Vessel average 574.3 Average linear power density, Kw/ft 6.2 NORTH ANNA PLANT:
Total core heat output, Mut 2775 Hsat generated in fuel, X 97.4 System operating pressure, psi 2250 Total coolant flow rate, gym 278400 Coolant temperatures,
'F (J 100% power)
Nominal inlet 549.5 Average rise in core 69.4 Average rise in vessel 66.6 Core average 586.1 Vessel average 582.8 l
Average linear power density, Kw/ft 5.4 t
'l l
w e,
,w
-w a
PAGE 32
(
TABLE 2-2 SINGLE LOOP MODEL CONTROL VOLUME DESCRIPTION Volume Mixture Temperature ID Description Type Transport Delay 1
Vessel upper plenum H
No 2
Reactor hot leg H
Yes 3
S/G inlet plenum H
No 4
S/G tube volume 1 H
No 5
S/G tube volume 2 H
No 6
S/G tube volume 3 H
No 7
S/G tube volume 4 H
No 8
Pump suction piping H
Yes 9
Reactor coolant pump H
No 10 Reactor cold leg H
Yes 11 Douncomer H
Yes 12 Vessel lower plenum H
Ho 13 Core bypass H
Yes 14 Core section 1 H
Ho 15 Core section 2 H
No 16 Core section 3 H
Ho 17 Pressuri=1r N
Ho 18 Pressuri=er surge line H
Yes 19 S/G secondary side T
No Abbreviations:
S/G - steam generator H - homogeneous equilibrium M - two-phase non-equilibrium T - tuo-phase equilibrium
PAGE 33 TABLE 2-3 SINGLE LOOP MODEL JUNCTION DESCRIPTION Two-Phase Fanning Junction Friction Valve ID Description Type Multiplier Index H/V 1
Vessel outlet nozzle Normal Baroczy No V
2 Hot leg outlet Normal Baroc=y Yes H
3 S/G inlet planum Normal Baroczy No H
4 S/G tubes Normal Baroc=y No H
5 S/G tubes Normal Baroc=y No V
6 S/G tubes Normal Baroczy No H
7 S/G-pump suction Normal Baroc=y No H
8 pump intake Normal Baroc=y No H
9 Pump discharge Normal Baroczy No V
10 Vessel inlet nozzle Normal Baroc=y No V
11 Downconer outlet Normal Baroczy No H
12 Bypass inlet Normal Baroc=y No H
13 Lower plenum - core Normal Baroczy No H
14 Core internal Normal Baroczy No H
15 core internal Normal Baroc=y No H
16 Core - upper planum Normal Baroc=y No H
17 Bypass outlet Normal Baroc=y No H
18 Cold leg spray intake Fill Baroc=y No V
19 Przz. spray Spray Baroc=y No H
20 Przz. - surge line Normal Baroc=y No H
21 Surge line - hot leg Normal Baroc=y No H
22 Feeduater fill Fill Baroczy No V
23 S/G outlet Fill Homog.
Yes H
24 PORV 1 Fill Baroczy No H
25 PORV 2 Fill Baroczy No H
PAGE 34 TABLE 2-3 (cont.)
SINGLE LOOP MODEL JUNCTION DESCRIPTION Two-Phase Fanning Junction Friction Valve ID Description Type Multiplier Index H/V 26 S/G atm. steam relief Fill Homog.
No H
27 Przr. safety valve Fill Baroczy No H
28 Steamline safety valve 1 Fill Homog.
No H
29 Steamline safety valve 2 Fill Homog.
No V
Motes All junctions hava single-stream compressible flou except junction 21 which is incompressible flow.
Abbreviations:
PORV - power operated relief valve atm. - atmospheric S/G - steam generator Przt. - pressuriner Homog. - homogeneous V - vertically distributed junction area H - horizontally distributed junction area
^
PAGE 35 TABLE 2-4 SINGLE LOOP MODEL TRIP DESCRIPTION Trip ID Cause of Trip Activation Trip Action l
l 1
End of transient time End calculation 2
High flux (normalized power)
Scram i
3 overtemperature delta-T Scram 4
Overpower delta-T Scram 5
High pressuriser pressure Scram 6
Low pressurizar pressure Scram 7
High pressurizer level Scram 8
Low coolant flow Scram 9
User specified time
- Close loop isolation valves 10 Lou backup,) eater setpoint Turn pressuri=er heaters on 11 High backup heater setpoint Turn pressuricer heaters off 12 User specified time
- Shut off reactor coolant pumps 13 Transient time = 0 see Trip initiali=ation 14 User specified time
- Uncontrolled rod withdrawal 15 User specified time
- Scram 16 High pressuriser pressure Open p0RV # 1 17 Low pressurizar pressure Close PORV 4 1 18 High spray setpoint Open PORV 4 2 19 Low spray setpoint close p0RV # 2 80 High S/G pressure Open atm. steam relief valve I
21 Lev S/G pressure Close atm. steam relief valve i
22 High S/G pressure Open S/G safety valves 23 Low S/G pressure Close S/G safety valves 84 High pressuriser pressure Open pressurizer safety valves 25 Low pressurizer pressure Closa pressuriner safety valves I
l
PAGE 36 TABLE 2-4 (cont.)
SINGLE LOOP MODEL TRlp DESCRIPTION Trip ID Cause of Trip Activation Trip Action f
26 User specified time'*
Turbine trip 27 Low power End calculation 28 Low-lou steam generator mass Scram 29 Low-lou steam generator mass Auxiliary feedwater on i
30 Scram Turbine trip Notest Not applicable for most transiants.
i Abbreviations 8 PORV - power operated relief valve atm. - atmospheric S/G - steam generator i
1 l
l t
-.,---,,,,-,--,--,e,
PAGE 37 2.1.2 RETRAM Hot Spot Model With a conservative prediction of the core average power history during a
postulated rod ejection event obtained from the RETRAM Single Loop
- Model, a RETRAM Hot Spot Model is used to predict the thermal-hydraulic rosponse of the fuel for the hot spot location of the core. The Hot Spot Madel describes a segment of a single fuel rod at the location where the paak core power occurs during the transient. Using the core average Power history as a
basis for determining a conservative peak power history for the hot spot location, this information is input to the Hot Spot Model as a
driving function.
The resulting fuel and clad tamperatures and enthalpy history for the hot spot as predicted by the cedel are then used to ensure that the fuel melt and clad embrittlement limits have not been exceeded for the transient.
The Vepco Hot Spot Model consists of 2 control volumes, 2 flou junctions and a
single heat conductor. The volumes are stacked one on top of the orher as shown in Figure 2-2.
The cross sectional area of each volume equals that of a single channel. The lower volume models a one foot high hot channel section of the core surrounding the one foot high heat conductor.
Ups'tream from this hot volume is a one-foot high unheated, time-dependent volume which serves as a reservoir for the flow from the hot channel. The hot channel's volume, flow area and hydraulic diameter cre based on the nominal unit cell dimensions for either the Surry or North Anna nuclear
- cores, depending on which unit is undergoing cnalysis. Both control volumes are homogenous. pressure and enthalpy are specified for both volumes for the initial pre-transient core average
PAGE 38 conditions for the appropriate unit and initial power level.
Both junctions specify single-stream, compressible flow and use an icoenthalpic model for choking, (should the model predict choked flow at the junction.)
The junctions have a hori=ontally distributed junction crea and assume no two phase flou wall friction multipliers. Junction 1, which feeds into the bottom of the hot volume, is a fill junction which cupplies flow to the hot volume and has the same flow area. Junction 2, which connects the hot volume to the~ time-dependent volume, is a normal junction.
Fluid properties tables (dynamic viscosity, specific heat and thermal conductivity) are specified for the coolant based on a pressure of 2250 paia which is the nominal operating pressure for both the Surry and North Anna cores. These tables are used by the control system models to cc1culate post-DNB heat transfer as discussed below.
The fluid properties are derived from Ref.
7.
Material properties for the fuel, fuel-clad
- gap, and Zircaloy clad along with an enthalpy table for the fuel as a
function of temperature are built into the model. These Property tables are essentially the same as those provided in the RETRAN Single Loop Model with the exception of covering a higher temperature j
range and allowing for changes which occur upon melting of the fuel. The material properties are derive ( from Ref.
8.
Three core materials regions are used to represent the fuel pellet, the holium filled gap and the =ircaloy cladding. The fuel pellet consists of i
10 concentric mesh spacings of equal radial si=e, the gap of a single mash spacing and the cladding of 3
mesh spacings of equal si=e. A l
l
PAGE 39 pcrabolic power distribution is assumed through the pellet (Ref.
2.
Soction 15.4.6.2.1.2) based on a fuel enrichment which is conservative (high) for the fuel in the reload core loading under analysis.
The heat transfer correlations used for the surface of the heat conductor are modeled with the RETRAM code's control system. Initially the Thom correlation for nucleate boiling (Ref.
4.
Vol. 1) is used.
- However, at 0.1 seconds into the transient, departure from nucleate boiling (DNB) is assumed to occur and a trip centrol in the model switches the heat transfer correlation to the Bishop-Sandberg-Tong film boiling heat transfer correlation, (Ref. 10.) These two heat transfer correlations are summari=ed in Table 2-5 while Figures 2-3 and 2-4 present an outline of the control block strategy used to model the correlations.
Additional control blocks are used to calculate a fuel average temperature and enthalpy for the hot spot as described in Table 2-6.
Initially the values of the gap heat transfer coefficient and nucleate boiling heat transfer coefficient are adjusted to yield the desired steady state temperature distribution throughout the fuel pin. At 0.1 soconds into the transient the value of the hot spot gap heat transfer coefficient is changed to a very large value (10000 Btu /ftz.hr *F) to nodel a closure of the gap at the hot spot location, (Ref. 17.)
The RETRAM option to calculate the heat generated from an exothermic Zircaloy-water reaction between the coolant and the cladding is included in the model. Since this reaction only occurs in the presence of steam, the inlet enthalpy of the fill junction emptying into the hot control e
,n..-
n-n.
PAGE 40 volume is deliberately ramped to 1000 Btu /lb over a 0.2 second interval to induce steam quality in the volume. (Additional details of this catal-water reaction option are presented in Vol. 1 of Ref. 4.)
Imu um
PAGE 41 FIGURE 2-2 VEPC0 RETRAN HOT SPOT MODEL n
1.0 ft.
2u i r l $ !, l1
.a Y
4 y
ji g Fuel 4 1.0 ft.
JIj', +- Mesh Points
"' d!"'.it 4
5
- e 1 1
'!'3 1^
l
- t d.
Legend: Junction (flow direction shown)
Control Volume Q
Heat Conductor--Fuel Gap x '. N.
Clad
F-I PAGE 42 TABLE 2-5 HOT SPOT MODEL HEAT TRANSFER CORRELATIONS I.
Thom Subcooled Boiling correlation (Transient time 0.0 to 0.1 seconds.)
hnb = qu/(Tu - To) hnb = nucleate boiling heat transfer coefficient, Btu /ft2-hr
- F qu = uall heat flux, Btu /ft2-hr Tu = uall temperature,
'T 1/2 where Tu - Tsat + 0.072*expC-P/1260)*qu Tsat = coolant saturation temperature,
'T coolant pressure, psia P =
Tc = coolant temperature,
'T um ais
PAGE 42 TABLE 2-5 (cont.)
NOT SPOT MODEL HEAT TRANSFER CORRELATIONS II. Bishop-Sandharg-Tong Film Boiling Correlation (Transient time > 0.1 seconds.)
0.8 1.23 0.68 0.068
= 0.0193(DG/ )g (Cyp /k)g
(
/f )
({j/$)
C hp.'k )g h = film boiling heat transfer coefficient, Btu /ft*-hr *r D = equivaleat diameter, it k = thermal conductivity, Btu /ft-hr *r
.z.
- Cp = specific heat, Btu /lb
- F j( = viscosity, lb/ft-hr G = mass velocity, lb/ftz-hr 7
= saturated steam density, 1b/ft3 f,
= saturated liquid density, lb/ft3 fg = bulk density, lb/ft3 Subscript f refers to properties at the film temperature, Tfilm, which is defined as 0.5(Tu + Tsat) where Tu = uall temperature,
'T Tsat = coolant saturation temperature,
'T l
l
~
l
~
PAGE 44 i
FIGURE 2-3 RETRAN CONTROL MODEL FOR THOM CORRELATION
=
2 qu MAX CONS (=0.5)
Ensure heat flux > 0.5 BTU /hr it.
e i r sa XPO XPO 0.5
-P/1260
-P/1260 x 0.072 e
i, MULT Calculate wall superheat Tsat SUM Calculate Tu Tu s,
-Tc SUM Motet see Table 2-5 for Parameter definitions.
qu DIV
- hnb
PAGE 45 1
I FIGURE 2 - 14 RETRAM CONTROL MODEL TOR BISHOP-SANDBERG-TOMG CORRELATION Tsat SUM Tu Calculate x 0.5 Tfilm Tfilm
,r TNG TNG TNG Calculate k vs.
CP vs.
- vs.
Fluid Tfilm Tfilm Tfilm Properties E
k Cp 1/p 1/ f u
DIV 4--
MULT
,r p
1/q MULT DIV sr XPO XPO 4--
DIV G
1.23 0.8 xD se
,e XPO XPO
-0.68 0.068 s,
MULT x.0193
,e MULT MULT Note:
See table 2-5 for Parameter definitions.
k MULT h
x 1/D
PAGE 46 TABLE 2-6 HOT SPOT AVERAGE TUEL TEMPERATURE AND ENTHALPY I.
Average Tual Temperature Assume the fuel pellet is modeled in the radial dimension by n equally spaced concentric rings bounded by n+1 mesh points (nodes).
Let the tcaparature at node i be given by Ti, where i 1,2,....n+1.
(Therefore,
=
T1 is the fuel centerline temperature and Tn+1 is the pellet surface tamparature.)
The area weighted average fuel temperature T is given by
))(T4)(2i-2)
+
(1/2)(Tn+1)(2n - 1)]
T=
(1/n ) [(1/2)(TI)
+
i=2 For the RETRAM Hot Spot Model, n=
10.
Therefore, 10](((Ti)(2i-2)
+
(19/2)(T1131 T=
(1/100) ((1/2)(T1)
+
i=2 II.
Average Tuel Enthalpy From a table of enthalpy versus fuel temperature, the value of the fuel anthalpy at node i, hi, is found for each corresponding value of Ti.
The same area weighting as given above for T is then used to derive the overage fuel enthalpy h.
That is,
[(hi)(21-2) h=
(1/100) ((1/2)(h1)
+
+
(19/2)(h11)]
i=2
PAGE 47 2.2 Calculational Technique The analysis of the rod ejection transient by Vepco may be grouped into thrae major phases as follows:
1.
Generation of physics data from steady state physics calculational methods for use by the RETRAM models, 2.
An analysis of the reactor coolant system response to the transient using the RETRAM Single Loop Model with the point kinetics option to predict the core average power history, and 3.
A thermal-hydraulic analysis'of the core hot spot location using the RETRAM Hot Spot Model.
All three phases incorporate assumptions in their cnalytical techniques to provide a
conservative prediction of the results of a rod ejection transient.
e
PAGE 48 2.2.1 Steady State Physics Analysis In order to model the core physics conditions for a specific reload dosign with the RETRAM Single Loop Model, the following core physics parameters are required 1.
ejected rod worth, 2.
maximum total power peaking factor (Fq) before and after the rod is ejected, 3.
Doppler power defect, 4.
delayed neutron fraction, 5.
moderator temperature coefficient, 6.
total trip worth less the most reactEde stuck rod, and 7.
prompt neutron generation time.
The first four parameters are the most critical in determining the coverity of the transient. The transient is minimally sensitive to the romaining three parameters as will be shown in Section 3.
The maximum total power peaking factor is not used directly in the point kinetics analysis performed with the RETRAN Single Loop Model, but is important in formulating a
reasonable description of the Doppler roactivity feedback to be expected during the transient (see Section 2.2.3 below),
and is of critical importance in predicting the thermal-hydraulic response of the hot spot.
The use of the term
" rod" refers to a single rod cluster control assembly (RCCA);
that is, a
RCCA consists of the entire cluster of individual rodlets which move as a single unit within a single assembly.
I PAGE 49 For Westinghouse plants, a rod insertion limit for the various control red banks is defined as a function of core power level. This limit is not such that operation of the plant with the control banks inserted above their respective limits insures that the core maintains an cdequate shutdown margin and acceptable power distribution. A search is porformed with the physics codes to determine the location of the highest worth ejected rod with the banks at their appropriate insertion limits for the desired core conditions. For example at hot zero power (HZp) core conditions only the D and C banks will typically be inserted inte the core while at hot full power (HFp) only the D bank will be inserted.
(See Refs. 1 and 2 for a description of the control rod bank nemenclature and geometry for Vapco's nuclear units.) Rod worths are ec1culated with frozen feedback as described below.
The parameters mentioned above are calculated for each reload core using Vopco's steady state physics codes as described in Refs. 11, 12 and 13.
Due to the lou sensitivity of the transient to the prompt neutron lifetime (see 3.2.1.4),
a generic value is used in the RETRAM point kinetics analysis although this value is checked with the calculated roload value to assure that no large deviation has occurred. Generic values of the minimum total trip reactivity are also used in the RETRAM point kinetics
- model, but these values are always compared to the minimum availabla reload calculated values of the total trip reactivity to ensure that the generic values are conservative.
The remaining parameters when used by the RETRAM models are modified by their appropriate nuclear reliability factors in a direction which is
PAGE 50 conservative for the rod ejection transient. For example, the ejected rod worth is increased by 10%, which leads to a higher core reactivity insertion and thus produces a more severe transient than would otherwise bo expected.
These nuclear reliability factors are documented in Ref.
14.
Additional conservatism is insured by calculating all physics parameters at steady state conditions using the
" adiabatic assumption."
