Citation
Optimum Boiling Water Reactor Fuel Design Strategies to Enhance Reactor Shutdown by the Standby Liquid Control System

Material Information

Title:
Optimum Boiling Water Reactor Fuel Design Strategies to Enhance Reactor Shutdown by the Standby Liquid Control System
Creator:
FENSIN, MICHAEL LORNE ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Boron ( jstor )
Conceptual lattices ( jstor )
Eigenvalues ( jstor )
Exhibit cases ( jstor )
Gadolinium ( jstor )
Geometry ( jstor )
Reactivity ( jstor )
Reactor cores ( jstor )
Reactor design ( jstor )
Trucks ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Michael Lorne Fensin. Permission granted to University of Florida to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/7/2004

Downloads

This item has the following downloads:

fensin_m ( .pdf )

fensin_m_Page_100.txt

fensin_m_Page_062.txt

fensin_m_Page_093.txt

fensin_m_Page_020.txt

fensin_m_Page_069.txt

fensin_m_Page_054.txt

fensin_m_Page_063.txt

fensin_m_Page_036.txt

fensin_m_Page_083.txt

fensin_m_Page_085.txt

fensin_m_Page_057.txt

fensin_m_Page_084.txt

fensin_m_Page_031.txt

fensin_m_Page_097.txt

fensin_m_Page_065.txt

fensin_m_Page_064.txt

fensin_m_Page_046.txt

fensin_m_Page_038.txt

fensin_m_Page_060.txt

fensin_m_Page_030.txt

fensin_m_Page_012.txt

fensin_m_Page_094.txt

fensin_m_Page_068.txt

fensin_m_Page_028.txt

fensin_m_Page_053.txt

fensin_m_Page_011.txt

fensin_m_Page_079.txt

fensin_m_Page_073.txt

fensin_m_Page_075.txt

fensin_m_Page_078.txt

fensin_m_Page_008.txt

fensin_m_Page_042.txt

fensin_m_Page_033.txt

fensin_m_Page_024.txt

fensin_m_Page_091.txt

fensin_m_Page_102.txt

fensin_m_Page_040.txt

fensin_m_Page_007.txt

fensin_m_Page_019.txt

fensin_m_Page_051.txt

fensin_m_Page_039.txt

fensin_m_Page_009.txt

fensin_m_Page_034.txt

fensin_m_Page_059.txt

fensin_m_Page_015.txt

fensin_m_Page_004.txt

fensin_m_Page_048.txt

fensin_m_Page_041.txt

fensin_m_Page_047.txt

fensin_m_Page_052.txt

fensin_m_Page_021.txt

fensin_m_Page_014.txt

fensin_m_Page_098.txt

fensin_m_Page_101.txt

fensin_m_Page_027.txt

fensin_m_Page_080.txt

fensin_m_Page_049.txt

fensin_m_Page_006.txt

fensin_m_Page_058.txt

fensin_m_Page_013.txt

fensin_m_Page_056.txt

fensin_m_Page_035.txt

fensin_m_Page_071.txt

fensin_m_Page_088.txt

fensin_m_Page_037.txt

fensin_m_Page_087.txt

fensin_m_Page_066.txt

fensin_m_Page_016.txt

fensin_m_Page_081.txt

fensin_m_Page_077.txt

fensin_m_Page_044.txt

fensin_m_Page_082.txt

fensin_m_Page_061.txt

fensin_m_Page_023.txt

fensin_m_Page_076.txt

fensin_m_Page_032.txt

fensin_m_Page_099.txt

fensin_m_Page_067.txt

fensin_m_Page_096.txt

fensin_m_Page_074.txt

fensin_m_Page_070.txt

fensin_m_Page_089.txt

fensin_m_pdf.txt

fensin_m_Page_095.txt

fensin_m_Page_092.txt

fensin_m_Page_072.txt

fensin_m_Page_090.txt

fensin_m_Page_050.txt

fensin_m_Page_005.txt

fensin_m_Page_018.txt

fensin_m_Page_086.txt

fensin_m_Page_026.txt

fensin_m_Page_055.txt

fensin_m_Page_022.txt

fensin_m_Page_029.txt

fensin_m_Page_017.txt

fensin_m_Page_003.txt

fensin_m_Page_010.txt

fensin_m_Page_045.txt

fensin_m_Page_025.txt

fensin_m_Page_043.txt

fensin_m_Page_002.txt

fensin_m_Page_001.txt


Full Text












OPTIMUM BOILING WATER REACTOR FUEL DESIGN STRATEGIES TO
ENHANCE REACTOR SHUTDOWN BY THE STANDBY LIQUID CONTROL
SYSTEM















By

MICHAEL LORNE FENSIN


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2004

































Copyright 2004

by

Michael Lorne Fensin


































This document is dedicated to the memory of my late grandmothers Bernice Anker and
Edna Fensin.