The ediabatic assumption asserts that any fuel damage which might occur during the transient takes place in a small time interval immediately following the ejection of the rod and before the thermal-hydraulic isedback effects of the core become important. This freezing of the core's feedback leads to larger values of the total power peaking factor and ejected rod worth than would otherwise be expected in the transient.
An analytical assessment of this effect is presented in Section 4.4 of this report where a comparison is presented between steady state peaking factors calculated without thermal-hydraulic feedback and transient paaking factors calculated with thermal-hydraulic feedback using a three-dimensional kinetics code.
For each core reload analysis, key steady' state physics parameters are used in determining the severity of each transient, values of these parameters derived for the reload core are compared to thos9 "ced in a roference safety analysis. When any reload parameter is not b..pr.ded by the value used in the reference analysis, the affected transient (s) must bo reevaluated based on the new key parameter values.
Additional details on the use of the Vepco physics design codes in
PAGE 51 f
enalyzing the safety of a reload cycle and in their application to the red ejection transient in particular are provided in Ref. 15.
4 e
PAGE 52 2.2.2 core Average Transient Analysis The core average transient analysis is performed using the RETRAN Single Loop Model with the point kinetics physics option.
Core power is distributed among three equal si=e control volumes stacked axially one upon the other.
Each contains a single heat conductor, core power is distributed axially among the core volumes with an approximate cosine chapes that is. 25% of the power and reactivity feedback effects occur in the upper and lower core volumes while the remaining 50% occurs in the middle core volume.
The fuel rod is modeled with tuo radial concentric' fuel regions, a single gap region, and a single clad region.
Initial reactivity insertion due to the ejection of the control rod is icplemented by modeling a linearly increasing reactivity insertion into the core from zero to the total worth of the ejected rod over a time interval of 0.1
- seconds, (Ref.
17.)
The value used for the total integral worth of the ejected rod is the value provided by the physics dosign codes increased by the Nuclear Reliability Factor of 10% for conservatism, f
Nagative reactivity insertion from reactor trip is initiated by the high Power trips of the RETRAN Single Loop Model. For HFp cases the trip cotpoint is 118%
of full power (including setpoint and instrument crrors) while for HZP cases the trip setpoint is 35% of full power (cgain including setpoint and instrument errors). A 0.5 second trip dolay is assumed between the time the trip setpoint is reached and the time trip reactivity insertion begins, (Refs. 1 and 2). A conservative octimate of the total trip reactivity integral rod worth is provided for
l PAGE 53 the appropriate core condition.
This is typically 4000 pcm for a HTp condition and 2000 pcm for a
HZp condition. (A "pcm" is equal to a P3rcent mille of reactivity.)
This total integral trip worth is cultiplied by a conservative curve of normalized trip worth versus time to provide an estimate of the actual negative reactivity inserted at a porticular time following trip, (s'ee Figure 2-5).
Roactivity feedbach effects due to a change in the coolant density are codeled by a
moderator temperature coefficient. Reactivity feedback offects due to a change in the fuel temperature are modeled by a table of Doppler defect versus fuel temperature.
Both the moderator tamperature coeffidient and Doppler defect values are provided from the physics design codes and are modified by the appropriate Nuclear R3 liability Factors in a conservative direction, (i.e.,
+3 pcm/'F for the moderator temperature coefficient and less 10% for the Doppler defect.) In addition the Doppler defect values are multiplied by a Power Waighting Factor (pWF) to more accurately model the additional feedback offect obtained from actual three-dimensional geometry versus the point kinetics geometry of the RETRAM code. A description of the derivation end use of the PWF is provided in Section 2.2.3 below.
All reactivity parameters input to the RETRAM code are in units of dollars (s).
- Hence, tbc values obtained from the physics design codes cre converted to dollars by dividing by the value of the delayed neutron fraction. The delayed neutron fraction is provided by the physics design codes for the appropriate core conditions and modified by a Nuclear R0 liability Factor in a conservative direction.
PAGE 54 The value of the gap heat transfer coefficient (HGAP) used in the fuel pin model is held constant throughout the point kinetics calculation.
Since the correlation between core average fuel temperature and power lovel (necessary to construct a
table of Doppler defect values as a function of the fuel temperature) is dependent on HGAP, the problem crises of an appropriate value of HGAP to use. Steady state calculations unge performed with the RETRAM Single Loop Model at various core power lovels while varying the value of HGAP.
Correlations were derived botween power level and core average fuel temperatures as a function of HGAP. Heat transfer from the fuel increases or decreases with increasing or decreasing values of HGAP, respectively. Lower values of HGAP result in a
higher relative core average fuel temperature for a given power lovel. Since higher fuel temperatures result in greater negative Doppler roactivity feedback, a high value of HGAP appears to be conservative for the rod ejection transient. This effect is examined in Section 3 where the RETRAM point kinetics model is shown to be only moderately sensitive to the value of HGAP used. A value of 1000 Btu /ft2-hr *F is used for
{
HGAP in the Point kinstics calculation.
In addition to a prediction of the core average power history for the transient, the RETRAM point kinetics model also calculates the normalized core energy release.
This parameter is obtained by integrating the normalized power as a
function of time using an integrator control block.
The transient is analyzed at four different core conditions for a given roload cycle. These conditions are
l PAGE 55 l
1.
Beginning-of-life (BOL), hot zero power (HZP) 2.
Beginning-of-life (BOL), hot full Power IHTP) 3.
End-of-life CEOL), hot zero power (HZP) 4.
End-of-life (EOL), hot full power (HFP)
These conditions are the current licensing basis for the Surry and North Anna nuclear units, (Refs. 1 and 2.)
6
FIGURE 2-5 exac ss ORMALJ2ED TayP REACTJVJTY CU i
1. 0 --
0.9k N
0.6 l
I N
O R
0.7 N
A L
0.6d Z
f E
O
)
T R
0. 5.:
J P
R E
0.4 :
A C
T I
V 0 32 1
T Y
0. 2 --
01 00 e
A I
0.0 0.4 0.8 1.2 1.6 2.0 2.4 TJnE AFTER TRJP iSEC)
l PAGE 57 2.2.3 power Weighting Factor The RETRAM point kinetics model calculates the Doppler reactivity foedback during the transient from a
table of negative reactivity insertion versus core average fuel temperature.
This table is ocnstructed from the results of Doppler reactivity calculations parformed with a
steady state diffusion code such as PDS07. The table provides a reasonable set of values for the core during a pre-transient core conditions that is, with the ejected rod still inserted into the core.
Upon the ejection of the rod, an extreme skeuness results in the radial flux and temperature shapes about the position of the ejected rod. This condition violates the basic assumption behind the applicability of the point kinetics approach--i.e.,
that the spatial flux shape throughout the core does not vary appreciably with time. Use of the typical point kinetics models produces results for this transient that can be very conservative. Accurate modeling of the core's actual reactivity feedback for such a case requires that the spatial kinetics effects, especially in the radial direction be included.
However, the spatial kinetics GPProach has the disadvantage of requiring considerably greater computer rosources than the point kinetics appkoach.
An acceptable solution to simulating the radial kinetics effects with the point kinetics model may be found in the application of a pouer W2ighting Factor (pWF) to the Doppler reactivity feedback table used by the point kinetics model. To understand the concept behind the use of a
- PWF, consider the case where one actually knows the Doppler temperature i
s PAGE 58 i
i coefficient for the post-ejected rod core condition. Let this value be denoted as DTC_0UT. This is the value of the Doppler reactivity feedback which should actually be reflected in the point kinetics Doppler roactivity table. Likewise, assume the value of the Doppler temperature ocefficient for the core with the ejected rod still inserted, (i.e.,
the pre-transient value), is known. Let this value be denoted as DTC_IN.
In
- general, a
Doppler reactivity coeffizient is the change in core roactivity, DELRHO, resulting from some unit change da the core average fuel temperature, DELTF. That is, DTC = DELRH0/DELTF For the case of the ejected rod being fully withdrawn from the core, this equation may be written as f
DTC_0UT = DELRH0_0UT/DELTF_0UT Likewise, for the pre-transient case, we have DTC_IN = DELRHO_IN/DELTF_IM since the unit changes in fuel temperature for the two cases are normally
- equal, (i.e.,
DELTF_0UT equals DELTF_IN), the ratio of the Doppler temperature coefficients for the two different rodded configuration may be written as:
DTC_QUT/DTC_IN = DELRH0_0UT/DELRH0_IN The power Weighting Factor is defined as:
l l
PAGE 59 pWF = DELRH0_00T/DELRHO_IM
~
uhere the assumption is made that the change in core average fuel temperatures over which the resulting changes in reactivities are calculated is the same for both cases.
Given a method for calculating DELRH0_OUT and DELRH0_IM, and thereby the PWF, for various rod ejection eve..t conditions, it would be advantageous to find a
correlation between the PWF and some other parameter of the rod ejection transient. If such a correlation exists, the value of the PWF for a
particular rod ejection case may be determined based on the corresponding value of the parameter to which the PWF is correlated.
Since the value of DELRH0_IM, or more correctly, the Doppler reactivity table for the pre-transient condition, is readily available from steady state diffusior.
code calculations, the value of the PWF may be applied to the table to more correctly predict the true Doppler reactivity foedback to be used by point kinetics in modeling the transient. This approach has the advantage that DELRH0_0UT need not be explicitly cciculated for each rod ejection analysis.
Two-group neutron perturbation theory was used to calculate the value of the PWF for various post-ejected rod core conditions for both Surry and North Anna core loadings. Changes in the core average reactivity due to a
perturbation of the fuel temperature were calculated for the various radial fuel temperature distributions.
Values of nuclear macroscopic cross sections and normal and adjoint flux shapes required for the calculations were provided by radial full core, coarse mesh calculations
l l
PAGE 60 parformed by the PD207 One-Zone Model, (Ref. 12.) Details on the use of noutron perturbation theory and its application to calculating point kinetics parameters are provided in Ref. 16.
The resulting values of pWFs were then correlated with the maximum values of the post-ejection radial power peaking factors (Fxy) as calculated with the two-dimensional PD207 Discrete Model, (Ref. 11.) The rosulting least squares fit to the points is presented in Figure 2-6.
The value of the weighting factor increases with increasing Fxy due to the stronger feedback effect.
The value of Fxy is in turn strongly correlated with the worth of the ejected rod. This curve is used to do' rive the value of th3 pWF to be used for a specific rod ejection enalysis.
The Doppler defect curve used in the point kinetics RETRAN codel is then weighted by this value of the PWF to provide a more rossonable estimate of the Doppler reactivity feedback effect during the transient.
Since the value of the post-ejection Fq is routinely provided for the transient, (suitably increased by a Nuclear Reliability Factor,NRF), the value of the post-ejection Fxy used to derive the appropriate pWF to be used is calculated bys Fq/(NRF x 1.55)
Fxy
=
where NRF removes the additional conservatism of the Nuclear Reliability Factor (1.075 for the Vepco models) and the factor of 1.55 is a generic value of the axial peaking factor for a typical cosine axial power distribucion shape.
Some additional points of clarification in the use of the pWF follow.
PAGE 61 I
L Mo benefit is taken for negative Doppler reactivity feedback effects which result from the change in the axial flux shape during the ejection of the rod. Hence, the application of the PWF is conservative from this standpoint.
For a
HYP case, the worth of the ejected rod is relatively low. Hence, the value of the post-ejection radial peak power is also relatively low cnd typically yields a
value of the PWF near unity.
As will be demonstrated in the sensitvity studies of Section 3, such a low value of PWF has little effect on the transient results. Therefore, the PWF is of I
loss importance for a HFP case.
For a
HZP case, the importance of the PWF is increased. The relatively high ejected rod worth usually encountered results in a large value of Fxy and hence a
PWF value which has a more significant impact on limiting the initial power excursion following the rod ejection.
The application of a PWF derived from perturbation theory to the point kinetics model is an improvement in the prediction of what the actual Doppler reactivity feedback would be for the transient; however, best ostimate analysis of the Doppler reactivity feedback is not implied. To demonstrate that the use of the PWF produces a result that is still conservative relative to a more accurate prediction of the transient, comparisons with rozults from a three-dimensional space-time kinetics codel are provided in Section 4.4.2.
These comparisons will not only verify the conservatism of the use of the PWF, but of the point kinetics codel in general for the rod ejection analysis of Vepco nuclear units.
FIGURE 2-6 Exces2
(
POWER WEIGHJTNG FR.CTOR l
4.00-x 3.75-3. 5 0 --
3.25-P O
x W
3.00-E R
W 2.75-E I
G H
2.50-T x
I N
G 2.25-F R
x x C
2.00-T 0
R 1. 7 5 --
x 1.50-g x
1.25-x 1.00-0 1
2 3
4 5
6 7
8 9
10 11 12 RRD18L PERKING FRCTOR e
e u
i A
PAGE 63 2.2.4 Hot Spot Transient Analysis Ao described in Section 2.1.2, in order to calculate the thermal-hydraulic response of the hot spot core location to the ejection of the rod, the power history of the het spot is required. Ideally this would be modeled by multiplying the total power peaking factor (Fq) history of the hot spot by the core average power history calculated l
uith the point kinetics RETRAN model.
In place of the hot spot Fq history, the pre-and post-ejection values of Fq as calculated by the physics design codes are available. Since these values have been calculated using the " adiabatic assumption", they will be higher (i.e.,
conservative) compared to the actual values of Fq oxpected to occur throughout the initial part of the transient. A conservative hot spot Fq history is constructed by assuming that initially the value of Fq is the steady state pre-ejection value f
provided by the physics codes. This value is then linearly increased to the post-ejection value of Fq over a time interval of 0.1 seconds and hold there for the remainder of the transient. (This is the same time interval assumed for the complete ejection of the rod from the core.)
This Fq power history is then multiplied by the core average power history to provide the hot channel power history for the Hot Spot model.
Upon reactor trip, the insertion of the scram banks into the core will ccuse additional perturbations to the core power distributions both radially and axially. It is possible for the value of Fq at a later time in the transient to actually exceed the post-ejection value of Fq uhich 10 assumed to occur at 0.1 seconds into the transient. However, because
PAGE 64 i
of the relatively low value of the core average power at such a time, the resulting hot channel power will be appreciably lower than during the first part of the transient when the core average power is at a f
ecximum.
In addition, the location of the hot channel in the core will change during the transient whereas the hot spot analysis assumes a single location throughout the transient.
An actual plot of the hot channel power during the transient as predicted by a three-dimensional space-time kineticF model is Provided in Section 4.4.2 and shows the power in the hot channel to be steadily decreasing following the early paak due to the ejection of the rod.
In
- summary, the power history driving function input to the RETRAN Hot Spot Model consists of tuo components: a conservatively predicted core average power history and a
conservative total peaking factor power history.
f At 0.1 seconds into the transient, the hot channel is forced into DNB by switching from a
subcooled nucleate boiling surface heat transfer coefficient, (Thom correlation, Ref 4.
Vol.
1),
to a film boiling surface heat transfer coefficient (Bishop-Sandberg-Tong correlation, Raf. 10.) Conservatisms applied to the latter correlation are the use of
{
a safety factor to reduce the calculated heat flux and the assumption of a constant bulk coolant density.
The large increase in core power level is expected to lead to a rapid oxpansion of the fuel pellet. This reduces the pellet-clad gap size, resulting in an increase in the gap heat transfer coefficent. At 0.1 soconds into the transient, the value of HGAp in the Hot Spot Model is 1
PAGE 65 increased from its initial value to a relatively high value of 10000 Btu /ftz-hr *r to reflect this closure of the gap.
From the hot channel calculation., values of the fuel temperatures at the I
boundaries of each of the 10 fuel pellet regions, the clad temperatures
(
nt the inner and outer clad surfaces and the pellet average enthalpy are obtained as a function of time. These parameters are used to assess the f
fuel response against the acceptance criteria presented in Section 1.3.
Typically, if the maximum amount of fuel melt at the hot spot is less than
- 10X, the pellet would be expected to maintain its configuration end, therefore, no unacceptable radiological consequences or core damage are expected to occur. An upper bound on the percentage of fuel melt is dorived from the ratio of the cross sectional area of the portion of the pollet where melting occurs to the total pellet cross sectional area.
The cross sectional area faz melting is determined by noting those l
radial concentric fuel nodes whose temperature at some point in the
- transient exceeds the assumed temperature for fuel melt. At BOL case this melting temperature is assumed to be 4900 'r while for an EOL case o
temperature of 4800
'T is assumed. Table 2-7 presents the table used to compute the maximum fraction of fuel melt for the RETRAN Hot Spot Model.
(
PAGE 66 TABLE 2-7 HOT CHAMMEL FUEL MELT FRACTION TABLE
(
Highest Numbered Node n Maximum melt for which Tn < Taelt fraction (X) l 1
0 2
1 3
4
(
4 9
5 16 Notes
- l Tn = fuel temperature of node n Taelt = fuel melting temperature
= 4900 'r for BOL case 4800
'F for EOL case
=
l RETRAN Hot Spot Model contains 11 fuel nodes, (10 concentric l
fuel rings.)
(
l
(
l
r l
PAGE 67 2.2.5 System Overpressure Analysis Included in the acceptanco criteria is the limitation that the maximum roactor pressure during the transient will be less than the value that will cause ths stress to exceed conservatively defined stress limits.
This problem is resolved by a pressure stress calculation and has been cddressed generically for the North Anna and Surry nuclear units as f
d0 scribed in Refs.
1, 2
and 17.
Since it was concluded that no violations will occur for Vepco nuclear units due to the rod ejection transient, the system overpressure analysis is not performed as part of the Vepco methodology.
I l
k 9
=
e_
--__m
PAGE 68 r
[
2.2.6 Radiological Concerns AD with the case of the system overpressure analysis, the radiological concerns for the rod ejection transient for the Surry and North Anna nuclear units have been addressed generically, (Refs.
1, 2 and 17.) In cosessing the fission product release, it is assumed that all of the rods which experience DNB release their entire gap inventory of fission f
products to the coolant. Vepco's additional acceptance criteria that the tamparature for clad embrittlement will not be exceeded and that the fuel pellet configuration will be maintained guarantees that the condition of the fuel rod losing its integrity from entering DNB is a very conservative assumption.
The nature of the rod ejection event causes a large hot channel factor to occur enl>
in a
very localized region of the core, i.e.,
in the inmediate vicinity of the ejected rod. Fuel census analysis as provided
(
in Raf. 17 showed that for the Worst case investigated, the average fuel rod failed to reach DNB, and even for those fuel rods reaching DNB no oxcessive release of fission products was to be expected, f
In
- summary, the localized nature of the event coupled with the small number of rods expected to reach DNB and the large conservatisms inherent in the analysis ensures that meeting the Vepco imposed limits of allowable fuel melt and clad embrittlement temperature for the hot channel assures that the radiological limits for the event as specified in Regulatory Guide 1,77 (Ref. 3) u111 he met.
I L
PAGE 69
['
SECTION 3 - SENSITIVITY STUDIES l
3.1 Introduction f
A sensitivity study was performed for both the RETRAM Single Loop Model cnd RETRAM Hot Spot Model to quantify the impact of uncertainties in i
core parameters and modeling assumptions on the models' predictions for the rod ejection transient.
This section provides a summary of the ctudy's results.
The study is divided in three parts:
(1) neutronics parameter sonsitivities for the point kinetics calculation, (2) thermal hydraulic censitivities for the point kinetics calculation, and (3) thermal hydraulic sensitivities for the hot spot calculation. The first two Parts were performed with the RETRAM Single Loop Model for a Surry plant.
Tor these two
- parts, the four bounding cycle conditions were cnalyzed for each sensitivity, (i.e.,
BOL HZP, BOL HFP, EOL HZP and EOL HFP.)
The sensitivities for the hot spot calculation were analy=ed with the RETRAM Hot Spot Model for typical HZP and HFP hot spot power histories only at BOL, since for this calculation, with the exception of the assumed temperature for fuel
- melt, the cycle lifetime is not explicitly reflected in the input.
{
l l
l PAGE 70 i
L 3.2 Sensitivity Study Results 3.2.1 Point Kinetics Neutronics parameters f
Sonsitivity studies were performed for the point kinetics calcula' ion to casess the impact of various neutronics parameters on the predicted core everage power history.
For the rod ejection event, the core average Power history may be divided for sensitivity analysis purposes into two cojor time domains. The first of these domains runs from the beginning of the transient until approximately 0.5 seconds into the transient.
During this time the core average power rises rapidly, obtains its noximum value, and is turned around by the negative reactivity insertion due to Doppler feedback.
The transient results are particula' ly consitive to the core average power histories in this first domain, as a significant increase in the power value will lead to higher transient hot spot temperatures and thus produce more severe transient results.
The second
- domain, starting at approximately 0.5 seconds into the transient, models the' reactor shutdoun with the insertion of the control rod banks and the effects of moderator reactivity feedback. During this time interval, the core average power level is significantly lower than during the first domain.
Tchle 3-1 presents a summary of the results of the sensitivity study for the neutronics parameters for the four different cycle conditions. The P3ak normalized core power level (where unity is equal to full power) is presented for each case along with the total energy released during the first five seconds of the transient.
This latter quantity is of
(
PAGE 71 I
icportance for HZp cases in demonstrating the sensitivity of a
perameter.
This is due to the narrowness of the initial core average pswer peak which occure during the first time domain for HZp cases. What f
10 in reality a
small impact on the transient results may show a difference in the peak normali=ed power between the nominal and consitivity cases of several full power levels. Hence, the total energy release (which is the integral of the core power history) at some later point in the transient is used as a more representative quantity for demonstrating the magnitude of the sensitivity for HZp cases. The total energy release is also useful in assessing the sensitivity of a parameter whose effects are not noticable until the second time domain, (o.g.,
trip worth.)
The first entry in the table for each cycle condition is for the nominal ccsa with no perturbation to any of the neutrenics parameters.
Comparison of the sensitivity values for the peak normali=ed power and onergy release for each case with those of the nominal case provides a concise summary of the relative sensitivity of each case.
(
3.2.1.1 Doppler Reactivity Feedback The first sensitivity investigated is that of the Doppler reactivity foedback.
This is modeled in the RETRAM point kinetics calculation through the input of a table of Doppler defect as a function of fuel tamparature. Since the Doppler defect values in the table are multiplied by an appropriate Power Weighting Factor (pWF), comparisons are shown batween the pWF used for the nominal case and a PWF of unity.
s f
L PAGE 72 For the two HZP
- cases, the necessity of the use of the PWF becomes obvious. Without the PNF the total energy release is tripled for the BOL ocse and nearly doubled for the EOL case. Figure 3-1 presents plots of core average power histories for the tuo HZP cases with and without the
(
PWF applied.
Corresponding plots of the total energy releases are presented in Figure 3-2.
For comparison, additional curves are plotted with the Doppler reactivity feedback curve used in the nominal cases docreased by 10%
(typical of the uncertainty assumed in a Doppler foedback calculation by steady state physics models.)
The 10%
uncertainty in the Doppler defect shows little sensitivity compared to
- ^
that resulting from deletion of the PWF.
For the HFP cases as summarized in Table 3-1, reducing the PWF to unity, although it increases the peak normalized power and total energy
- release, shows a much lower sensitivity than the HZP cases. This is as would be
- expected, since for a HFP case the ejected rod worth, and, therefore, the resulting power peaking and flux redistribution, are analler than for a HZP case. Hence, PWF values which are near unity have little impact on the transient predictions. Figure 3-3 presents the normalized power history comparisons for the two HFP cases.
3.2.1.2 Moderator Reactivity Feedback Maderator reactivity feedback effects in the RETRAM point kinetics eniculation are modeled by inputing a moderator temperature coefficient (MTC).
The assumed uncertainty in the value of the MTC calculated with Vopco steady state physics codes is presently +3 pcm/'F. The value for
f L
PAGE 73 the MTC input to the nominal cases was increased by this uncertainty for the sensitivity study.
The result was a minor increase in the core everage power prediction as would be expected for the insertion of an cdditional positive reactivity feedback effect. Figure 3-4 presents a j
comparison of the normalized power histories for the two BOL cases. The offect of the MTC is shown to be most prominent early on in the transient where the greatest change in core power level, and therefore, the greatest change in moderator temperature, occurs. A lag occurs botween the time of peak power and the onset of the effect of the coderator feedback since the moderator temperature responds more slowly to changes in core power than the fuel temperature.
3.2.1.3 Delayed Neutron Fraction The delayed neutron fraction is used in estimating the dollar worth of the reactivity effects in the solution of'the point kinetics equation.
The value of the delayed neutron fraction decreases with core burnup.
Docreasing the delayed neutron fraction value by 5%, (the assumed design uncertainty),
represents a
reduction in the percentage of delayed noutrons in the core and thereby implies a core with faster response to
/
changing conditions.
In addition, the dollar uorth of the tuo dominant s
roactivity effects for the transient, the ejected rod worth and the Doppler reactivity feedback, are both modified by perturbing the value of the delayed neutron fraction. In effect, reducing the value of the dolayed neutron fraction by 5%
increases the dollar uorth of both parameters by 5%. As described in Section 3.2.1.1 above, the transient
i PAGE 74 r
chous little sensitivity to an increase in the Doppler reactivity
[
foedback of the magnitude of 5%.
However, as documented in Section
(
3.2.1.5
- below, a
5% increase in the ejected rod worth is significant, f
The ejected rod worth is therefore the stronger of the two competing roactivity effects.
This results in a higher predicted core average pcuer with decreasing delayed neutron fraction. As is demonstrated in a comparison of BOL and EOL cases with similar core parameters, the EOL ocses with significantly lower delayed neutron fractions show higher predicted core average powers.
The same effect appears in the consitivity results which show slightly higher predicted powers than the nominal
- cases, (Table 3-1).
Figure 3-5 presents a comparison of the power histories for the tuo BOL cases.
3.2.1.4 prompt Neutron Generation Time An increase in the mean value of the prompt neutron lifetime is expected to slow the rate of initial increase in core average power during the transient
- since, on the average, each prompt neutron will now survive longer in '@he core before it is absorbed.
As shown in Table 3-1, f
increasing the value of this parameter by SE-6 seconds (from a nominal value of 18E-6 seconds) typically decreases the predicted peak normali=ed power although the total energy release shows little censitivity.
Likewise, a decrease in the mean value of the prompt neutron lifetime is oxpected to hasten the rate of initial increase in the core average Power during the transient, with each neutron being ab= orbed sooner. As l
Y L
PAGE 75 r
choun in Table 3-1, decreasing the prompt neutron lifetime by SE-6 (from a
nominal value of 18E-6 seconds) increases the predicted peak normalized power phile the total energy release again shows little consitivity. The effects of these changes is more pronounced for the HZP cases than the HFP cases.
Since the assumed design uncertainty in this parameter is only 5%, the
(
transient can be expected to show little sensitivity to even large voriations in the value used by the point kinetics model. Therefore,'a generic value of 18E-6 seconds is used for all analyses.
3.2.1.5 Ejected Rod Worth The assumed design uncertainty for rod worth is 10%. Increasing the ojected rod u' orth by 10% shows the greatest impact on the power history of all the swasitivities investigated.
As
- expected, the increased f
roactivity insertion due to the withdrawal of a greater rod worth from
~
the core results in a higher peak normalized power. Figure 3-6 presents the power history comparisons for the BOL cases.
3.2.1.6 Rod Ejection Time In the nominal cases, the rod is ejected by modeling a linear reactivity insertion into the core over a
time interval of 0.1 seconds. Two consitivity cases were investigated for both a shorter rod ejection time (0.05 seconds) and a
longer rod ejection time (0.2 seconds.) Little impact resulted from either sensitivity. Typically, for the shorter rod ojection
- time, the peak core average power occurred earlier in the
..a
._..w
r PAGE 76 L
transient (see Figure 3-7 for a BOL comparison), while Io'r the longer
{
red ejection time the peak power occurred correspondingly later in the transient.
(Figure 3-8). The lower peak powers resulting for the longer
(~
transient time for the HTp cases is a result of the negative feedback j
offects having more time to mitigate the effects of the positive roactivity insertion from the ejection of the rod. No significant impact en the integrated anargy release was observed.
3.2.1.7 Trip Delay Time A
trip delay time of 0.5 seconds is assumed in the nominal cases.
Sonsitivity studies were performed by increasing this value to 1.5 coconds.
Since the peak normalized power for the rod ejection transient typically occurs before the scram banks begin movement, increasing the trip delay time has no impact on the peak normali=ed power as shown in Table 3-1.
However, the delay in negative reactivity insertion slows the rate of decrease in the core pouer level leading to a greater energy release throughout most of the period through which the scram banks are inserting.
Figure 3-9 presents the BOL power history comparisons for f
this sensitivity.
3.2.1.8 Trip Worth Ao with the ejected zod worth, the assumed design uncertainty of 10% uas cyplied to the trip worth for the sensitivity study. In this case, the ubrth of the trip banks were decreased by 10% in order to cause an increase in the core power level.
As with the
" trip delay time"
PAGE 77 i
'scusitivity
- study, the power history predictions of the sensitivity ccses were identical to those of the nominal cases during tne first time domain.
As shown in Table 3-1, the sensitivity of changing the trip f
worth is minor. Figure 3-10 presents comparisons of the power histories for the BOL cases.
3.2.1.0 Initial Zero power Level For the nominal zero power cases, an initial normalized core power level of 1.0E-9 was assumed.
This power level was increased to a value of 1.0E-3 Ci.e.,
0.1% full power) for the sensitivity study. The transient showed little sensitliity to this change as shown in Table 3-1.
Figure 3-11 presents the power history comparisons for the two HZP cases. For the sensitivity case, the peak normali=ed power is reached earlier than for the nominal case. This would be as expected since the sensitivity case is initially starting at a higner power level.
3.2.1.10 Time Step Size The time step si=es used in obtaining the time-dependent numerical solutions to the equations in the RETRAM model are chosen according to the expected rate of change of core conditions during the transient.
- Thus, early in the transient where the most rapid changes in core power and temperatures are taking
- place, the time step si=e is chosen rolatively small compared to later in the transient where the rate of j
change of core pczameters is slower. Ideally, the smaller the time step size, the greater the confidence in'the accuracy of the solution.
PAGE 78 The sensitivity of the time step si=e was investigated by decreasing it by a factor of approximately 0.2 over the nominal value used throughout the transient.
The results as presented in Table 3-1 show little sensitivity to this magnitude of ret.uction.
3.2.1.11 Beta Yield Fractions The RETRAM point kinetics model has a built-in standard set of beta yield fractions for the reactor kinetics calculations. The sensitivity of this parameter was checked by inputting a set of beta yield fractions representative of Surry reloads in the RETRAN model. The results, as shown in Table 3-1, shou little sensitivity. Figure 3-12 presents a comparison of the power histories for the BOL cases.
PAGE 79 TABLE 3-1 POINT KINETICS HEUTRONICS SENSITIVITY STUDY I.
BOL HZP Studies:
Peak Energy Norm.
Release Parameter Sensitivity Power (t=5 sec)
Nominal case values Not applicable 51.1 2.23 Power weighting factor (PWF)
Changed from 2.4 to 1.0 127.
6.69 Doppler reactivity feedback Decreased 10% (PWF=2.4) 56.5 2.52 Moderator reactivity feedback Increased +3 pcm/'T 51.6 2.34 Delayed neutron frs.ction D.ecreased 5%
58.8 2.30 Prompt neutron generation time Increased 5.0E-6 sec 41.5 2.23 Prompt neutron generation time Decreased 5.0E-6 sec 73.5 2.24 Ejected rod worth Increased 10%
81.8 2.61 Rod ejection time Changed to 0.05 sec 52.5 2.24 Rod ejection time Changed to 0.2 sec 53.2 2.23 Trip delay time Increased to 1.5 sec 51.1 2.59 Trip worth Decreased 10%
51.1 2.26 Initial =ero power level Changed to 1.0E-3 52.2 2.24 Time step si=e Multiplied by 0.2 49.9 2.22 Beta yield fractions Surry Reload values 51.3 2.28 Notes:
Peak Horm. Power = peak normali=ed power occurring during transient, (1.0 is equivalent to full power level.)
Energy Release (t=5 sec) total core energy release up to a
=
transient time of 5.0 seconds, (units of full-power-seconds.)
I h
PAGE 80 s
1 TABLE 3-1 (cont.)
L POINT KINETICS NEUTRONICS SENSITIVITY STUDY
(
II.
BOL HFP Studies:
Peak Energy Norm.
Release Parameter Sensitivity Power (t 5 sec)
___=______
Mcminal case values Not applicable 2.02 3.50 Pcuer weighting factor Changed from 1.2 to 1.0 2.04 3.63 Moderator reactivity feedback Increased +3 pcm/'T 2.02 3.58 Dolayed neutron fraction Decreased 5%
2.13 3.52 i
Prompt neutron generation time Increased 5.0E-6 sec 2.02 3.50 Prompt neutron generation time Decreased 5.0E-6 see 2.02 3.50 Ejected rod worth Increased 10%
2.25 3.67 Rod ejection time Changed to 0.05 sec 2.04 3.49 Rod ejection time Changed to 0.2 see 1.94 3.52 Trip delay time Increased to 1.5 sec 2.02 4.80 Trip worth Decreased 10%
2.02 3.57 Time step si=a Multiplied by 0.2 2.02 3.49 Bota yield fractions Surry Reload Values 2.03 3.S1 Notes:
Peak Norm. Pouez a peak normalized power occurring during transient, (1.0 is equivalent to full power level.)
Energy Release (t=5 sec) total core energy release up to a
=
transient time of 5.0 seconds, (units of full power-seconds.)
PAGE 81 t
TABLE 3-1 (cont.)
POINT KIMETICS MEUTRONICS SENSITIVITY STUDY t
I III.
EOL HZP Studies:
Peak Energy Norm.
Release Parameter Sensitivity Power (t=5 sec)
Maminal cass values Not applicable 80.2 1.87 Pcuer weighting factor (PWF)
Changed from 2.3 to 1.0 139.
3.65 Doppler reactivity feedback Decreased 10% (PWF=2.3) 87.2 2.02 Moderator reactivity feedback Increased +3 Pcm/'F 81.2 1.91 D31ayed neutron fraction Decreased 5%
87.0 1.91 Prompt neutron generation time Increased 5 0E-6 sec 62.7 1.87 Prompt neutron generation time Decreased 5.0E-6 see 110.4 1.88 Ejected rod worth Increased 10%
116.
2.14 Rod ejection time Changed to 0.05 sec 78.4 1.87 Rod ejection time Changed to 0.2 sec 80.1 1.87 Trip delay time Increased to 1.5 sec 80.2 2.06 Trip worth Decreased 10%
80.2 1.89 Initial cero power level Changed to 1.0E-3 75.9 1.88 Time step si=e Multiplied by 4 2 77.9 1.86 Beta yiel'd fractions Surry Reload Values 80.3 1.89 Notes:
Peak Norm. Power = peak normali=ed power occurring during transient, (1.0 is equivalent to full power level.)
total core energy release up to a Energy Release (t=5 sec)
=
transient time of 5.0 seconds, (units of full-power-seconds.)
e PAGE 82 TABLE 3-1 (cont.)
POINT KIMETICS MEUTRONICS SENSITIVITY STUDY IV.
EOL HTP Studiest Peak Energy Norm.
Release Parameter Sensitivity Power (t=5 sec)
Nominal case values Not app'.1 cable 2.60 3.02 Power weighting factor Changed from 1.2 to 1.0 2.63 3,05 Moderator reactivity feedback Increased +3 pcm/*F 2.60 3.03 Dolayed neutron fraction Decreased 5%
2.79 2.99 Prompt neutron generation time Increased 5.0E-6 see 2.57 3.02 Prompt neutron generation time Decreased 5.0E-6 sec 2.61 3.01 E acted zod worth Increased 10%
3.04 3.10 j
Rod ejection time Changed to 0.05 sec 2.67 3.00 Rod ejection time Changed to 0.2 sec 2.43 3.03 Trip delay time Increased to 1.5 sec 2.60 4.06 Trip worth Decreased 10%
2.60 3.07 Time step si=e Multiplied by 0.2 2.59 3.00 Beta yield fractions Surry Reload Values 2.60 3.01 Notes:
Peak Norm. Power = peak normali=ed power occurring during transient, (1.0 is equivalent to full power level.)