ACKNOWLEDGMENTS

I would like to acknowledge Dr. Samim Anghaie for chairing my committee,

supplying a connection to Global Nuclear Fuels of America, and providing excellent

tutoring and advice as my graduate advisor. I would also like to thank Dr. Bob Coldwell

Dr. Edward Dugan, Dr. Alireza Haghighat, Dr. David Hintenlang, Dr. Travis Knight, Dr.

Alan Jacobs, Dr. Tim Olson, Dr. Benard Mair, Proffessor Jim Tulenko and Dr. William

Vernetson for providing me countless hours of instruction in all areas of nuclear

engineering and mathematical computation during my graduate studies.

I would like to acknowledge Global Nuclear Fuels of America for the sponsorship

of its computer codes, time and efforts. From Global Nuclear Fuels of America I would

specifically like to thank Dr. Mehdi Asgari, Kenneth Gardner, Roland Jackson, J.D.

Kavaal, Thomas Marcille, V.W. Mills, Dr. Brian Moore and Tony Reese for supplying

intriguing knowledge and guidance during the course of the study.

I want to thank my family for being a constant source of support and pushing me to

completion. Without their support none of this would have been possible.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IS T O F T A B L E S ................. ........................................................................ ... v ii

LIST OF FIGURES ...................................... ........ .......... ............ .. viii

ABSTRACT ........ .............. ............. ...... ...................... xi

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

The B oiling W ater R eactor System ........................................ ......................... 4
Boiling W ater Reactor History .................................. .....................................8
History of Fuel Bundle Developm ent..................................................................... 10
T he SL C S E vent ............................................................................................ ....... 12
P roje ct S c o p e .....................................................................................13

2 M ODEL AND M ETHODOLOGIES ........................................ ...... ............... 15

Standby Liquid Control System and Shutdown Margin...........................................15
M o d e lin g T o o ls ..................................................................................................... 1 7
T G B L A 6 ............................................................................................................. 1 7
P A N A C 11 ................................................................ ............................18
Utilized Temperature States, Boron Concentrations and Lattice Types...................19
Measurement of SLCS and SDM during the Lattice Development Stage .................20
Fuel Bundle G eom etry .................. ................................. .... ........21
Thermal Limit Design Considerations..................... ..... ......................... 24

3 MAXIMIZING HOT-COLD BORATED k- DIFFERENCE UTILIZING
ENRICHM ENT ................................................ ...................... ........ 28

Hom ogeneous Enrichm ent Distribution ....................... .............. ............... ....28
Determining the Most Limiting Lattice Axial Zone and Void Concentration ....29
Understanding the Exposure Dependent HUCU### Curve..............................31
Enrichment and Boron Concentration Effects.........................................34
Pow er Peaking D istribution........................................... .......................... 35









H heterogeneous Enrichm ent D distribution ........................................ .....................39
Localized Enrichment Perturbation .............. ............................................. 39
G ross Enrichm ent Perturbation ........................................ ........ ............... 41

4 MAXIMIZING HOT-COLD BORATED ko, DIFFERENCE UTILIZING
GADOLINIUM ................................... ..... .. ...... .............. 44

Spatial Self-Shielding Effects of Gadolinium Rods on HUCU###............................44
The Effects of Increasing the Amount of Gadolinium Rods on HUCU### ..............47
The Effects of Increasing the Gadolinium Concentration on HUCU### .................50
The Importance of Gadolinium Rod Location............................................... 51
Fuel Lattice D esign Conclusions ..................................................... ..... .......... 55

5 FULL CORE SLCS M ODELING ....................................... ........................... 57

Enhancing SLCS by Perturbing the Location of Gadolinium Rods...........................59
Axial Power Shape Characteristics................... .............................. 65
Enhancing SLCS through Axial Power Shaping Utilizing Enrichment
D ifferencing ............... .... ............ .................................... ....... 66
Enhancing SLCS by Means of Axial Power Shaping Utilizing Gadolinium
In se rtio n ......................................................................... 7 4

6 C O N C L U SIO N S ....................... .... .......................... ................ ...... ......... 83

7 FUTURE W ORK.......................... ........... .. ........... ... ...... 86

L IST O F R E FE R E N C E S ......................................................................... ....................88

BIO GRAPH ICAL SK ETCH .................................................. ............................... 90















LIST OF TABLES


Table page

3-1 Gross enrichment perturbation scheme for figure 3-10......................................... 41

5-1 Critical eigenvalue at specified exposure points for the gadolinium rod location
perturbation cases that exhibited greatest enhancement in SLCS..........................62

5-2 Exposure dependent MFLPD for the gadolinium rod location perturbation cases
that exhibited greatest enhancement in SLCS...................... ................................. 63

5-3 Exposure dependent MFLCPR for the gadolinium rod location perturbation cases
that exhibited greatest enhancement in SLCS...................... ................................. 64

5-4 Critical eigenvalue at specified exposure points for the gadolinium rod location
perturbation case and enrichment differencing case that exhibited greatest
enhance ent in SL C S. ........................................ ........................ 71

5-5 Exposure dependent MFLPD for the gadolinium rod location perturbation case and
the enrichment differencing case that exhibited greatest enhancement in SLCS. ...72