Energy Release (t=5 sec) = total core energy release up to a transient time of 5.0 seconds, (units of full-power-seconds.)
FIGUPE 3-1 SENSITIVITY STUDY - ll2P DOPPLER REACTIVITY FEEDBACK (ENERfiY PELEASE)
END OF LIFE BEGINNING OF LIFE 1000.00-8000 00-l p
100.00-f 800.00-I l
N i
N f
A 10.00-i 1
0 A
10.00 R
L I
L I
i E
g O
E ll O
P O
l.00-i P
0 t.00-E E
E A
0 80-g 0 40-e 7
]
0.04-n 0.04-00 05 1.' O l.5 2.0 2.5 3.0 3.5 4.0 4.5 50 ilM 15EC) g ilME 15ECl S TRA : NOMINAL CASE IPMFat.35 SOUARE: SENSITIVlif CR5E IPurst 04 ETAA: NOMINAL CASE IPNFat.48 1AJANGLE: NOMINAL CA5E ILE55 1048 SOUARE: SEN5lilV111 EASE IPNFal.08 IRIANGLE& NOMINAL CASE ILE55 10XI 7,gu,c 3,i, fl0VRE 3-IA
Flfit'RE 3-2 l
SENSITIVITY STUDY - HZP D0PPLER REACTIVITY' FEEDBACK (ENEMY RELEASE)
END OF LIFE BEGINNING OF LIFE 7-3 5-
?
-e 3 3-8-
i 30 l
t-
l N
g l5-2 4-l l
L l4 I
2.i.
a O
g E
0 R
1.5-f3" i
A I
R a
E 3.2-f A
e E
E
!I" 0 9-g 0.5-I-
"U 0 3-O t rl 0-0.0-li g
0.0 05 1.0 l.5 20 2.5 30 3$5 40 4.5 50 00 05 3.2 l[8 24 30 355 42 40 IIME 15ECl fint 65EC)
STAR: NOMjhAL CASE (PNFa!.38 SouARE: SEM5tilVilf CASE IFuFal.08 ETAR NOMINAL CASE IPhFat.48 IRIANGLE: NonlNAL CASE (LESS BOXI SOUAAC : SEMEstaVlt? CASE (PNFal.Ol IRIANCtts NOMINAL CA&E (LE55 1048 FIOURE 3-28 F100RE 3-2R
FIGl'P.E 3-3 SENSITIVITY STt'DY - IIFP 00PPLFit PEACTIVITY FEEDBACK I?
BEGINNING OF LIFE END OF LIFE l
e j
2 75-ll 3 50-l.8-ll t.25-l l.5-t.00-1 4-m e 4 15-A N
ii M
0 A
A L
I.50-M l.2-A 3
L y
\\>
E 9
g w
l.25-E l.0-h O
P O
P u 8 00-'h 1
0 E
u O. e -~
A E
A 0 15-i O.Si 0 50-e
?
l 0. 4 -.
O.25-O 0.00,
g y
0 2J 0.0 0.5 1.0 9.5 20 25 3.0 3.5 4.0 4.5 50 m
tant 85ECl 0.0,
0.0 0.5 1.0 l.5 2.0 3.5 3.0 3.5 40 4.5 5.0 I
TIME 15ECl 1
STAR: NOMINAL CASE IPWFst.tl
)
50UAAEa SEN5818vliT CASE IFWFal.08 STAR: montNAL EASE IFWFal.21 50uRRE : SEN58IIVIIV CASE IFufsl.08 FIGUAE 3-30
FIGURE 3-4 I
SEllSITIVITY STl'DY - TDEPATnP REACTIVITY FEED 0ACK BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER
.00 00-r.0-ll ii I I l.0-l l.5-10.00-i 8.4-l N
N o
e R
R S l-L l'
L 8
8
..Os i
O i
0 P
P o
a M
M 0.5-'
E E
4 R
0.5-0 10-0..-
0.2-f3 0.08-
=
0:0 0.s i.0 i.s
- 0 r.s 3.0 s.
4:0
..s s.O nne inci tth!
!E 500AAt 1
3 PCn/ DES FI r Co.E 3...
,,0u.E 3...
FIGl?RE 3-5 i
SEliSITIVITY STUDY - DELAYED NEUTR0il FRACTIO 1 BEGINNING OF LIFE, HOT ZERO POWER BEGINN!NG OF LIFE HOT FULL POWER t.35-l 2.00-3 15-l 10 00-8.50-0 0
A I
A l
A l.25-L i
e t
8.00-g O
I.00-h l
0 i
r 0
0 1
l 0 15-A l
0.50 O.60-l 0 25-0.00-o l
N O.0 0.5 l.0 1.6 f.0 2.5 30 3.5 4.0 4.5 5.0 IlME 18fCI 0.01-OO 05 1.0 3.5 2.0 2.5 3.0 35 4.0 4.5 5.0 TIME 15ECB STAA: NOMINAL EA5E SOUARE: SEN5fIlVIIT EASE ILESS EXI flGunt 3-5A 50uRRE:
It AS LE55 EXl FIGURE 3-55 L__
g
w
~
FIGURE 3-6 SENSITIVITY STUDY - EJECTED R0D Ll0RTH BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER 100.00-2.50-ll s.r5-l l
.00--
10 00-i.25-N N
e O
A 1 50-A M
n A
A L
L i
i i
I i
I l.25-l 1.00-ii 0
P ii 0
1.00-'h a
E E
A A
i 0.15-0 80-0.50-i i
nl 0 25-o g
0.00-g 0 08-00 0.5 l.0 35 2.0 2.5 3.0 35 4.0 4.5 50 00 0:5 i.0 i:5 ::0
.5 3.0 3:s
.0
. :..:0 flat IEEcl ilME 15EC B SIRA: NonlNAL CASE ETAA: NonlNAL CASE 500AAE SEN5813V11T CASE IPLUS 10XI 50unAE SEN58tiVIIT CASE IPLUS 1021 F100AE 3 54 F100AE 3-58
TIGl?RE 3-7 j
SE'ISITIVITY STUDY - DECREASED TI!!E OF EJECT!0tl BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER 400.00-20-5 0-1 10.00-3.5-l.4-1 O
R n
n 0
p ll n
A; i.2-2 B.00-h t
ll 3
0 i
l g
g.0-'ii l
r 0
M P
O j
it u
0.0-A 0 80-O.4-l i
l "U
0.1-f m
0.08.
0.0,
0.0 0.5 t.0 3.5 2.0 2.5 30 3.5 4.0 45 5.0 II5 0.0 0.5 8.0 2.0 2.5 30 3.5 4.0 4
0 flME 15ECl fint 16tta STAA: NoninnL CASE 80.8 SECl 500RRE: SEN8lIIVIIT 5tuGT 10 05 SECl stan: NonlNat CASE 80 8 SECB SOUARE: SEN88tivlTV CASE 80 05 SECl FlGURE 3-7A F100AE 3-10 I
FIGURE 3-8 SENSITIVITY STUDY - !?lCREASED TI'!E OF EJECTIO!!
l BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE. HOT FULL POWER l
100.00-i.._
l l.5-10.00-0 4-Il gg Il
'i 0
8 A
5 g
R l.2-t 0
I ii l 5 0-h l
1 00-0 a
0 P
P l
l 0...
E I
R R
O.5-0.50-t 0.4-0.2-2
]
a it 0.01-0.0-e 0.0 0.5 l.0 3.5 2.0 2.5 3.0 3.5 4.0 4.5
$$0 00 0.5 l.0 1.5 2.0 2.5 3.0 35 40 4.5 50 0
TIME IEECB flNC 55Ecl SOU I vi A5 '80 EC1 rIOo.c 3...
riou.c...
FIGURE 3-9 J
SENSITIVITY STUDY - TRIP DELAY TI'1E 1
BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER 100 00-
~
2 0-iI l.5-30.00-l.4-A N
O 8
A a
n.ui L
L 8
i I.00-3 l
[
g I.0- o P
p 8
0 M
M 0.8-'
E E
A A
0.5-0.l0-0.42 V
0.3-eq 43 0.01.
0.02 00 05 l.0 15 2.0 25 3.0 35 4.0 4.5 5.0 0.0 0.5 1 0' l.5 20 35 3.0 3.5 4.0 4.5 50 IlME 15ECl flME 15ECl 51AA: WOMINAL CASE 10.5 SECl SIAA: NOMINAL CASE 10.5 SECB 500 ARES SEM51TIVIIV CASE II.5 EECI 500AAEs SEN51tlVIIV CASE 11 5 SECB FIGUAE 3-9A FIGURE 3-98 l
l
I I
l FIGURE 3-10 SENSITIVITY STUDY - TRIP 110RTli BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER 500.00-2.G- }
ii 8.8-1 5-10.00-3.4-a A
A l.t-1
\\
I I
l.00-5.0-o g
P P
0 l
0 l
g 0.8-N R
l Q
0.5-0 10-0.4-0.2-y 0.01-to 0.0 0.5 l0 l.5 20 2.5 3.0 355 4.0 4.5 5.0 flME IEECl IIME SEECl SIAR: NOMINAL CASE ST AR : NOMINAL CA5g SOUARE s SEN53 fivitV CASE ILE5510XI SOUARE: SEN51Tivilf CASE ILES$ 10X3 FIOURE 3-808 FIGURE 3-80A y
a-,
FIGURE 3-11 SEllSITIVITY SleDY - I!!ITIAL ZERO POMER LEVEL BEGINNING OF LIFE, HOT ZERO POWER END OF LIFE, HOT ZERO POWER 600.00-1 0-w int il I
1 f
10.00-80 00-N O
R M
A ll l
8 00-I.00-0 i
l ll
'l N
l 0 10-0 80-U l
=
.j rn 0 08-0 08 l
0.0 0.5 10 1.5 2.0 25 30 35 40 4.5 50 0.0 0.5 1.0 15 20 25 3.0 35 4.0 4.5 50 IIME 15ECI flME 15ECI STARS NOMINAL CASE li.0E-SS
&I3R3 NOMINAL CASE II.0E-Si 50UAREa SEN58tiVlIT CASE 80.0018 SOURRE : 5(N5lIlvlif CR6E 10.0011 FIGURE 3-llR FIGURE 3-188 l
l l
~ ' -
I
~
FIGURE 3-12 SENSITIVITY STUDY - BETA YIELD FRACTIONS BEGINNING OF LIFE, HOT 2ERO POWER BEGINNING OF LIFE. HOT FULL POWER 600.00-3 0-ll I
8.s-i t.5-80 04-n.4-M ii O
il A
i.t.
L L
l 8 00-i t 0-h
[
i r
0 N
0..-
y A
0.5-0.80-04-y'D 01-rri 0 0s-0 0$5 4.0 l$5 2.0 tl5 30 35 4.0 4.5 5.0 0.0 05 4.O t.5 2.0 3.5 30 3.5 4.0 45 50
+
IINE e5ECl linE 85ECS stAA: NOMINAL CASE GREIRAN CODE VALUE51 5tAA: NonlNAL CA5E 4REIRAN CODE VALUE55 SOUARE s SEN58 tivilt CASE 8 5arry Reload values) 50uAAt s SEM58 tiVIII CASE (Surry Reload values)-
FIGUAE 3-828 FICUAE 3-lIR
PAGE 95 r
l 3.2.2 point Kinetics Model Thermal Hydraulic Parameters Four parameters were evaluated in assessing the sensitivity of thermal hydraulic parameters on the RETRAM point kinetics calculation. These unre:
1.
gap heat transfer coefficient 2.
fuel pin geometry description 3.
core inlet temperature 4.
system pressure Table 3-2 presents a summary of the results in the format of Table 3-1 f
discussed previously, 3.2.2.1 Gap Heat Transfer Coefficient In the RETRAM point kinetics model, the fuel-clad gap heat transfer coefficient (HGAp) is modeled by using a constant conductivity for the gap material throughout the rod ejection transient. For sensitivity study purposes, the value of HGAp was decreased by 50% from its nominal value of 1000 Btu /ftz-hr
'F to 500 2
Btu /ft -hr
- F.
The results as
{
Presented in Table 3-2 show little sensitivity.
An described in Section 2, a change in the value of HGAp is expected to result in off-setting effects on the transient. A reduction in HGAp loads to a
decrease in the rate of heat transfer from the fuel to the coolant with a
resultant increase in the rate of fuel heatup. This increased rise in fuel temperature in turn results in a greater negative
}/
Doppler reactivity feedback effect which leads to a more rapid reduction in core power which in turn tends to reduce the fuel temperature.
PAGE 96 Figure 3-13 presents a comparison of the power histories for the tuo BOL Cases.
3.2.2.2 Fuel pin Geometry Description The standard RETRAN Single Loop Model describes the fuel pin with tuo concentric fusi pin regions, a gap region and a cladding region. For the purpose of the sensitivity
- study, the number of concentric fuel pin regions was increased from two to, six. As shown in Table 3-2, this change had little impact on the transient predictions. Figure 3-14 presents a comparison of the power histories for the two BOL cases.
3.2.2.3 Core Inlet Temperature The initial core inlet temperature was increased 4
'F from the nominal values.
(4
'T is typical of the c3 certainty assumed in the initial coolant temperature for a safety ana
.)
A change of this magnitude showed little effect as seen in Table 3-2 3.2.2.4 System pressure Typical uncertainty assumed for system pressure in a safety analysis is 30 psia. The initial system pressure in the point kinetics analysis was reduced by this magnitude to investigate the sensitivity to system Pressure.
As with the core inlet temperature, no noticeable impact on the transient result was found, (Table 3-2.)
PAGE 97 i
L TABLE 3-2 POINT KINETICS THERMAL HYDRAULIC SENSITIVITY STUDY
{
I.
BOL HZP Studies Peak Energy Norm.
Release f
Parameter Sensitivity Power (t=5 see)
{
Neminal case values Not applicable 51.1 2.23 Gap heat transfer coefficient Reduced 50%
51.0 2.02
- of fuel pin meshes changed from 2 to 6 51.1 2.15 j
Core inlet temperature Increased 4
'T 51.1 2.23 System pressure Increased 30 psia 51.1 2.23 II.
BOL HTP Studies peak Energy Norm.
Release Parameter Sensitivity Power (t=5 sec)
Nominal case values Not applicable 2.02 3.50 Gep heat transfer coefficient Reduced 50:
2.02 3.45
- of fuel pin meshes Changed from 2 to 6 2.02 3.45 Core inlet temperature Increased 4
- F 2.02 3.50 System pressure Increased 30 psia 2.02 3.50 Notes Peak Norm. Power = peak normali=ed power occurring during transient, (1.0 is equivalent to full pouer level.)
Energy Release (t=5 sec) total core energy release up to a
=
transient time of 5.0 seconds, (units of full power-seconds.)
k PAGE 98 l'
TABLE 3-2 (cont.)
POINT KINETICS THERMAL HYDRAULIC SENSITIVITY STUDY l
I.
EOL HZP Studiest Peak Energy Norm.
Release Parameter Sensitivity Power (t=5 see)
Naminal case values Not applicable 80.2 1.87 Gap heat transfer coefficient Reduced 50%
80.1 1.86 4 of fuel pin meshes Changed from 2 to 6 80.2 1.87 Core inlet temperature Increased 4
- T 80.1 1.87 i
Systen pressure Increased 30 psia 80.3 1.88 II.
EOL HFP Studies Peak Energy Norm.
Release Parameter Sensitivity Power i.t=5 sec)
Mcminal case values Not applicable 2 60 3 02 Gap heat transfer coefficient Reduced 50%
2.60 3.06
- of fuel pin meshes changed from 2 to 6 2.60 3.05 Core inlet temperature Increased 4
'T 2.60 3.02 System pressure Increased 30 psia 2.60 3.02 Notes:
Peck Norm. Power = peak normali=ed power occurring during transient, (1.0 is equivalent to full pouer level.)
Energy P.elease (t=5 sec) = total core energy release up to a transient time of 5.0 seconds, (units of full power-seconds.)
.~
~
FIGURE 3-13 s
SENSITIVITY STUDY - GAP HEAT TRMSFER BEGINNING OF LIFE, HOT ZERO POWER BEGINNING OF LIFE, HOT FULL POWER to.. 0-t..-
l ll l
j i
il l
g.e-I so.co-t.6-t.4-N O
N a
A e
n A
A n
5.2 L
i p
2 3 00-t I
e E
t.o-n e
0 P
N O
r N...-
A i.
e.,0 o.4-o M,
o.2-m e
...I
- W o.o o.s i.o i.s 3.s 3..
3.s 4.s s'..
O.o-
,,,,,,,,,,,,,,,,, 3,,
inne esect SI AA : NoninAL CA50 SouARE: 5tN5818vilt CASE ILESE 50ZI STAA: NonlNAL CASE FICuat 3 33A FIGUAC 3-138
l FIGUP.E 3-14 SEllSITIVITY STUDY - POINT KI!!ETICS FUEL PIM GE0 METRY BEGINNING OF LIFE, HOT FULL POWER BEGINNING OF LIFE, HOT ZERO POWER 500.00-2 02 ii Il l.E q
O.6-10.00-i l.4-N O
N
'l R
I.2-g A
M L
g il i
(
I I
t 1 0-Il I
l.00-D E
D 0.5-g E
I i
R 0 5-l 0 50 0 4-i l
0 2-
- 2
~
O.0 05 l.0 n.5 tbr 25 30 35 4.0 4.5 5.0 0.00-C 0.0 0.5 10 1.5 20 25 3.0 3.5 4.0 4.5 5.0 STARS NOMINAL CR5[ llWO FUEL AEGIONS$
SOUAREs SEM5tiltiff CREE 15lE FUEL SEGION5l ETAA MonlNAL CASE IfWO FUEL RfC10N51 500AAts 5(N55fivilf CASE 8512 FUEL REGleM5B FIGURE 2-148 FIGURE 3-14A i
Y
PAGE 101 3.2.3 Hot Spot parameters Sensitivity studies were performed with the RETRAN Hot Spot Model to evaluate the impact of various thermal hydraulic and neutronics parameters on the temperature and enthalpy history of the hot spot location.
For a given nuclear unit at a given initial power level, the only differences in the RETRAN Hot Spot Model between a BOL and EOL case are the assumed temperature of fuel melt and the input hot spot power history.
Therefore, the sensitivity studies were performed only at BOL conditions f'or both a =ero power and full power case.
The sensitivities investigated were:
1.
total power peaking factor 2.
power history shape 3.
gap heat transfer coefficient 4.
fuel pin geometry description 5.
inlot temperature 6.
system pressure 7.
mass flow rate 8.
metal-water reaction option Table 3-3 presents a summary of the sensitivity study results for the Hot Spot Model. The percentage of fuel melted, maximum fuel centerline temperature and clad temperature, and maximum fuel avarage enthalpy are compared for each sensitivity case with those predicted for a nominal case since these are the key parameters in assessing the potential for fuel damage and radiological release for the transient.
L PAGE 102 a
(
3.2.3.1 Power peaking Factor The core average power history predicted with the point kinetics model 10 weighted by a hot spot total power peaking factor (Fq) shape before input to the hot spot calculation.
The design uncertainty on Fq as predicted by Vepco's steady state physics models is presently 7.5%. The power histories for the two sensitivity study cases were increased by this factor (7.5%).
The results as presented in Table 3-3 show an increase in the temperatures and average enthalpy for both cases.
3.2.3.2 power History Shape One of the point kinetics neutronics sensitivies investigated was the time of rod ejection which is nominally set at 0.1 seconds for the rod to be completely ejected from the core. To model a similar type of consitivity with the Hot Spot Model, the hot spot power history was displaced 0.1 seconds later in time; i.e.,
the time of the peak power end all later powers were increased by 0.1 seconds. As seen in Table 3-3, the sensitivity was minimal.
3.2.3.3 Gap Heat Transfer Coefficient For a
HTP case, the gap heat transfer coefficient (HGAp) is initially cdjusted to produce a target initial radial temperature distribution for the fuel pin. At 0.1 seconds into the transient, this value is increased to 10000 Btu /ft2.hr *F and held there for the remainder of the transient to model the expected.
closure of the gap. Since the initial value of HGAp has no impact on the pin's initial radial temperature distribution I
PAGE 103 for a
HZp case, the value of HGAp is maintained at 10000 Btu /ft2-hr
'T for the entire transient in order to simplify the analysis. (Additional consitivity studies showed this change to have no significant impact on the Hot Spot Model's predictions for HZp cases.)
For the. sensitivity investigation, the values of HGAp used in the nominal analysis were reduced by 10%. This would be expected to lead to higher fuel temperatures since the ability of the fuel to transfer heat to the coolant has been reduced. The results presented in Table 3-3 show little significant impact from such a reduction.
o 3.2.3.4 Fuel pin Geometry De s c riptiori j
l 4
The fuel pin geometry of the Hot Spot Model is described with ten M
concentric fuel
- regions, a
single gap region, and three concentric cladding regions.
To test the sensitivity of this geometry, the number of fuel regions was reduced from ten to six. The results shown in Table 3-3 show little impact on the model's predictions.
An additional note is required on the percentage of fuel melt which is shown for the HFP case in Table 3-3.
For the sensitivity case this is listed as
<11.1 t[hile for the nominal case it is given as <9 %.
Although the acceptance criteria for percentage of fuel melt is <10 %,
the six node sensitivity case value of <!1.1 % does not necessarily imply that the acceptance criteria was violated. Due to the closeness of the comparisons of the average fuel enthalpy and fuel temperature histories between the nominal and sensitivity cases, the percentage of r
fuel melt for the sensitivity case is indeed <9 %,
(see Figure 3-15.)
PAGE 104 r
L Additional confirmation is presented in Figure 3-16 which provides a f
comparison of the radial fuel temperature distribution through the pallet for the sensitivity and nominal cases at the time of maximum fuel colt.
The simple technique used for determining percentage of fuel melt with the RETRAM Hot Spot Model merely puts upper and louer bounds on the parcentages.
As shown in Table 2-6, the estimate of maximum fuel melt boing <9 % for the ten fuel region nominal case is derived from the fact that the highest numbered node for which the transient fuel temperature did not oxceed the assumed melting point was node 4.
For the six fuel region sensitivity
- case, this table is invk1'id. The highest numbered fuel node for the six fuel region case which did not exceed the assumed colting temperature was node 3.
Applying the methodology used to derive Toble 2-6 to the six region case gives a parcentage of fuel melt somewhere between 2.8 M and 11.1 3.2.3.5 Inlet Temperature The initial inlet temperature to the hot spot volume was increased by 5
'T for the sensitivity study. As expected the maximum temperatures for both cases showed a slight rise over the nominal values (Table 3-3), but overall the effect was insignificant.
3.2.3.6 System pressure The initial system pressure was decreased by 30 psia. This produced a i
slight increase in the maximum clad temperatures for the sensitivity
PAGE 105 L
ocses (Table 3-3), but again the overall effect uao' minimal.
(
3.2.3.7, Mass F1,ou Rate Roducing the mass flow rate throughout the Hot Spot Model by 5% produced the most i$phet on the clad temperature where it would be expected to chou the most noticable effect (Table 3-3). The impact on the pellet conterline temperature, as shown in the HZp case
- results, was insignificant.
a 3.2.3.8 Me al-Water Reaction The RETRAN Hot Spot Model includes a calculation of the additional onergy release due to reaction between the coolant the the Zircaloy cotal of. the c1&dding. Turning off this option reduced the temperatures and enthalpy predicted by the model as expected (Table 3-3). Again the consitivity was sore pronounced in the clad than in the fuel.
l 4
l w
f PAGE 106 TABLE 3-3 NOT SPOT SENSITIVITY STUDY I.
HZP Studiest Max.
Max.
Max.
Fuel Tcl Tclad Enthalpy Parameter Sensitivity Melt
(*F)
(*F)
(Btu /lb)
Msminal Case Values Not Applicable 0
4017 2460 266 Pcuer Peaking Factor Increased 7.5%
0 4205 2609 285 Power History Shape 0.1 sec delay of peak 0
4018 2463 266 Gcp Heat Transfer Reduced 10%
0 4018 2456 266 Coefficient 1 of Fuel Pin Meshes changed from 10 to 6 0
4012 2471 266 Inlet Temperature Increased 5
'T 0
4020 2462 266 System Pressure Reduced 30 psia 0
4018 2474 266 Mass Flow Rate Reduced 5%
0 4019 2496 267 M3tal-Water Reaction Turned off 0
4014 2372 263 Abbreviations
% Fuel Melt = Maximum % pellet melting at hot spot Max. Tcl
= Maximum pellet centerline temperature at hot spot Max. Tclad
= Maximum cladding temperature at hot spot I
t
r-l PAGE 107 TABLE 3-3 (cont.)
HOT SPOT SENSITIVITY STUDY II.
Hrp Studies Max.
Max.
Max.
I ruel Tcl Tclad Enthalpy
________I Nominal Case Values Not Applicable
<9 4903 2285 317 power peaking Factor Increased 7.5%
<9 4904 2358 328 power History Shape 0.1 sec delay of peak
<9 4903 2297 318 Gcy Heat Transfer Reduced 10%
<9 4904 2304 322 Coe f ficie r.t 4 of Fuel pin Meshes Changed from 10 to 6
<11.1 4902 2296 316 Inlet Tamperature Increased 5 'T
<9 4903 223<
317 System pressure Reduced 3r psia
<9 4903 2298 417 Moss Flow Rate Reduced SX
<9 4903 2319 318 Matal-Water Reaction Turned off
<9 4903 2239 315 Abbreviationst X Tual Malt = Maximum % pellet melting at hot spot Max. Tcl
= Maximum pellet centerline tamperature at hot spot Max. Tclad
= Maximum cladding temperature at hot spot o
As described in Section 3.2.3.4, the actual X of fuel melt for the sensitivity case is <9%.
FIGURE 3-15 SENSITIVITY STUDY - HOT SPOT FUEL PIN GEONETRY FULL POWER HOT SPOT CENTERLINE TEMPERATURE FULL POWER HOT SPOT AVERAGE ENTHALPY 320-5000-380-4900-300-4000-C E
4100-200-A E
L n
[210-N 4600_
I A
L f
260-n 4500-P 8
259-
[
4400-S 240-
[
E i
[4300 230-O I
220-ggno.
280-4100-200-
=
=
=
0 8
2 3
4 5
6 1
8 9
to 4000,
0 I
2 3
4 5
6 1
8 9
to gimE ISCCI fint 85ECl N
A ST AR s NOMINAL IEN fu[L A[Cl0N CA5[
EIAas NC.tlNAL f[m FU(L Atolog cAgg 500AA[s SEN5tilVIIV EIE fu(L A(CION CA5[
50UAA[a $[m51IIVilt $11 fu(L REC 304 CA?"
g ficuRE 3-45A FIGUAE 3-858 W
v
PAGE 109 FIGURE 3-16 l
[
SENSITIVITY STUDY -- HOT SPOT PELLET TEMPERATURE O!STRIBUTION 5000-47E0-4500-4250-F U
E 4000-L TE 3750-l E
R 3500-R T
U R 3250-E l
0E 3000-G F
2750-2500-2250-2000 g
.g.
.g.
.g-
.g.
.g.
.g.
.g.
.g.
3 0.0 0.1 0.2 03 0.4 0.5 0.6 0.7 0.8 0.9 1.0 NORMRLIZE0 PELLET P.R01US STAR = 10 NODE PELLET GEOMETRY SQURRE = 6 N00E PELLET GEOMETRY
PAGE 110 L
SECTION 4 - VERIFICATION COMPARISONS 4.1 Introduction The purpose of this report is to document Vepco's analytical capability for performing reload core safety and licensing analysis for the rod ojection transient.
As verification of this capability, appropriate results and comparisons are provided for a representative series of enalyses of licensing transients.
A description of the licensing methodology and codes presently used by the vendor (Westinghouse) is provided below. These methods are currently cyproved for the licensing of Vepco reload cores and were implemented by Vopco to analyre six specific cases of the rod ejection transient--the four cases reported for the reload analysis of Surry 1 Cycle 5 (BOL HZp, BOL
- HFP, EOL HZP and EOL HFp), (Ref. 19), and tuo BOL cases (HZP and HFP) documented in the surry positive moderator coefficient submittal (Ref 22).
Rosults from the Vepco analyses performed with the vendor's codes and tochniques are compared to the vendor's results for the same analyses to domonstrate Vepco's ability to replicate vendor results. A comparison is then made between the licensing results obtained with the vendor cathodology (using vendor codes) to those obtained with the Vepco tethodology using the RETRAM code.
A comparison is also presented between the Vepco calculated results using the vendor and Vepco methodologies and results obtained by Vepco
f I
l PAGE 111 using the vendor's three-dimensional space-time kinetics code. This
(
comparison demonstrates the conservatism of the Vepco approach to the rod ejection analysis.
W 4
\\
PAGE 112 I
l' 4.2 Vendor Licensing Methodology The vendor licensing analysis of the rod ejection transient parallels that of Vepco in that the analysis may be broken into tuo main phases:
1.
a core average nuclear power transient calculation, and 2.
a transient thermal-hydraulic:s analysis of the core's hot spot.
Unlike Vepco, which uses a point kinetics model to derive a core average nuclear power
- history, the vendor uses a
one-dimensional (1-D) opace-time kinetics
- model, the TWINKLE code. (Ref. 18.) TWINKLE is a space-time neutron diffusion code which uses nuclear macroscopic cross noctions generated by the vendor's steady state physics design codes to solve the neutron diffusion equatio..: lor two energy groups. The code's goometry may be specified in one, two or three dimensions with up to 2000 spatial points. Six delayed neutrcn groups are assumed. A detailed Eultiregion, transient fuel-clad-coolant heat transfer model is included for predicting pointuise Doppler and moderator feedback effects.
The vendor uses the 1-D axial geometry configuration for the. rod ojection analysis (Ref. 17). Input for the model is specified to bound the conditions for a given nuclear power plant. This approach has been eccepted for providing the core average power for the transient. Thr; dolayed neutron
- fraction, Doppler reactivity feedback, trip insertion characteristics, trip setpoints, and total trip reactivity worth are conservatively adjusted to fit the core conditions and plant characteristics for the specific analysis under consideration. The coderator feedback effacts are adjusted by changing the value of the
PAGE 113 coluble boron concentration. As with the Vepco methodology, a weighting
(
factor is applied to the Doppler reactivity calculation to more escurately model the effects of severe flux redistribution in a
three-dimensional geometry.
Dnspite having an explicit representation of the axial core geometry, the ejection of the rod is not modeled by perturbing the nuclear cross noctions in a space-time dependent manner. Instead the code's eigenvalue 10 ramped over a 0.1 second time interval by a value equivalent to the casumed ejected rod worth. Upon reactor trip. (with an appropriate trip time delay),
a modeling of the trip banks being inserted into the core is performed.
This accounting of the axial effects is the major difference between the TWIHKLE model and the RETRAM point kinetics model where in the latter the scram is based on a generic normalized trip roactivity table.
The vendor thermal-hydraulics analysis of the core hot spot is performed with a
detailed fuel and clad transient heat transfer code, FACTRAN, (Ref. 9.) This code computes the transient temperature distribution in a cross section of a
metal clad uranium dioxide fuel rod, and the heat flux at the surface of the rod, using as input the normalized core average power history predicted by the TWINKLE code and the local coolant conditions for the actual plant and core conditicns under consideration.
The fuel rod geometry is input for the particular plant b3ing analyzed.
A Eircaloy-water reaction is modeled, and all material properties are represented as functions of tempurature. The fuel rod casumes an initial parabolic radial power generation.
I L
PAGE 114 The code uses the Dittus-Boelter or Jens-Lottes heat transfer correlations to determine the surface heat transfer characteristics b3 fore the onset of DNB. AT 0.1 seconds into the transient, the code is forced into DNB by specifying a conservative DNB heat flux, and the Bishop-Sandberg-Tong correlation is used to determine the film boiling hnat transfer coefficient. The gap heat transfer coefficient is adjusted in order to provide agreement of the full power steady state temperature distribution with the distribution predicted by Westinghouse design fuel hoat transfer codes. As the transient progresses, the gap heat transfer coefficient is ramped from its initial value to a very high value to model the gap closure expected to occur during the temperature transient (Ref. 17).
As with the Vapco methodology, the pre-and post-ejection design maximum total power peaking factor (Fq) is input. The value of Fq used by the code is assumed to increase from the pre-ejection value to the Post-ejection value over a
time interval of 0.1 seconds, and remain there for the duration of the transient.
(
In summary, the analytical techniques in modeling the transient of both Vopco and the vendor are similar. Table 4-1 presents an overvieu of the two methodologies.
Table 4-2 presents the results obtained for the six ocuparison cases using the TWINKLE /FACTRAM codes with vendor methodology Parformed by both the vendor and by Vepco. The results from the vendor cnalysis are provided in Ref.
19 for the Surry 1 Cycle 5 reanalysis ecses and Ref. 22 for the Surry +MTC cases. The close agreement between the results demonstrate; Vepco's ability to perform the rod ejection
PAGE 115 enalysis using the TWIHKLE/FACTRAM codes and vendor methodology.
k PAGE 116 r
L TABLE I4-1 COMPARISON OF VENDOR /VEPCO LICENSING METHODOLOGIES f
I.
Core Average Power History Calculation Item Vepco Vendor a
Principal code RETRAM TWINKLE
(
Physics model Point kinetics 1-D space-time kinetics Rcd ejection Reactivity ramp Reactivity ramp Doppler feedback Reactivity vs. temp.
Calculated from nuclear table cross sections Modexator feedback Moderator temp.
Calculated from nuclear coefficient cross sections f
Trip reactivity insertion Reactivity vs. time Calculated from nuclear table cross sections D31ayed neutron groups 6
6 Doppler reactivity Yes Yes w2ighting factor Fuel heat transfer model Yes Yes HGAP Constant Variable
[
[
[
[
1
l PAGE 117 r
TABLE 4-1 (cont.)
COMPARISON OF VENDOR /VEPCO LICENSING METHODOLOGIES
(
II.
Hot Spot Thermal-hydraulics Calculation Item Vepco Vendor Principal code RETRAN FACTRAM
(
Number of fuel pellet 10 6
regions f
Fq history Ramped over 0.1 sec Ramped over 0.1 sec Zircaloy-water reaction Yes Yes l
Pre-DNB heat transfer Thom Dittus-Boelter or correlation Jens Lottes Post-DNB heat transfer Bishop-Sandberg-Tong Bishop-Sandberg-Tong correlation
{
Forced DNB at 0.1 sec Yes Yes Mnterial properties Function of temp..
Function of temp, f
G0p closure modeled Yes Yes
{
l l
{
m___
P' AGE 118 I
(
TABLE 4-2 VENDOR /VEPCO ANALYSIS RESULTS USING VENDOR METHODOLOGIES (TWINKLE-FACTRAM CODES)
The vendor calculated value is followed by the Vepco calculated value soparated by a slash (/).