5-6 Exposure dependent MFLCPR for the gadolinium rod location perturbation cases
that exhibited greatest enhancement in SLCS...................... ................................. 73

5-7 Critical eigenvalue at specified exposure points for the gadolinium insertion into
the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited
greatest enhancem ent in SLC S.......................................... ........................... 76

5-8 MFLPD at specified exposure points for the gadolinium insertion into the DOM
(case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest
enhance ent in SLC S. ........................................ ........................ 79

5-9 MFLCPR at specified exposure points for the gadolinium insertion into the DOM
(case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest
enhance ent in SLC S. ........................................ ........................ 80
















LIST OF FIGURES


Figure p

1-1 B W R pressure vessel system ......................................................................... ...... 5

1-2 A typical BW R fuel assem bly and fuel rod..................................... .....................6

1-3 A four fuel assembly group with cruciform control blade.............. ... .................7

2-1 A cross sectional view of the modeled fuel bundle.................... ... ............... 23

2-2 The geometric setup of the fuel lattice axial zones. ...............................................24

3-1 Exposure dependent HUCU660 for the DOM at 3.95% enrichment....................... 29

3-2 Exposure dependent HUCU660 at varied void fraction and axial zone for the C
lattice at an enrichm ent of 3.95% ................................. ........................ ......... 31

3-3 Exposure dependant HUCU### curve. ....................................... ............... 33

3-4 Beginning of cycled HUCU vs. enrichment in the DOM, at 40% void fraction,
for a C lattice. ..........................................................................34

3-5 The power peaking distributions at 5 GWD/STU, 3.95% enrichment, DOM, C
lattice, and 40% void fraction vs. temperature state and boron concentration .........35

3-6 The power peaking distribution at 5 GWD/STU, CU660, DOM, C lattice, and
40% void fraction versus enrichment................ ............................. ............... 38

3-7 The power peaking distributions at 5 GWD/STU, HU, DOM, C lattice, and 40%
void fraction vs. enrichm ent .................................. ............... ............... 38

3-8 Localized enrichment perturbation map............................................ .............40

3-9 Exposure dependent HUCU660 for different localized enrichment perturbation
pattern s. .............................................................................40

3-10 An example of a gross enrichment perturbation map..........................................41

3-11 Exposure dependent HUCU660 at 40% void fraction, in the DOM, with a C
lattice. .....................................................................................4 2









3-12 Exposure dependent HUCC at 40% void fraction, in the DOM, with a C lattice....42

4-1 Four sam ple clum ped geom etries....................................... .......................... 45

4-2 Corresponding 0 GWD/STU gadolinium worth for the patterns displayed in
figure 4-1. .............................................................................46

4-3 Corresponding exposure dependent gadolinium clumping effects on HUCU660
for the patterns displayed in figure 4-1. ...................................... ............... 46

4-4 The effects of increased number of gadolinium rods on the gadolinium worth at
0 G W D /STU ..............................................................................................48

4-5 The effects of the number of gadolinium rods inserted on HUCU at
0 G W D /STU ..............................................................................................49

4-6 The effect of increasing the gadolinium concentration for 14 gadolinium rods on
gadolinium worth at 0 GW D/STU. ............................................... ............... 50

4-7 The effect of increasing the gadolinium concentration of 14 gadolinium rods on
H U CU at 0 G W D /STU ................................................. ............................... 51

4-8 Gadolinium rod place ent diagram .............................................. .....................52

4-9 Gadolinium worth versus location for 0 GWD/STU, 7% gadolinium
concentration, ........................................................................54

5-1 Reference base core fuel bundle loading map .......................................................58

5-2 The perturbation diagram for the gadolinium rod perturbation cases....................60

5-3 Exposure dependent SLCS for the reference base case, the case in which the
perturbation was made to only all of the fresh low enrichment bundles (case 2),
and the case in which the perturbation was made to only all of the fresh high
enrichm ent bundles (case 1). ........................................ ......................................61

5-4 Exposure dependent SDM for the gadolinium rod location perturbation cases.......65

5-5 The base case hot axial power shape with superimposed cold axial power shape...66

5-6 Cold axial power shape perturbation diagram ......................................................67

5-7 Exposure dependent SLCS for the gadolinium location perturbation case
(case 2) and axial power shape perturbation utilizing enrichment differencing
case (case 11) that exhibited the greatest enhancement in SLCS.............................68

5-8 SLCS enhancement utilizing different magnitudes of enrichment differencing
betw een the D OM and VAN .................................. ............... ............... 69









5-9 The hot axial power shape for maximum SLCS enhancement utilizing
enrichm ent differencing. ........................................ ............................................70

5-10 Exposure dependent SDM for the gadolinium rod location perturbation case and
axial power shaping utilizing enrichment differencing case..................................74

5-11 Exposure dependent SLCS for the gadolinium location perturbation case
(case 2), axial power shape perturbation utilizing enrichment differencing case
(case 11), inserting a gadolinium rod in the PSZ (case 26) and inserting a
gadolinium rod in DOM (case27) that exhibited the greatest enhancement in
SL C S ...................................................................................... 7 5

5-12 The hot axial power shape of the most power peaked fuel bundle caused by
inserting a gadolinium rod into the PSZ........................................ ............... 77

5-13 The hot axial power shape of the most power peaked fuel bundle caused by
inserting a gadolinium rod into the DOM .................................... .................78

5-14 Exposure dependent SDM for the gadolinium rod location perturbation case,
axial power shaping utilizing enrichment differencing case and the axial power
shaping utilizing gadolinium place ent case................................ ............... 81















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

OPTIMUM BOILING WATER REACTOR FUEL DESIGN STRATEGIES TO
ENHANCE REACTOR SHUTDOWN BY THE STANDBY LIQUID CONTROL
SYSTEM

By

Michael Lorne Fensin

August 2004

Chair: Samim Anghaie
Major Department: Nuclear and Radiological Engineering

Licensing a commercial nuclear reactor core involves a stringent amount of

calculations that demonstrate the capability of safe reactor shutdown in the instance of an

emergency transient event. In a boiling water nuclear reactor the control blades and

standby liquid control system are the two independent redundant safety systems utilized

for shutting down the reactor. In past fuel designs lower power cores with smaller cycle

lengths resulting from lower core average enrichment caused shutdown by the control

blades to be the most limiting strategy of the two modes for reactor shutdown. This led

to a lesser focus on fuel design strategies for standby liquid control system margin

(SLCS). Advanced modem core designs involve higher powers and increased cycle

lengths resulting from higher core average enrichments, therefore causing SLCS to now

become a more significant shutdown parameter. This study characterized the most

limiting fuel design parameters for maximizing the margin for safe reactor shutdown

utilizing SLCS while maintaining the demanded cycle energy requirements.









This study examined perturbation effects of certain parameters in the fuel lattice

development stage and the 3-dimesional reactor core simulator modeling. Lattice

enrichment perturbation response was examined first to determine the effects of average

and local enrichment perturbations on SLCS. Lattice gadolinium perturbation response

was next investigated to determine the optimum concentration, number of rods, location

and degree of clump of gadolinium rods necessary for enhancing SLCS. Axial power

shape perturbation utilizing both gadolinium and/or enrichment differencing in certain

axial zones of the fuel bundle was then examined on the full core level to determine the

optimum strategy for maximizing SLCS.

This study concluded that the necessary strategy for maximizing SLCS depended

upon the exposure point at which SLCS was most limiting. Certain perturbations

utilizing gadolinium exhibited maximized beginning of cycle SLCS; however, these

strategies involved a modified operating strategy to meet the beginning of cycle

operational requirements while not maximizing the limiting end of cycle SLCS. Axial

power shaping utilizing enrichment differencing maximized end of cycle SLCS and

increased beginning of cycle SLCS margin but to a lesser magnitude than the gadolinium

perturbation. In all cases the amount of improvement to the margin was limited by a

maximum value. Therefore if the desired magnitude of improvement needed is within

the achievable limits of the examined techniques, the choice of optimum strategy for

enhancing SLCS to a desired value depends upon the magnitude of necessary

improvement at the most limiting exposure points.














CHAPTER 1
INTRODUCTION

Boiling water nuclear reactor cores are major sources of revenue for power

producing utilities. If the utility is able to maximize the amount of energy output from the

nuclear reactor while minimizing the cost of the reactor operation then the utility will

realize an increase in profit. A utility may choose to maximize the energy output from

their nuclear reactors in one of three ways. Either the utility may increase the operating

cycle length of the reactor thereby increasing the amount of energy per cycle and

increasing the amount of time between refueling outages, or the utility may choose to

increase the power level of operation thereby increasing the amount of available

distributable energy at any given time, or the utility may choose to utilize a combination

of both practices [2]. In either of the operational techniques the utility must increase the

installed reactivity in the reactor core in order to meet the desired goal.

Increasing the amount of installed reactivity in a given cycle as compared with a

previous cycle in order to aggressively improve reactor power output is termed

"aggressive core loading." Aggressive core loading strategies involve higher fuel batch

fractions with fuel bundles of higher average enrichment in order to increase the installed

positive reactivity in the reactor core [2]. Greatly increasing the core power output in the

internal core locations greatly increases the neutron flux and thus greatly increases the

exposures of the interior bundles. A twice-burned fuel bundle in the reactor core is at its

least reactive state in bundle life and therefore in ordinary core loadings twice-burned

fuel bundles can be an excellent power suppressor utilized to flatten the power









distribution in the internal core locations; however, in aggressive core loadings the high

neutron flux in the interior of the core may causes twice-burned fuel bundles to exceed

thermo-mechanically limited peak exposure. Therefore in aggressive core loadings after

bundles have burned two cycles they must be moved to the core periphery in order to

inhibit surpassing peak exposure [4]. This leads to only fresh and once burned

assemblies loaded in the interior core locations.