Surry 1 Cycle 5 Values Parameter l BnL HZP BOL HFP EOL HZP EOL HFPl Fuel Pellet Melting (X) 0/0
<10/<10 0/0
<10/<10 Mox. Fuel Average Temp. (*F) 3430/3432 4185/4210 3373/2949 3844/3794 Max. Clad Average Temp. (*F) 2500/2520 248S/2482 2317/2169 2151/2167 Max. Fuel Enthalpy (cal /g) 145/145 185/186 142/121 166/164
__ Surry +MTC Values _
Parameter l BOL HZP BOL HFP l Max. Fuel Average Temp. (*F) 2883/3394 3639/3630 Mcx, Fuel Center Temp. (*F) 3353/3918 4958/4874 f
Max. Clad Average Temp. C'F) 2123/2488 2013/2079 l
)[
l l
w
1 L
PAGE 119 I
L 4.3 Verification With Licensing Analyses r
{
Six cases of the rod ejection transient were compared for benchmarking the Vepco methodology with that of the vendor. Assumptions fc; each comparison calculation were matched as closely as possible between the f
two methodologies.
{
the greater flexibility of specifying reactivity effects in the Due to Vapco RETRAM
- model, the steady state values of the Doppler defect (without any additional weighting factor),
moderator temperature coefficient, and total trip rod worth were first calculated by Vepco using the TWINKLE code; these calculations used nuclear cross section input chosen to give conservative values for these parameters. These i
vclues were then used in the corresponding RETRAM point kinetics cciculation.
An exception was the reactivity feedback weighting factor used for each methodology.
For the Vepco calculation, the weighting factor (pWF) was obtained f: om Figure 2-6 based on the value of Fq casumed for each analysis. The corresponding weighting factor used in TWINKLE was derived from the data in Ref. 19 for the Surry 1 Cycle 5 roanalyses cases and from Ref. 22 for the Surry +MTC cases. Table 4-3 Presents a summary of the input conditions for the six cases.
L PAGE 120 1
TABLE 4-3 VERIFICATION COMPARISON CASES f
SICS SICS S1CS SIC 5 S+MTC 5+MTC BOL BOL EOL ivu LOL BOL Parameter HZP HFP HZP HFP HZP HFP Ejected rod worth (Pom) 900 300 840 300 920 160
(
D21ayed neutron fraction
.0055
.0055
.0044
.0044
.0059
.0059 Pre-eject.Lon Fq NA 2.55 NA 2.55 NA 2.55 Post-ejection Fq 15.2 5.46
'15.2 5.46 15.2 4.8 Zoro to full power Doppler
-1088
-1088
-850
-850
-1088
-1088 dofect (pcm)
RETRAN moderator temperature 7.7 5.4
-20.0
-30.8 7.7 5.4 coefficient (pcm/*F)
Number of operating pumps 2
3 2
3 2
3 TWINKLE reactivity feedback 2.725 1.3 2.32 1.3 2.725 1.2 usighting factor RETRAN Power Weight!.ng Factor 2.95 1.4 2.95 1.4 2.95 1.25 Motest S1C5 = Suv:ry 1 Cycle 5 S+MTC =
Strzy Posit' ave Moderator coefficient NA = not ctyplicable pcm = Percent mille.
(
PAGE 121 i
L 4.3.1 Core Average power History Rosults from the nuclear power transient analyses for the six cases are graphed in Figures 4-1 through 4-6.
Each figure presents a normali=ed power history comparison from the TWINKLE 1-D and RETRAM point kinetics
{
colculations and the corresponding normalized energy release curves. As with the sensitivity study results presented in Section 3.
a value of
(
1.0 on the " normalized power" axis represents a value of 100% ft.1 power while a
value of 1.0 on the
" energy release" axis represents F
cne-full-power second of energy. For the HZp cases, both the TWINKLE and RETRAM models assumed an initial normalized full power level of 1.0E-9.
The thrae HZp
- cases, (Tigures 4-1, 4-3 and 4-5),
show excellent cgreement between the power histories. The minor differences occurring ofter the initiation of the trip can be explained by the different Ecthods used in modeling the trip reactivity insertion. Whereas RE'; RAM uses a
table of normali=ed trip worth insertion versus time which is i'dentical for each
- case, TWINKLE calculates the change in core roactivity through an axial variation of nuclear cross sections with time.
This is a more realistic modeling of the insertion of the control bcnks and takes into account the impact of the axial flux shape on the f
rate of change of reactivity.
Although the maximum normalized core power predicted by the tuo r.ethods say differ for a
specific case, the effect is not significant due to sharpness of the power peaking portion of the curve. This is verified in the plot of total energy release where both models. track reasonably wall.
After the trip insertion
- begins, the normali=ed power curve 1
PAGE 122 f
predicted by the TWINKLE code asymptotically approaches a lower power lovel relative to that predicted by RETRAM and this accounts for the oventual divergence of the energy release curves.
For the three HFP. cases, (Figures 4-2, 4-4 and 4-6), the divergence bstween the two models is more pronounced. Typically, TWINKLE predicts a higher peak power level and maintains the power at a relatively high f
value longer than the RETRAM model. As with the HZp cases, a deviation in the predictions is noticable during the trip portion of the transient with TWINKLE once again settling out at a lower relative power level.
The energy release curves show reasonable agreement at the most critical part of the transient between the two models with a larger divergence'in the energy release predictions occurring only after the power level of the core has dropped to a relatively low value.
Comparison of the. trends inherent in the three HTp cases show the effect of the delayed neutron fraction and ejected rod worth on the transient.
The BOL and E0L cases for the surry 1 Cycle 5 reanalysis, (Figures 4-2 and 4-4), both assumed the same ejected rod worth. The only significant difference between the two cases was in the delayed neutron fractions, (0.0055 for BOL versus 0.0044 for EOL), with the higher value causing the power curve to hang. (this is due more to delay in the core response than to changes in core conditions.) That this effect is due entirely to f
the delayed neutron fraction is especially obvious from the RETRAM cnalysis uhere it was the only parameter which changed between the two calculations.
A comparison between the two BOL HFP cases, (Figures 4-2 and 4-6),
1 r
L PAGE 123 demonstrate a
similar trend. Here, the difference in the values of the dolayed neutron fractions is less pronounced than in the previous
- oxample, but the 5+MTC case has a significantly lower ejected rod worth than the surry 1 Cycle 5 case. This results in less reactivity insertion into the core which in turn causes less of a rise in fuel temperature end therefora a
less pronounced negative Doppler reactivity feedback effect to turn the transient's power excursion around.
Ccnsidering the differences in the two models used, the power history predictions compare well.
Both show the same characterics for the transients and demonstrate reasonably close agreement in actual nognitude.
f l
l l
l e
ii-
M.
n Fir.tlRE 4-1 NUCLEAR POWER TRANSIENT 51C5 BOL HZP NORMALIZED POWER HISTORY NORMALIZED ENERGY RELEASE 3**
too.oo-g 1.6 3.e N
o
+(***)
io.oo-A t.2 t(
j o
I.o A
g N
N A
E L
a o.e f
2 I.co-t o
a e
o..
o u
p E
5
,A E
o.4 o.le' o.2
~
N o.o o.s i.e i.s z.o a.s 3.o s.s 4.o a.s s.o Inc iste WESTINGHOUSE PROPRIETARY CLASS 2 o.o o.s i.o i.s 2.o a.s 3.o 3.s 4o 4.s s.o isnt i5cci,
g siaa: iwime i-o aut o
si..: iwi m e i.o nooct sMC8 AtlRAN FolNI RINCIICE MML m
soVAAL. A[l8MN FolMI RIMEI8CE noo(L I 8U8E 4*IO FloWRC 4-IA
FIGURE 4-2 NUCLEAR POWER TRANSIENT - 5105 BOL HFP NORMAL 12ED POWER HISTORY NORMALIZED ENERGY RELEASE 3.s :
e.a s-
+(a,c) l l
e 4
3.o-3.1s-
.)
a i.se-n a
s a
8 L
n l
a i. s
.i L
t*
.oo.
l'*
o P
a e
c i.s t o.is-A a
e i.:
L o.s -
s e
o.as-e.s
=
=
~
Q.)
o.oo-e.o a.s i.o i.s 3.o a.s 3.o 3.s
..e
..s s.o 0.0 lint istCl o..
o.s i.e i.s a.s s.s
..e e.s s.e u ne istC g
cn WESTINGHOUSE PROPRIETARY CLASS 2 m
M sipW: INinett 3-0 n00(L sima s Iginatt t.o 9 00(L un seUAGE t ACISAN F08ml nim (Ilts nSOfL stuARCs AtlRAN FelNI ElNtfits n00tL FIGuer 4-:A FICuat 4-se O
~
!IUCLEAR POWER TP.Nis!EllT - 51C5 EOL liZP NORMALIZED POWER HISTORY NORMAllZED ENERGY RELEASE 100.00-
- .8' In l.6 j
i l.4 M
80.0o-n A
5.3 L'
I l
a 0
A O
l.0 4
[
A n
L e
r 8
A 0.8 8
8.00-c
(
0 3
.i A
E 0.6 0
t M
g I...
A c
0.00-o.s 0.Q u.
e.o a.s..e i.s z.o a.s 3.o s.s 4.s-4.s 5.0 N-sont istci 4
a.ot-HESTitlGHOUSE PROPP.IETARY CLASS 2 0.o a.s i.o i.s z.o a.s 3.o-3.s
..e 4.s s.o isnt neci
,v, 5388I IWgunLE g-p n00fL sOUARC IA f
IN N MOOCL 600AAC.s A[IAAN FSIMI M M I M W R fICUAC 4-3R T100M i-30
~
NUCLEAP. POWEP. TPXiSIEHT - 51C5 E0l. lifP
~
WESTINGHOUSE PROPRIETARY CLASS 2 NORMALIZED ENERGY RELEASE NORMALIZED POWER HISTORY
]
3 25I 3.00-3.00-2 15-2 15 t.50-
+(a,c) i 2 50- s' 2.tSJ a
0 3 25 a
s -2.002 n
A L 2 00 e
I 4 3 75 i
t t0 4.15 i
i 1 50-
.t ii
[
4,50 M-h 8
- l.25 C
3 T
t.25 g
(R 'g.00 n
E t
[ I.00 0.15-0.15 0.50-
.I 0.50 0.25-
=
. =
=.
x
~
0.25 0.00.
0.0 05 30 35
.t.0 25 3.0 35 4.0 4.5 5.0 lane 85tCI O.0
.5
.0
.5
- .0 n.5 30 3.5 4.0
.5 50 8*88
?.
f lint GEEC8
[
j Elmar TW3 putt 3 0 noott e
E M E: REI N reist MIM ilC5 M L y
3 EDUASE lpA I N MC MODEL flGURE 4-48 g
0 Flount 4-en
_ - - ~
FIGURE 4-5 NUCLEAR P0kkR TRAllSIEllT - S+NTC 80L HZP NORMALIZED ENERGY REI. EASE NORMALIZED POWER HISTORY i. 0 '-
800 00-i.s.
l e
f, 3.4 -
n 0
A n
40.00-a i.
+(a.C)
Li I(
0 80-a 0
r n
"a r,
a 0.0 y
a n.00-g i
t 0.s D
L E
j O
M
.t c
0..
a l
1 0.io-5 0. D i ',
0.0 0.s i.0 i.s 3.0 a.s 30 2.s s.O 4.s s.0 ffME (SECl HESTINGHOUSE PROPRIElARY CLASS 2 8 8' 0.0 05 1.0 n.5 2.0 25 3.0 3.5 4.0 4.5 S.0 linC tlEtt SIMA: INiputt I-0 nooit 590A01: REIAnu 703MI MlWCilEE n00(L IWIMutt 3-0 ft00f t sipR:
REIRAM folNI RINElits n00EL flGU8E 4*58 y
s 50uhas:
FIGUAE 4-5A C.
m m
FIGUP.E 4-6 HUCLEAR POWER TP.ANSIEllT
-S*f1TC BOL JiFP MESTINGHOUSE PROPRIETARY CLASS 2 NORMALIZED ENERGY RELEASE NORMALIZED POWER HISTORY 3.00.
l I.4-z.n n.i-s.x-2.50
+(a,c) 5 8-8 85-i.0 i,
e A
n I.00 0
0.5-a A
t n.
3 2
0.5-I l.15 L
t 8
0 0.1-E h
O 0.5-E f
n O
C 1 25 83 0.5-
.T g
0.4 E
l.00 g
0.3-0.t-3 0.5-0.50-0.0-0.25 C.0 05 10 3.5 20 2.5 3.0 3.5 4.0 4.5, 5.0 fint l&ECl
- N.
0.0 0.5 3.0 3.5 2.0 35 3.0 35 4.0 4.5 5.0 lint t$tC8 s
SipA in88 Matt 3-0 n00f t o
I SOURAC 1A IEN 5 MODEL
- rico.c 4 5.
=
FIGUAE 4-En I
k f
PAGE 130 s
f l
4.3.2 Hot Spot Analysis R3sults of the comparisons between the FACTRAN and RETRAN Hot Spot Model predictions for the hot spot transient are presented in Figures 4-7 through 4-12 and Table 4-4.
All FACTRAM calculations ucre performed bcsed on the appropriate TWINKLE 1-D model predicted core average power histories, while RETRAN Hot Spot Model calculations urre performed based
(
cn the predicted RETRAM point kinetics power histories. Both hot spot medals used identical assumptions as to the Fq history curve input; i.e.,
between a time of zero and 0.1 seconds the Fq value was linearly reaped from the pre-ejection value to the post-e3ection value and held there fo2 the remainder of the transient.
Figures 4-7 through 4-9 Plot the fuel centerline transients at the hot spot location.
The curves for the HZp cases demonstrate good agreement t
and follow the trends predicted by the TWINKLE /RETRAM core average power
(
history predictions--i.e.,
RETRAN predicts a higher fuel temperatvre than FACTRAN later into the transient just as it predicts a higher power t
lovel than TWINKLE at that point. In none of the HZP cases did any fuel salt occur.
The corresponding HFP cases (Figures 4-7 through 4-9) demonstrate more daviation, especially in the tuo Surry 1 Cycle 5 cases where fuel melt occurred.
For the RETRAM cases, upon reaching the assumed fuel melting t3mperature, (4900
'T for the BOL case and 4800
'F for the EOL case),
the temperature remained constant for some time due to the high heat of fusion required to melt the fuel, (Ref. 8). Although FACTRAN assumed the came fuel melting temperatures as RETRAN, it apparently has a smaller
pAGE 131 value for the heat of fusion of uranium dioxide. Thus the FACTRAM
(
temperature can at first level off at the molting point, and resume its rise sooner than the RETRAN Hot Spot model temperature. The main differences between the two predictions appear to be in the material properties tables used by the two codes. To check this, the TWIKKLE power history used by FACTRAN uns input to the RETRAN Hot Spot Model.
The RETRAM predicted temperature transient characteristics uere consistent uith those predicted by the RETRAN Hot Spot Model using a RETRAM point kinetics power history input.
Figures 4-10 through 4-12 are comparisons of the outer clad temperature transients predicted by the vendor and Vepco models. Again agreement is better for the HZP
- cases, although for all cases the curves show excellent convergence late in the transient.
The slight deviations between the predictions at the very start of the transient (where the FACTRAM predicted clad temperature actually decreases in value) are due to modeling differences between the tuo models of the gap and surface heat transfer coefficients before the onset of film boiling.
Table 4-4 is a
summary of the maximum values of the hot spot temperatures, average enthalpies and percentage of fuel melt for the two models.
As described
- above, tuo Surry 1 Cycle 5 HFp cases experience fuel melt.
For these
- cases, the FACTRAM temperature rises above the assumed melting temperature while the RETRAN temperature has yet to overcome the heat of fusion input into that model's material properties tables.
The FACTRAM code predicts an actual percentage of fuel melt while RETRAM predicts only an upper boundary based on Table 2-6.
m.,
PAGE 132 1
.e o
1l a
3 I
W g' "
m a
a w
o E
A u
.s
~,
w y
e.,
E_
_J o
.e-
'=
as D
A 37
- in a
u I'"*
e w
'a
.O.
w
=
e
. O.
~
m
&zw m
m
.e O
. o p.
o o
o o
o o
o O
O O
o o
o o
o o
m w
o.
o.
o.
o.
e a:
T D
wwa=mses=aw==w r e.wm a= 3e w ee.d b
>=
w c.:
c.
=
w C
o in
. a g
w w
F.
e
'4 wm m
.o
.a n o ew
>=
=
J a
u g
W 0
(L3 C
q
.,,J w
~3 w
a m
o wt.
A W
e A
S <,;;; ;
H
~
u o
o w
w
.E.wsi 3 c.
m o
y
'a
@C g
t9 s
O E
Z
. o.
. *a N
6 6
6 6
S' 6
6 6
6 6
6 6
6*
8 o
o a
o.
.o
.o o.
- o. -
o o
.o.
w e
ww 3a heg age.3 w whs E hWS Sm 360s# ow@ h
FIGil9E 4-8 I
l I
l HOT SPOT FUEL CENTERLIllE TEt1PERATURE TRNtSIEftTS - 51C5 E0L CASES l
i t
t HOT FULL POWER l
I HOT ZERO POWER l
l l
4000-f 4300-1 4850-l 3900-4400-r 9
1600-l 4150-C 3300 E
t t
4100-53000-E 8
t i
B 4650-3100-
"I N
t 1 4600-E 1 IA00 R
tn 4550-E 3:00-9 a
g 4500-a
- u. 1800-k I
t 0 6450-r 0 1500-g*
o r 4400-F 4100-4350-000-4300-500 g-m 0.0 0.5 1.0 9.5 2.0 - 35 3.0 1.5 4.0 4.5 50 00 0.5 s.O a.5 3.0 35 1.0 3.5 40 4.5 5.0 etSO-M 34-IInf 46tti
[
IInt 15ttl Sims. thCIR>m E00C w
5popst. ettsom pel 6rel n00tL SlpA: FAtlppu E00t
$6uart. ellamp n01 SPol n00tL flGust a.94 FJCutt 4-SA 3
l l
FIGURE 4-9 HOT SPOT FUEL CtNTERLINE TE!!PERATURE TRANSIENTS - S+ffTC CASES l
l
\\
l HOT FULL POWER I
HOT ZERO POWER 4900-
)
4100-4050-3900-4000-3500-4150-1300-N l4100-N g1000-F l
8 4650-k1100-i i
l4600-f400-
?
g4550-gisCO-o a500-u 1800-E E
0 4450-O 4500-G G
F 4400-F 8200-4350-900-V 4300-a 606, g l
g 4250-W A
100-9.0 0.5 t.0 -
6.5 f.0 2.5 3.0 3.5 4.0 4.5 5.0 0.0 0.5 s.O t.5 f.0 f.5 3.0 3.5 4.0 4.5 5.0 t;#ig 35fCI flRf 55tti 50unwr:ats[amNO: Sr01 n00ft.