Most of the gadolinium in a once burned fuel bundle is completely burnt up at the

end of the previous cycle therefore these bundles are at the most reactive state. Inability

to suppress this energy output decreases the available margin to shutdown the core in an

emergency situation. Due to the aggressive core loading geometry, the only available

power suppression comes from the fresh fuel bundles that are loaded into the core.

However, with increased energy placed in the fresh bundles by increasing average

enrichment in order to meet the increased power demand, these fresh bundles will have

decreased power suppression capabilities. Therefore with decreased power suppression

capabilities, the reactor core becomes more limiting in emergency shutdown capabilities.

One of the shutdown systems is the standby liquid control system and the margin in

which the reactor core is shutdown utilizing this system is the standby liquid control

system margin (SLCS). If the reactor core becomes more limiting in SLCS due to the

decreased power suppression capabilities of the core loading strategy, it becomes

paramount to then determine the optimum bundle design utilized in order to improve

SLCS in an aggressive core loading environment.

A reactor core is never licensed without being able to meet all the necessary

shutdown criteria, thermal limit characteristics, and cycle length requirements. Therefore









operating reactor cores do not encounter a failure to meet SLCS because the reactor

would not be allowed to operate if the reactor could not meet the SLCS requirements.

Inability to meet SLCS with a certain reactor core fuel bundle configuration is realized

and mitigated in the design phase of the reactor fuel cycle.

The designer has many options to improve SLCS but any one option may endure a

list of consequences some of which may result in extreme economic concern. The

designer may request that the reactor cycle length be decreased; obviously if the utility

wishes to increase profit by increasing power output this option is not acceptable. The

designer may request the utility to increase the boron concentration or enrichment of B10

in the boron solution utilized by the standby liquid control system. However, this course

of action is limited by increased aggravation caused from Nuclear Regulatory

Commission (NRC) licensing, availability for the utility to plan for the change

economically, ample time to complete the concentration increase in time for the next

cycle loading, capabilities of the installed accumulator tank to support the concentration

increase and ability to keep the boron soluble in solution. The designer may choose to

increase the amount of fuel bundles loaded in the cycle and load more fuel bundles with a

smaller enrichment; however, this may cost the utility more than what was budgeted and

therefore is not a viable option.

The action that is demanded by the utility is for the designer to create a core design

that meets the utilities budget and does not increase the amount of bundles that are loaded

into the reactor core. Therefore the designer must design a fuel bundle with inherently

better SLCS characteristics. In order to accomplish this task efficiently the designer must









know the set of limiting design parameters that may be utilized to enhance SLCS, and

have a list of effective techniques that utilize the advantages of those parameters.

The purpose of this study was to determine the fuel bundle design parameters that

were most limiting in achieving the maximum possible enhancement for SLCS, and to

determine the maximum amount of available improvement to SLCS by utilizing certain

enhancement techniques that take advantage of those design parameters. This will

demonstrate the feasibility of designing a fuel bundle and core operating strategy that has

the ability to meet SLCS without incurring the costs of adding extra fuel bundles in the

design or decreasing cycle power output requirements.

The Boiling Water Reactor System

The boiling water reactor (BWR) system is a nuclear system that boils water

creating steam that is converted into power. The entire BWR system is composed of a

reactor pressure vessel system, a turbine system, a generator system, a condenser system

and the auxiliary control and heat removal systems. Steam is created by boiling water in

the reactor pressure vessel system. The high quality steam then passes through a turbine,

and causes the turbine shaft to rotate. The turbine shaft is connected to a generator and as

the turbine shaft rotates the generator converts the mechanical energy of the rotating

turbine shaft into electrical energy. Once the steam leaves the turbine system, it is sent to

the condenser to be condensed into a sub-cooled fluid and pumped back into the reactor

pressure vessel system.

The reactor pressure vessel system is of primary concern to the nuclear reactor core

designer. Figure 1-1 is an illustration of a typical BWR reactor pressure vessel system.

The reactor pressure vessel system consists of the control drive mechanisms, the active








5



fuel, the jet pumps, the moisture separators and steam driers as well as other inlets for


emergency core cooling [16].




STEAM DRYER LIFTING LUG

VENT AND HEAD SPRAY





S-+ STEAM DRYER
C ASSEMBLY

STEAM OUTLET A S B


STEAM SEPARATOR
ASSEMBLY



FEEDWATER INLET
CORE SPRAY INLET
i j FEEDWATER SPARGER


LOW-PRESSURE COOLANT I .
INJECTION INLET CORE SPRAY LINE

CORE SPRAY SPARGER --
TOP GUIDE


JET PUMP ASSEMBLY CORE SHROUD


S1 CONTROL BLADE
FUEL ASSEMBLIES


CORE PLATE
JET PUMP/IECICULATION RECCULA
WATER OUTLET

CONTROL ROD GUIDE TUBE i


-.. I SHIELD WALL


CONTROL ROD DRIVE
HYDRAULIC LINES


IN.COE FLUX MONITOR



Figure 1-1. BWR pressure vessel system [16].











The active reactor fuel length is approximately 12 feet though actual fuel length


may vary according to different types of fuel assembly product lines. The fission power


of the reactor converts the sub-cooled coolant into steam. The steam then travels through


the steam separators and driers to create a high quality steam that is then send to the


turbine. Each fuel assembly is composed of the fuel and water rods, intermediate spacer


grids, upper and lower tie plates, and flow nozzle. Figure 1-2 displays a typical BWR


fuel assembly as well as a typical fuel rod. Each fuel assembly is encased in a zircaloy


fuel box in order to limit flow between adjacent fuel assemblies. This allows the flow to


any given fuel assembly to be orificed to maintain a constant exit steam quality as well as


limit instability in core thermal performance [16].














TIgE PLTE
FUEL CbLAHNG
BUNDLE

----- ~ ^^'(P1NS QN


F', ,, *.: ----- --.
FU*I "AN-1L
PLENUM




TYPICAL OP 4)







Figure 1-2. A typical BWR fuel assembly and fuel rod [16].






Fuel assemblies are loaded into the reactor in groups of four with a cruciform B4C
control blade loaded in the center of the grouped bundles. Displayed in figure 1-3 is a
typical layout of four fuel bundles with a cruciform control rod loaded in the center of the
grouped bundles. The fuel assemblies are diagonally symmetric with themselves and are
loaded so that there exists 1/8 bundle symmetry with the four grouped bundles.


o000o0000 00000000
00000000 00000000
00000000 000.0000
ooooooo oooooooo
OOOOOOO OOOOOOOO
0oo000000 o0000000o


OOOOOO'O OOOOQOO
00000000 00000000
00000000 00000000


00000000 OOOOOOO
OOOOOOO 00000000
0000000O 0000000
00000000 00000000
00000000 00000000
00000000 00000000


FOUR-BUNDLE FUEL MODULE
O FUEL ROD
O WATER RODS
0 TIE RODS


Figure 1-3. A four fuel assembly group with cruciform control blade [16].









By applying this loading scheme, symmetry may be utilized in modeling the fuel

bundle thereby easing the computation time of the fuel assembly parameters [14]. Each

BWR fuel assembly will contain fuel rods at certain enrichments that may vary radially

and axially as well as gadolinium rods which may vary in placement and concentration

radially and axially. The modeling of SLCS will first involve modeling a 2-dimensional

slice of a single fuel assembly at certain axial heights and then modeling the entire 3-

dimensional reactor utilizing the information from the 2-dimensional model.

Boiling Water Reactor History

The first two light water cooled nuclear reactor systems commercially available for

power production were the boiling water reactor and the pressurized water reactor

(PWR). The concept of the commercial PWR was created from the technology

developed for submarines by the navy nuclear program [7]. BWR development occurred

at Argonne National Laboratory and the Nuclear Energy Division of General Electric

(GE) [9]. The PWR concept is generally characterized as a system in which the bulk

coolant is sub-cooled and contains boron. The system utilizes an indirect dual-cycle that

uses a steam generator. The BWR concept is generally characterized as a system that has

boiling in the reactor core, with the bulk fluid containing no boron, utilizing a direct cycle

for power conversion (the demonstration BWR/1 plants utilized a dual cycle).

The first BWR experiments conducted at Argonne National Laboratory utilized the

BORAX in 1953. BORAX-III produced steam-generated electricity for the town of

ARCO, ID in 1955. The experimental boiling water reactor (EBWR) was developed in

1957 and ran until 1967. This reactor produced 100 MWt and was utilized to

demonstrate the BWR concept for electricity generation utilizing a variety of fuel

enrichments. The GE Valecitos boiling water reactor (VBWR) was the first commercial









nuclear power plant to be licensed by the United States Atomic Energy Commission

(USAEC). Utilized as an experimental reactor, VBWR examined BWR fuel cycle

technology and determined the stable modes of operation. In 1955 Dresden-1 became the

first commercial BWR specifically constructed for commercial power [12].

Dresden-1 was a dual cycle plant and fell under the category ofBWR/1. BWR/1

designs were basically prototype designs of both dual and direct cycle utilized as

demonstration plants that were custom made to meet individual utility specifications.

Dual cycle BWR plants eventually fell out of favor because of the enormous capital cost

involved in utilizing a steam generator. Ouster creak was the first attempt at

standardizing the BWR and marked the beginning of the BWR/2. In 1963 BWR/2 plants

were developed incorporating a direct steam cycle as the chosen method of power

conversion. The reactor concept utilized internal steam separators and forced flow

circulation that pumped core flow through 5 variable speed recirculation pumps. In 1965

GE introduced the BWR/3, the Dresden-2 design, which incorporated the use of internal

jet pumps eliminating the need for external flow circulation loops. In 1966 the BWR/4

or Browns Ferry design was introduced. This design was similar to previous designs but

incorporated a 20% increase in the core power density improving power producing

capability of the reactor and thereby increasing its economic value. The year of 1969

marked the introduction of the Zimmer class of plants better known as the BWR/5.

These plants utilized an improved emergency core cooling and recirculation system.