50un' t. A lapu 608 af01 facung 4.ps FJGust 4 3A
PAGE 135
. o.
O e
. o.
N
. W.
y M
S:
E o
.o g
a o
A W3 =
o m
u o
W W
M
= -
N 4
'E'* "E D
E Ng i v
o :
E g p,
A
'a g
@C m
e O
y W
M
=
E m
l o
m D.
.W
.o NE o
E a
a s
4-i i
e i
e
- -P ' d o
o o
o o
u o
o o
o o
O P-o.
o o
o o
o o
o o
o o
m N
o e
e.
g y
N We N
a=
m es o
a w
Y
. D e* W E W ad E o
>WCR. WEE
- DEW OWC an.
D L.:
>==
Cr.. <
C' C
W 9:
L&
- ,W p
. o.
O m
Ju C
W
&D
. o.
O 4
J W.
e
'A f
M s
S:
=
O
.=
W-c.
O n
o o
^
W gE o
N. E O -
H.
o
=
O
$5 W
rz3 NEN8 N
.o C
a E;
e-O Z
~1 3
E
. o.
. N.
o r
,4 o e
o o
o o
o o
o o
o o
o o
o.
o o
o o
o o
o o
e ce C
e.
W w
N.
o.
eg og se a
m ed e
w e.
so
. D e= tad W W.J E o - W C b b E E e= 3 E dad o ind Q h
==
h f
1 PAGE 136
. O.
r t
- A; O
~E a
O C
M
.O w
5 O g
~
U g
d o
e 4
=O m
=s w
'M v.
w b
E$ k u
,a O
a g= -
aO w
b w=
m g
G Oa m
. O.
mD
- a C
O
-e N
N
. *g m
O O
O O
O O
O O
O O
O O
O O
O O
O e
e t-e e
c.e O
r-W Ft 8
W
- e
==
m e
e ee ce w
gir 2D Op wes usse a-wma.wase-sesw OwO m y
p.
k a
w
- e C1.
w Cw C
4 a
O
~4 wb
. O.
o 2
w a
- 2 a
\\
~
g u
M E
s
.9 g-a o
- ~5 1 4
. =
r.Q
,= --
Q
~
W s=
g
- * ~
..
- g M
gM '-
i t4 a
O
. g N
A Z
. O O
M e O 4
A 4
6 f
g h
0 8
8 O
=
8 8
4
=
0
~
,,,,,, os.O -ww w==-===
FIGURE 4-12 H0T SPOT FUEL OUTER CLAD TE!PERATUP.E TRANSIENTS - S+!!TC CASES HOT ZERO POWER HOT FULL POWER 2400-2800 '
2000-ON 1900-8000-2000-0 U l100 U
j i
1 E
E 4000 R 1500-C C
L 8530-'
L A
A 1500-l 0
0 5400-l I
E E
I, n 1300-n 8400-
{
l
- i=004 l
A i
1200-i y8800-A f
C E
l 8000-l 0 1000-(
E E
O 900-0 F
F 800-800' 100-500-500-Id, 500-400-;
0.0 0.5 n.0 n.5 2.0 2.5 3.0 35 4.0 4.5 5.0 ilnE 86Ecl 00 05 10 4.5 2.0 35 3.0 3.5 4.0 4.b 50 o
Tine 15ECl kg*
EDUARE; REIRAN HDT $ Of n00EL I N 0 5 I MOOR FIGURE 4 42A
PAGE 138 TABLE 4-4 FACTRAN/RETRAM Hot Spot Model Comparisons The FACTRAM calculated value is followed by the RETRAM calculated vclue separated by a slash (/).
Surry 1 Cycle 5 Values Parameter i BOL HZP BOL HFP EOL HZP EOL HFPI Fuel Pellet Melting (%)
0/0
<10/<10 0/0
<10/<10 Max. Fuel Center Temp, ('F) 3954/4003 5017/490,3 3390/4096 4889/4802 Mnx. Clad Temp. (*F) 2520/2484 2482/2292 2169/2568 2228/2172 Mox. Fuel EnthalPy (Btu /lb) 261/267 334/318 219/278 301/299
__Surry +MTC Values _
Parameter i BOL HZP BOL HFP I Fuel Pellet Melting (X) 0/0 0/0 Mex. Fuel Center Temp. (*F) 3918/3990 4874/4733 Max. Clad Temp. (*F) 2488/2471 2079/2036 Max. Fuel Enthalpy (Btu /lb) 258/266 280/275
PAGE 139 84.3.3 conclusions comparisons between results predicted with the Vepco and vendor octhodologies for the analysis of the rod ejection event for Vepco nuclear units showed similar trends and close agreement despite differences in the two models. For the cases analyzed in this report, both methods predicted a comparable degree of fuel melting for the cases in which it
- occurred, but for all cases it was under the 10% limit imposed by the acceptance criteria. Likewise, neither method predicted a violation of the acceptance criteria for the peak clad temperature or Paak average fuel enthalpy.
PAGE 140 4.4 Comparison to Three-Dimensional Space-Time Kinetics A
three-dimensional space-time kinetics code was used to verify the acceptability and conservatism of the point kinetics approach to predict the core average power history. As noted by Yasinsky, (Ref. 20), the escuracy and acceptability of the point kinetics approach is dependent both on the method used ano the particulars of the nuclear core under cnalysis. The use of a weighting factor to increase the magnitude of the Doppler reactivity feedback used in the point kinetics model (as is done with the Vepco methdology) is appropriate as indicated by the verification of the conservatism of the approach.
The TWINKLE code (Ref. 18) was used to analyze two rod ejection cases using three dimensional geometry:
1.
BOL, HZp. ejected rod worth of 1024 pcm 2.
BOL, HTP, ejected rod worth of 200 pcm The results from these analyses were compared to the results from the RETRAM point kinetics method and the TWINKLE 1-D method for similar transient conditions in order to verify the conservatism of the latter two methods.
e
_m__ _ _ _ _. _ _ _ _ _ _
t 1
l i
PAGE 141 4.4.1 Three-Dimensional Model A
three-dimensional (3-D) TWINKLE model was constructed for an initial Surry unit fuel loading (Cycle 1) for a core consisting of all fresh fuel assemblies. A quarter core loading map of the cycle is presented in Figure 4-13, and gives the initial fuel enrichment of the three batches (u/o U235) and number of fresh burnable poison rodlets (# BP rods) present in each assembly.
Appropriate nuclear cross section data describing the core loading was derived from vendor steady state core Physics design
- codes, and input to the TWINKLE code following the procedures outlined in Ref.
18.
A three-dimensional geometry was specified with one radial mesh point per assembly, 11 axial mesh points par
- assembly, one mesh point per top and bottom axial reflectors and a Icdial reflector region one mesh point deep at the nearest approach of a poripheral fuel assembly. A half core geometry model uas used with the full core being split along the vertical axis giving a total of 1989 nash points.
The location of the ejected rod is on the vertical axis through the center of the core and produces a radially symmetric pouer end flux distribution about the vertical axis. Therefors a half core goometry is sufficient for modeling the transient in three dimensions.
(See Figure 4-14.)
The patterns used for withdrawing the control rod banks from the core from HZp to HFP conditions cause the ejected rod to be located on an axis of symmetry.
The steady state radial power distr: butions at HZp and HFF for an all rods withdraun core configuration were normalized to those calculated by the vendor for the initial FSAR for the Surry units, (Ref. 21.) This l
l e
i PAGE 142 normalization was carried out by adjusting the fast diffusion coefficient for the radial reflector region and the macroscopic absorption cross sections for the various fuel batches until reasonable agreement was obtained with the radial power distributions published in Rof. 21. The resulting power distribution comparisons for initial steady state, all rods withdrawn core conditions are presented in yigure 4-15.
In a
similar
- fashion, the remaining critical cora parameters as predicted by the steady state 3-D TWINKLE model were normalized to values typical of those used in a rod ejection analysis. yor example, the zero to full power Doppler defect was normalized to a value of -1164 pcm by adjusting a multiplier to the fuel temperature component of the cocroscopic fast absorption cross section.
Typical values of the moderator temperature coefficient predicted by the model were obtained by Tdjusting the soluble boron concentration input to the code.
Control rod worths were not:malized by adjusting the thermal absorption cross section used to mode l the presence of control rods. Using the values provided by the steady state physics code for these cross noctions and ejecting the rod from the rod insertion limits specified for the initial Surry cycle resulted in ejected rod worths of such small magnitude that only minor power excursions were produced. To model transients which were more typical of those analyzed for reload cores, the insertion limits were deepened and the worth of the ejected rod increased by increasing the thermal absorption cross section of the D 4
bank.
This resulted in a
HZP ejected rod worth of 1021 pcm (percent mille) for an initial core configuration of the D bank fully inserted
h PAGE 143 cnd the C
bank inserted 43.5%
into the core. For the HFP case, an
(
ojected rod worth of 200 pcm resulted with the D bank inserted 56.5%
into the core. All other control and shutdown banks were initially out f
of the core.
For simplicity of modeling, reactor scrams were modeled by inserting the control banks (banks D,
C, B and A) less the ejected rod. The thermal cbsorption cross section for banks A,B and C was modified to yield conservative total trip worths typical of those used in the rod ejection
- analysis, (i.e.,
approximately 2000 pcm for the HZp case and 4000 pcm for the HFP case.)
Table 4-5 presents a summary of the steady state core physics parameters for the final 3-D TWINKLE model used for the benchmark cases.
The tuo transient cases were initiated by ejecting the chosen rod from the ' core at a
constant velocity over a 0.1 seconds time interval.
RSactor trip was initiated on the same trip setpoints assumed in the RETRAM point kinetics analysis and assumed a 0.5 seconds trip delay batueen activation of the trip and the start of rod motion. The trip f
rods entered the core using the same rod insertion model as is assumed in the standard TWINKLE 1-D analysis of the transient. Since the 3-D model predicts the effects of three-dimensional flux redistribution during the transient, no weighting factor was applied to the calculation of the Doppler reactivity feedback.
PAGE 144 i
FIGURE 4-13' SURRY UNIT 1 CYCLE 1 CORE LOADING PLAN (Eighth Core Geometry)
H G
F E
D l
1 08 l 1
1 I
I l
1 1
I I
I 09 l 2
l 1
I i
12 I i
4 Fuel w/o l
I I
Batch Assys U235 I
I I
l 10 1 1
1 2
l 1
l 1
53 1.85 I
i 12 l l
2 52 2.55 I
l l
l 3
52 3.10 1
I I
l l
11 1 2
l 1
1 2
l 1
l l
12 l l
12 l l
1 1
I I
I I
I i
l l
l 12 I 1
1 2
1 1
l 2
1 1
1 I
i 12 I i
12 I I
l I
I 1
I i
i f
I I
I l
l I
13 1 2
1 1
1 2
1 3
1 3
l l
12 l l
12 l 12 l l
l l
1 1
I I
I I
I l
l 14 l 1
1 3
1 3
l 3
l l
l 12 l l
l Legend:
i I
I I
i i
I I
I i
15 1 3
1 3
I i
x l
==> Batch #
l l
l l
x;t 1
==> # of BP Rods l
I I
I I
f
PAGE 145 FIGURE 4-14 RADIAL GEOMETRY FOR 3-D TWINKLE H
G F
E D
C B
A 01 l l
l I
I l
s...:...:...:...:...:
02 l D l l A l l
8 l
1 l
l l
8...:.........:
03 l l sal l
l l
i l
i l
l l
04 I I
l B l l C l l
1 I
I I
I I
I 05 I I
l l SBl l
l l
I I
I I
l l
l 1...:...:
06 i C l l D l l B l l A l Legend:
I I
I l_
i I
i I
05 I I SBl l
l l sal I
l l
I I
I l
l l
l l...:
1 I==> Fuel Mesh Point 08 I l
l C l l
l l D I l
1 l
l l
1 l
l l
l l
1...:
09 l l SBl l
l l sal l
l l
I I
i l
l l
l l...:
==> Reflector Mesh 10 lC i 1 D l l B l l A l Point i
I I
I i
l I
l...:...:
11 1 I
I I SBl l
l l
I I
I I
l l
l I...:...:
Control Banks:
12 I l
i B l l C l I
D I
I I
I I
I l...:...:...:
C 13 l l sal l
l l
B l
l l
l l
1...:...:...:...:
A 14 l X l l A l l
SB l
I I
I l...:
SA 15 I i
l X = Location of I
l l...:...i Ejected Rod
_.m.m__
___m___
PAGE 146 FIGURE 4-15 STEADY STATE RADIAL POWER DISTRIBUTIONS Beginning of Life. Hot Zero Power, All Rods Withdrawn H
G F
E D
11.1301 08 11.2101 l-6.6 I I
i 11.1s011.1201 1
09 11.22011.1901 1-3.3 l-5.9 l l
l l
11.12011.16011.0901 10 11.17011.18011.130l l-4.3 1-1.7 l-3.5 l l
l l
1 11.15011.10011.11011.0001 11 11.15011.12011.09011.0001 1 0.0 l-1.8 l 1.8 1 0.0 l l
l l
l l
11.10011.13011.030l0.97010.7801 12 11.09011.09011.01010.91010.7201
! C.9 1 3.7 l 2.0 l 6.6 I 8.3 l l
l l
l l
l 11.140.1.08011.04010.92010.6301 13 11.07011.04010.97010.91010.6201 1 6.5 1 3.8 1 7.2 l 1.1 1 1.6 I I
I l
l l
l 11.06011.12011.00010.6501 14 11.00011.12011.08010.6701 I 6.0 1 0.0 1-7.4 l-3.0 i Legendt i
I I
I i
10.93010.7101 lx.xxxl
==> FSAR Relative Power 15 10.94010.7601 ly.yyyl
==> TWINKLE 3-D Relative Power 1-1.1 l-6.6 l
l z.=
l
==> % Difference l
I I
I l
PAGE 147 FIGURE 4-15 (cont.)
STEADY STATE RADIAL POWER DISTRIBUTIONS Beginning of Life. Hot Full Power, All Rods Withdrawn R
G T
E D
11.1901 08 11.2401 l-4.0 I I
i 11.23011.1701 09 11.24011.2201 1-0.8 l-4.1 I
l I
l 11.16011.20011.1301 10 11.20011.20011.1601 l-3.3 1 0.0
~2.6 l
1 1
I i
11.18011.12011.13011.0201 i
.11 11.17011.15011.110l1.0201 1 0.9 l-2.6 l 1.8 1 0.0 l l
1 1
I i
11.11011.13011.04010.97010.7901 12 11.10011.10011.03010.93010.7501 1 0.9 1 2.7 1 1.'
1 4.3 1 5.3 I I
_I I
l l
l 11.12011.06011.01010.90010.6301 13 11.06011.04010.960l0.91010.6401 1 5.7 l 1.9 I 5.2 1-1.1 1-1.6 l 1
I i
l l
i 11.01011.06010.95010.6301 14 10.97011.07011.02010.6601 1 4.1 1-0.9 l-6.9 l-4.5 l Legend:
l I
l l
l 10.87010.6701 lx.xxxl
==> TSAR Power
(
15 10.89010.7301 ly.yyyl
==> TWINKLE 3-D Power
)
1-2.2 l-8.2 I i z.z !
==> % Difference i
I I
I I
l
PAGE 148 TABLE 4-5 3-D COMPARISON CASES BOL BOL Parameter HZP HFP Ejected mod worth (pcm3 1024 200 D31ayed neutron fraction
.0059
.0059 Pre-ejection Fq NA 2.55 Post-ejection Fq
- 10.6 4.33 Trip rod worth (pcm) 2210 4069 Zaro to full power Doppler
-1164
-1164 defect (pcm)
TWINKLE 3-D soluble boron 2260 1700 concentration (ppm)
RETRAM moderator temperature 2.9 1.5 coefficient (pcm/'F)
Mumber of operating pumps 2
3 TWINKLE 1-D reactivity **
1.74 1.2 foedback weighting factor RETRAM Power Weighting Factor 2.3 1.25 h
Notes:
Peak 3-D steady state nodal power weighted by. a generic pin-to-box ratio and uncertainty factor Weighting factor for 3-D TWIhKLE 1.0
=
1 NA = not applicable ppm = parts per million percent mille Pcm =
PAGE 149 4.4.2 Comparison Results In order to select the proper power weighting factor (PWF) for the RETRAM point kinatics analysis for comparison to the 3-D analysis, a hot spot total power peaking factor (Fq) for the condition of the rod being ejected must first be known. Since the core conditions being analyzed are not comparable to those which would typically be predicted for Surry Unit 1,
Cycle 1,
(i.e.,
the ejected rod worths have been arbitrarily increased),
values of Fq used in the comparison analysis were derived from the 3-D TWINKLE model. Steady state values of peak Oore nodal power at initial control rod core configurations less the ejected rod were calculated with the 3-D TWINKLE model for both HZP and HFP conditions.
(The HFP calculation assumtd froren thermal hydraulic feedback.) These values were increased by a 20% generic pin-to-box ratio to convert them from peak nodal powers to hot spot total power peaking factors. An additional 10%
was added for uncertainty. As presented in Table 4-5
- above, the assumed post-ejection total power peaking factors were therefore 10.6 for the HZP case and 4.33 for the HFP case. The PWFs for the RETRAN analysis were derived from these values of Fq based on Figure 2-6.