Flow control in these reactors was accomplished through use of valve control rather than

pump speed control allowing the plants to follow more rapid load change and decrease

the capital cost of the control system [12]. The BWR/6 incorporated the used of higher









efficiency steam separators and multi-hole jet pumps as well as improved power

flattening through enhanced coolant distribution and burnable Gadolinia zone loading.

The original BWR/6 fuel design incorporated an 8X8 fuel assembly rather than the

previously utilized 7X7 [12]. The current BWR technologies include the Advanced

Boiling Water Reactor (ABWR) and the European Simplified Boiling Water Reactor

(ESBWR) both of these concepts utilized passive safety features in order to alleviate the

need for complicated control systems. The history of the BWR is one that is based off of

evolutionary concepts in order to increase core power and eliminate external moving

parts that may fail during reactor operation.

History of Fuel Bundle Development

Over the years fuel bundle designs have undergone evolutionary changes in order

to improve the thermal performance of the fuel and the reactivity features. Each fuel

bundle design incorporated a key feature that allowed for increased margin in controlled

conditions, transients, and thermal limits. The original fuel bundle concept utilized by

the General Electric Company was the 7X7 fuel lattice design. These were fat fuel rods

with decreased surface area due to the large size of the fuel rods. The decreased surface

as compared to today's designs lead to an increase in the maximum linear heat generation

rate (kW/ft) in the fuel bundle leading to an increase in fuel duty [12].

With the creation of the BWR/6 came the inclusion of the 8X8 fuel lattice

assembly. Increasing the surface area of the fuel rods by decreasing there width and

increasing the amount of fuel rods in the assembly decreased the maximum linear heat

generation rate of the fuel rods. In 1988 a fuel design study by Motoo Aoyama, Sadao

Uchikawa and Renzon Takeda suggested that extended exposure of fuel assemblies was

possible if 9X9 fuel assemblies were utilized with and optimized internal water rod width









thereby increasing the non boiling area inside the fuel lattice to increase moderation

capability of the internal portions of the fuel lattice [1]. This design change increased

fuel lattice efficiency. Going to a larger amount of fuel rods in the lattice at smaller fuel

diameter decreased the linear heat generation of the fuel rod by increasing the surface

area of the fuel.

The current design utilized today by BWR vendors is the standard 10X10 fuel

lattice design. Though the water rod locations vary from vendor to vendor the idea is the

same; increase the moderation capabilities of the internal locations of the fuel lattice to

boost reactivity in the areas that are most suppressed in power.

Because the moderator density varies axially in the fuel assembly, the BWR has a

distinct axial power shape. Kazuki Hida and Ritsuo Yoshioka determined that there were

optimum axial enrichment distributions that minimized enrichment requirements subject

to thermal margins [9]. They proved that increasing enrichment in the top half of the

core actually decreases the uranium utilization. Therefore the interpretation of that study

determined that fuel utilization was a key design constraint in creating optimum fuel

bundles. In order to mainstream fuel designs industry moved away from axial

enrichment shaping and fabricated fuel rods of a single enrichment and utilized part

length control rods that increased the moderation capability in the top half of the fuel

assembly leading to an increase in fuel efficiency.

In 1997 Yasushi Hirano, Kazuki Hida, Koichi Sakurada and Munenari Yamamoto

created an algorithm for determining optimum enrichment loading schemes for fuel

lattices [10]. Holding the position and concentration of the gadolinium rods as the

constant, the algorithm optimized the enrichments in the fuel lattice in order to meet









certain thermal limit and local peaking factor criteria. The gadolinium configuration

utilized for the study was based off of a configuration that was supposedly optimized

only for shutdown margin (SDM) and certain thermal limit criteria based off of previous

fuel design experience. However, this method lacked the ability to place the gadolinium

and enrichment into the fuel lattice in such an optimum configuration such that all

controls were satisfied. If the chosen gadolinium configuration was only optimum for

SDM yet not also optimum for SLCS then this method was to only be successful in

designing a lattice to meet SDM. What this method was lacking was the rules for

understanding how the gadolinium needed to be configured to meet the SLCS, SDM and

thermal limit configuration for a given average enrichment.

The SLCS Event

The standby liquid control system is initiated during anticipated transient without

scram (ATWS). The following is a list of the events that occur in the SLCS event:

1. A transient even occurs in which it is necessary to SCRAM the reactor.

2. The reactor control blades fail to insert.

3. A calculated amount of steam is relieved from the reactor to the suppression pool at
a rate that will not violate containment.