Likewise, the reactivity feedback weighting factors used in the 1-D
/
TWINKLE analysis were based on these values of Fq.
Typically, the higher the value of Fq assumed for the analysis, the less sovere will be the power history predicted by the point kinetics calculation since a
higher value of Fq implies a higher PWF which in turn causes a lower power history curve to be predicted. The derivation of the Fq values was based en steady state 3-D TWINKLE calculations
i L
PAGE 150
(
l instead of the transient calculations for this
- reason, since'the transient calculation derives some benefit from thermal hydraulic faedback effects and therefore predicts lower Fq values than the steady ctate calculation.
This is the method used for core reload analysis where the Fq values are obtained with steady state physics codes ocsuming frozen thermal hydraulic feedback.
The peak nodal powers predicted by the TWINKLE 3-D model with the rod ejected from the core (cad before the initiation of the scram) were as follows:
HZP HFP Post-ejection steady state 8.04 3.28 Transient 8.39 2.59 In essence, the use of conservatively high values of Fq will yield lower RETRAM point kinetics power history predictions and therefore lead to a closer power history comparison uith the TWINKLE 3-D prediction than might otherwise be expected.
From the standpoint of the hot spot i
calculation, the lower core average power history will tend to lower the Predicted peak temperature and enthalpy predictions. But this is offset by the higher Fq values used to weight the core average power history which will in turn cause the peak temperature and enthalpy predictions to be more conservative.
The latter effect is the more dominant, onpecially for the HTT case where the pWF has little effect on the core k
overage power history prediction.
Figure 4-16 presents the core average power history comparisons. For both the' HZP and HTP cases, the RETRAN point kinetics predictions are conservative compared to the TWINKLE 3-D prediction thus verifying the
PAGE 151 acceptability of the point kinetics methodology in general, and the usage of the PWF in particular. The conservatism of the point kinetics approach is even more apparent in Figure 4-17 uhich compares the total energy release of the two models. Both figures shou reasonably close agreement between the RETRAN point kinetics predictions and the TWINKLE 1-D predictions.
The hot spot power histories input to the RETRAM Hot Spot Model were calculate'd using the method outlined in Section 2 based on the Fq values Provided in Table 4-5.
A similar method was used for the FACTRAN calculations based on the TWINKLE 1-D predicted core average power histories.
That is, the valuE'$f Fq as a function of time was linearly ramped form its pre-ejection value to its post-ejection value over a time interval of 0,1 seconds and maintained at the post-ejection value for the remainder of the transient. The core average power histor-is then multiplied by this Fq curve to derive a hot spot power history.
A different approach was used to derive the hot spot power histories for the FACTRAM calculations for the 3-D cases. For these cases, the hot spot power for a
specific time was found by multiplying the 3-D core average normali=ed power and the 3-D peak nodal power (increased by a 20% pin-to-box ratio and a 10% uncertainty factor.) This method was used to reflect the fluctuation of Fq throughout the transient. Typically, the Fq value at any time will be less than that assumed for the point kinetics methodology, However, because of the flux redistribution which results from the novement of the scram banks into the core, it is Possible for the peak normalized nodal power in the core to exceed that
PAGE 152 casumed for the post-ejection condition.
Use of a constant generic pin-to-box ratio to convert the peak normalized nodal power to a Fq value may result in a 3-D Fq prediction later in the transient which oxceeds that calculated for a post-ejection condition by steady state nothods.
The actual impact of this on the transient is mitigated by tuo offects:
(1) the actual core location of the hot spot in the core will bs expected to shift throuahout the transient, and (2) the actual power lovel of the core during the trip is relatively low compared to that at the transient peak. Therefore, the Fq transient shape input to the point kinstics analysis would still yield a
conservative hot spot power history compared to what would actually transpire in a three-dimensional core.
Figure 4-18 presents a comparison of the hot spot power histories input to the RETRAN Not Spot Model and the FACTRAM code for the three Esthodologies under comparison. Since the 3-D TWINKLE derived hot spot power history is appreciably lower than that of the RETRAM point kinetics calculation, it may be reasonably predicted that the results of the RETRAN hot spot analysis based on the RETRAN hot spot power history will be conservative compared to the FACTRAN analysis results based on the 3-D TWINKLE power history.
Tchle 4-6 presents the results of these analyses based on the hot spot Power histories presented in Figure 4-18. Plots of the fuel centerline cnd outer clad temperature transients for the three models are presented in Figures 4-19 and 4-20. As expected, the RETRAN point kinetics and TWINKLE 1-D based analyses are conservative compared to the 3-D analysis.
l l
l PAGE 153 TABLE 4-6 3-D HOT SPOT MODEL COMPARISON RESULTS I.
HZP Case 1-D TWINKLE 3-D TWINKLE Parameter
/FACTRAN RETRAN
/FACTRAM Fuel Pellet Melting (X) 0 0
0 Max. Fuel Center Temp. ('F) 4346 3872 2659 Max. Clad Temp. C'F) 2869 2387 1703 Max. Fuel Enthalpy (Btu /lb) 296 255 161 II.
HFP Case 1-D TWINKLE 3-D TWINKLE Parameter
/FACTRAN RETRAN
/FACTRAM Fuel Pellet Melting (X) 0 0
0 Mcx. Fuel Center Temp. (*F) 4819 4685 4419 Max. Clad Temp. C'F) 2037 1990 1851 Mnx. Fuel EnthalPy (Btu /lb) 274 268 235
fl O m
[
' m PAGE 154
. O.
11
<r w
.J
-e O
g M
4 w
a.o 6 4
.awO a-I
==
2:.
w w
cll
=
D V,
- - = = = - - - =-+
g
,g w
LA=
>=
Mw_-
.=
Oc=
P
- O m
M eC w
y w
D Z
. O.
w
.J
. mn.
wO M
C e.,6% o
.A
. O.
OO w g e
C e=
C ens c
==
O w
=-
O.a m w
a,
..a.
- m N
wa q
O W
W2K
.E.
J=2 x0
=
2k2 I
. O.
M G
ew 22 w
EW
=
- e= s en er.wo h
.=a ag7 o
.. a-Y w ac M
gm
. O.
a 3
a=
Ow
. an.
O w-m O.
i 4
e 5
C O
O O
O
=
0
, O, O.
O.
=.
O O
O O.=
,.s O
a Y
&MMW Y
l n
u.o PAGE 155 w
+
.. O..
. O.
a N
g n
O V
W.
C D
W J
kJ
. O.
d*E a
O-O v
k a
O-c sw O
o
=O O
Ce A
=,:..::
N
)-
w z'" tad y
.J w
a c
-a M
n:
O==
EO2 M
=
= + -
zu-a
)
2 o
g
'i-
~
-w c.
.= e 4.s a
E tadd,
- tPS CE E
bi o
m w
2 m
t.,-
s **
C.
o
=
M 3
e..
W
-V
- =4 Z
H W
D
..w C2 W
O
.~.7 Oo 6
cm o
e O
.n O
.n O
in o
.A O
O
-e o
e.
r en.
ce.
O.
P.
in.
N.
O Pe en.
O.
N g
g O.
e M
ge N
N ge
.o as O
O O
O W
La./
m zOure.,-Nwo wzwso s w s w a.n w A
LiJ y
C J
=
"I W
es A
CC v
+
i. o.
e O
En*.
W W
Z J<
. O.
p O
b N
.J D
= nn w
O a
O C
V, M
DOE' 3
M
. O.
v O-O M
O *= C C tad m
u zo C,
v g
w O-e 4,
Q
. w e,
.e. a m e.
-w W
w i.e z a C
o=K
=
ECE o
g
- O.
-Eb-L 2
m mz-o_
g w
A
.. w W
= In KWd W
ll;.,y-sa mo O
i=
o
=*
=
C c
O e==s en bW W
3 O
. O.
- O sn O
en O
sn o
en o
sn C
mn o
en.
ce.
O.
P*
in.
N.
O.
f'*
44 N.
O.
f'=
e ee sw sw N
.=
O O
O O
E O s E g.J-aew O ted E l d e o s=
m ind.J t.# C W w l
I
.,1 e
j
= i
. 8
' >$ s$
O 4
. 2 4
b 6
3 5
C R
f i 0 E
E0-N3 0
W Il 3
W L 0 O
B EE 0 C
TLM 1 E
NMM P
E IMI 4
OIW l
L 4 E PMT E
L
- n I
2 l
N
. Ru i
A:E G hEL U
I tRG F
EAN F
RUA Og 8
- Sg T
1 R
I A
O T
S H
S P
. 2 K
4 AM lO N
5 E
0 B
D 0;
3 0
8
, i I
')<
- 0 1
0 1
E 4
S t
4 YR N0A ALI E0 sP0 P0M E
O R
T U
S G
I I
H F
R 0
E e
, 5 W
O P
T
,5 4
O P
S
, 0 4
TOH
, 5 3
6 R
C 0
f 0
E f
3 E0-W l
N3 C
II E M L 8 O
, 5 6
EE l
TLK 8 P
1 2 E NNN M
lNI 4
l oIW O
I rWI E
R
, 0 R
I N :
U 3
A:E G E
REL I
IRG F
Z ERN T
, 5 RuA 0I n
- 5R R
T A
O T
H
, 0 E
8
, 5 0
l II 0
- 0 0
0 0
0 0
1 0
0 0
0 8
0 0
0 0
l 0
0 0
0 8
0 6
8 N0RnAL8 hoi SF0I 0fR F
ll (fll llil\\l I'
,l l!llll l
l-llllll
FIGURE 4-19 HOT SPOT FUEL CE!!TERLI!1E TE!PEPATURE TRMISIENTS D BE"OPtARKS l
HOT ZERO POWER HOT FULL POWER 4500-4800-4700-C 4600-g3500-T C
E E 4500-R N
I L
A 4400-N I
i N
n 2500-d P
I E
E R
M 4200-A P
i E
R A 4100-5 0
R E
E 4000-0 8500-O O 3900-1000-I
~
500 0.0 05 150 85 2.0 25 30 35 4.0 4.5 50
- g sine isECi 0.0 05 l.0 95 20 2.5 30 3.5 4.0 4.5 5.0 "M
O!
01 i R$! C Miltonil s t AR : FACTRAN CODE 18-0 POWER HISTORTI SOUARE:
TRIANGLES FACIRAN CODE 13-0 PCWER MI510RTI SOUARE: REIRAN Mot EF01 (PolNT NINffiC5 P0mts MfET0Ari islANGLE FACIRAN CODE 13-0 POWER MIEIORil
~
FIGURE 4-l90
O
'd PAGE 156
~
E O
==
O 4
E-
- E
>EO E ina >
..g O2W
Q
M 8bE M
EWE w tad E-2
. O.
haie > O O
g 2Wh OE h== O S Ee e W
>)
O. n ~,
g
~
M Q
. w c
g OO l=s 5
D who E
.O :
8-" 8 5:
A "3" C
=
Q EAR b
- W g
cn
- p. e-g g=-
O i
E
- 1"d m
O E
a= tas E GO E **
E m
M 7
g EE W
8
.-e m._
,..,. o d,
O O
O O
O O
O O
O O
O O
O O
O O
g O
O O
O O
O O
O O
O O
O O
O O
O O
e e
Po e.s.
d 9
M M.
.
O M
ge em me se Q
m W
Po e
tds m
==
==
N g
us OD>ws udEO wwEbwmE=35W OwO,
W C
D W
>==
'b W
CL CW
>=
. O c
aC
-J y
. w 9
c-W 3
H
. O.
Q g
E*--
- E B* E O
0 E tas e
LJ g
O2W Q
Q
- e. g n
en L E EWE H
ww O
O E=2
~*
Q-4 M
87 y
h=O E
ene E*
O eSI c
E Z
q,,
==== tna CO W
.c.
In* h O E N
8-" 8
_O u-E CE te.
ELE O
E te e=
. g E u e= e= E uom E E..
m E led E'* E CEJ
. O.
.e S e= 3
- ens E em E==
E
- e=
. m.
lae O
E E
d 8
~'~
- 9 O
e i
i i
s i
1" s
O=--E osEO
.EE_3E.
O,e, Q,
h
+*k
$gy
&n,
- k'j zzz, IMAGE EVALUATION
" @[f/
/$
q, #
i TEST TARGET (MT-3) j 4
t 4)y+
1.0 lf 2 E m su p::ms 2
u g
!!!N M
j,l p1 1.25 1.4 jg 150mm 6"
g'4* 4%
+$[>A,,
'tMih*[d
///77 p,
++
(MS!
(([/p%6,q'[@ g,
- o hk 2
IMAGE EVALUATION k/77p \\[ g/
4ff TEST TARGET (MT-3)
%'d $'
"% 4
+
1'0
'N E
$j2 9 El h
IBM i,i l.8 1.25 1.4 1.6
=
150mm 6"
4k,r?*,4 f[.4,., &
o cp "t-
l pAGE 159 l
SECTION 5 -
SUMMARY
AND CONCLUSIONS The Virginia Electric and power Company (Vepco) has developed a
mathodology using the RETRAM transient thermal hydraulics code to analy=e the control rod ejection event for Vepco's Surry and North Anna Nuclear power Stations.
The analysis is performed in two steps: (1) a point kinetics calculation to determine the core average power history, and (2) a thermal hydraulics analysis, based on the core average power history from step (1), to p'redict the fuel and cladding temperatures and average fuel enthalpy history of the core hot spot. Results from step (2) are used to confirm that safe plant operation acceptance criteria are met. These criteria are those in USNRC Reg. Guide 1.77 (Ref. 3) and the additional in-house limits discussed in Section 1.3.
Acceptability of the methodology has been demonstrated by comparison of selected analytical resul~ts obtained with the Vepco methodology to corresponding results obtained with a NRC approved vendor methodology.
The conservatism of the Vepco methodology, specifically the point kinetics calculation and its use of a
Doppler reactivity feedback ueighting factor with the very conservative hot spot static peaking, was demonstrated through the comparison of the Vepco results with those based on a
three-dimensional space-time kinetics model and a detailed hot spot thermal hydraulic model.
In Section 3,
a comprehensive sensitivity analysis was performed and reported for neutronic and thermal hydraulic parameters to verify the range of applicability. The overall good agreement between the two methodologies affirms that the Vepco methodology can be used for performing the reload safety analysis
PAGE-160 1
for the control rod ajection analysis for Vepco's nuclear power plants.
l
(
i f
6 o
um w-._
PAGE 161 SECTION 6 - REFERENCES 1.
Surry Pouer Station Units 1 and 2,
" Updated Final Safety Analysis Report," Virginia Electric and Power Company, 1982.
2.
North Anna Power Station Units 1 and 2,
" Updated Final Safety Analysis Report," Virginia Electric and Power Company, 1982.
3.
" Assumptions Used for Evaluating a Control Rod Ejection Accident for Pressurized Water Reactors," Regulatory Guide 1.77, USAEC, May 1974.
4.
J.
H.
McFadden et al.,
"RETRAN-02:
A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flou Systems," EPRI MP-1850. April 1981 (Vol. 1-3) and January 1983 (Vol. 4).
5.
"WREM:
Water Reactor Evaluation Model," NUREG-75/056, Revision 1,
May 1975.
6.
H.
A. Smith, " Reactor System Transient Analyses Using the RETRAN Computer Code," VEP-FRD-41, Virginit Electric and Power Company, March 1981.
7.
J.
H.
Keenan, et al.,
" Steam Tables:
Thermodynamic Properties of Water Including Vapor, Liquid, and Solid Phases," John Wiley & Sons, Inc.,
1969.
8.
"MATPRO - Version 11 (Revision 1), A Handbook of Materials Properties for Use in the Analysis of Light Water Reactor Fuel Rod Behavior," NUREG/CR-0497, TREE-1280, Rev 1.
USNRC, Feb. 1980.
9.
C. Hunin, "FACTRAN - A FORTRAN IV Code for Thermal Transients in a UO2 Fuel Rod," WCAP-7337, June 1972.
10.
L.
S.
Tong and J.
Weisman, " Thermal Analysis of Pressuri=ed Water Reactors," American Nuclear Society, 1970.
11.
M.
L. Smith, "The PD207 Discrete Model," VEP-FRD-19A, Virginia Electric and Power Company, July 1981.
12.
J.
R. Rodes, "The PD207 One Zone Model," VEP-FRD-20A, Virginia Electric and Power company, July 1981.
13.
W.
C. Beck, "The Vepco FLAME Model," VEP-FRD-24A, Virginia Electric and Power Company, July 1981.
14.
J.
G. Miller, " Nuclear Design Reliability Factors," VEP-FRD-45A, Virginia Electric and Power Company, October 1982.
15.
S.
A.
Ahmed el al., " Reload Nuclear Design Methodology," VEP-TRD-42, Virginia Electric and Power Company, April 1981.
l l
PRGE 162 16.
Roger A.
Rydin, " Nuclear Reactor Theory and Design," University Publications, 1977.
17.
D.
H.
Risher, "An Evaluation of the Rod Ejection Accident in Westinghouse Pressuri=ed Water Reactors Using Spatial Kinetics Methods," WCAP-7588, Revision 1-A, January 1975.
18.
R.
F.
Barry and D.
H.
Risher, " TWINKLE -A Multi-Dimensional Neutron Kinetics Computer Code." WCAP-7979, December 1972.
19.
Letter from C.
M.
Stallings, Vepco, to E.
G.
Case, NRC, Serial No. 108, Docket Hos. 50-280 and 50-281, March 15,1978.
20.
J.
B.
Yasinsky, "On the Use of Point Kinetics for the Analysis of Rod-Ejection Accidents," Nuclear Science and Engineering; 39, 241-256, 1970.
29.
D.
B.
Waters, "The Core Physics Characteristics of the Surry Unit 1 Nuclear Power Station," WCAP-7534, Rev.
1, July 1970.
22.
Letter from C.
M.
Stallings, Vepco, to K.
R.
Goller, NRC, Serial No. 553, June 5,1975 (Surry Positive Moderator Coefficient).
_ _ _ _ _ _.