4. The boron solution is injected into the core at a specified rate and concentration in
accordance with 10CFR50.62.

5. The reactor reaches an equilibrium shutdown condition.

In 1984 the Nuclear Regulatory Commission (NRC) issued 10CFR50.62,

"Requirements for reduction of risk from anticipated transients without scam events for

light water-cooled nuclear power plants" (ATWS rule). The law states:

Each boiling water reactor must have a standby liquid control system (SLCS) with
the capability of injecting into the reactor pressure vessel a borated water solution
at such a flow rate, level of boron concentration and boron-10 isotope enrichment,









and accounting for reactor pressure vessel volume, that the resulting reactivity
control is at least equivalent to that resulting from injection of 86 gallons per
minute of 13 weight percent sodium pentaborate decahydrate solution at the natural
boron-10 isotope abundance into a 251-inch inside diameter reactor pressure vessel
for a given core design. [17]

This is equivalent 660 ppm boron concentration in current reactor designs. The

model utilized for determining if SLCS will satisfy the criteria mentioned in 10CFR50.62

is a steady state point after the transient event has occurred. The purpose of utilizing this

method is to demonstrate that the reactor may be safely shutdown after the transient event

has occurred. Therefore the event modeled is a cold core (160C), borated to an

acceptable concentration that causes the reactor to be sub-critical by a specified amount.

Project Scope

This study was conducted at Global Nuclear Fuels in Wilimington, NC. The study

included both lattice physics analysis and full core modeling in the 3-d core simulator.

The lattice physics work was further subdivided into the enrichment phase and the

gadolinium phase. For a reference BWR/3 the following projects were undertaken:

1. Enrichment Phase

a. Analyze the effects of homogenous average enrichment perturbation on the
ability to maximize k, difference between the hot operating condition and cold
borated condition.

b. Determine the effect of localized heterogeneous enrichment perturbations on the
ability to maximize k- difference between the hot operating condition and cold
borated condition.

2. Gadolinium Phase

a. Ascertain the effects of gadolinium clumping on maximizing k- difference
between the hot operating condition and cold borated condition.
b. Resolve the effects of increasing the amount of gadolinium rods on maximizing
k- difference between the hot operating condition and cold borated condition.

c. Analyze the effects of gadolinium concentration on increasing k- difference









between the hot operating condition and cold borated condition.

d. Determine a methodology for placing gadolinium rods in order to improve k-
difference between the hot operating condition and cold borated condition.

3. Full Core Modeling Phase

a. Establish the maximum SLCS improvement utilizing an altered geometric
gadolinium placement within freshly loaded fuel bundles.

b. Determine the SLCS attainable from axial power shape perturbations utilizing
enrichment differencing in certain axial zones of the fresh fuel bundles.

c. Resolve the maximum SLCS gained from inserting extra gadolinium rods into
the freshly loaded fuel bundles.

d. Conclude the optimum design strategy for maximizing SLCS.














CHAPTER 2
MODEL AND METHODOLOGIES

In order to determine the necessary strategies for enhancing a margin of shutdown

it is necessary to have a clear definition of that margin. A model and tools to analyze that

model must then be selected that depict the physics of the problem as accurately as

possible. After the designation of a model and utilized tools, the design parameters that

are to be perturbed within the model must be determined. Finally, all other limiting

parameters must be clearly defined so that it may be determined if the improvement to

SLCS is feasible and will not cause the nuclear reactor to violate the thermo mechanical

limits of the fuel.

Standby Liquid Control System and Shutdown Margin

The two shutdown parameters utilized for reactor core licensing are SLCS and

SDM. SDM is a measure of the amount in which the reactor core is shutdown utilizing

all of the control blades excluding the highest reactivity worth control blade.

S kif kCHBWE
SDM = k k
keff (2.1)

CHBWE = controlled case highest worth blade excluded

If keff = 1 then

SDM = 1 k (2.2)


Therefore SDM is the reactivity needed to make the system critical or conversely viewed

as the amount of reactivity in which the system is shutdown utilizing all of the control

blades except for the highest worth control blade.









SLCS is a measure of the amount that the reactor core is shutdown utilizing a

homogenously dispersed boron poison solution.


SLCS= keff -kB.3)
keLC= ff (2.3)
keff

B = borated keff

If keff = 1 then

SDM = 1- k (2.4)

Similar to SDM, SLCS is the reactivity needed to make the system critical or conversely

viewed as the amount of reactivity in which the system is shutdown utilizing a

homogenously dispersed boron poison solution.

Both parameters depict the amount in which the reactor is safely shutdown;

however, the difference in geometry of the poison causes the physics involved in each

shutdown process to be significantly different. Because shutdown margin calculations

utilize a control blade, a heterogeneous poison located on the boundary of two sides of

the fuel bundle, the power distribution of the SDM case is expected to be skewed radially

across the fuel lattice with power peaking occurring in areas furthest from the control

blade. SLCS calculations utilize an evenly dispersed boron poison solution; therefore the

power distribution in the lattice is expected to follow the power distribution dictated by

the actual geometry of the fuel lattice.

The two different modes of control have two separate types of lattice reactivity

responses. Due to this significant difference in reactivity response, it is possible that the

ability to meet specified margin may be satisfied in one mode but not in the other.

Understanding the reactivity response to each mode's shutdown independently and then









utilizing the commonalities in maximized shutdown in each of the two modes ultimately

leads to a fuel design that has maximized margin in both cases.

Modeling Tools

TGBLA 6

TGBLA 6 is a static, multi-group, 2-dimensional, diffusion theory code with

transport corrections factors that assumes infinite lattice behaviors. The steady state

multi-group diffusion equation that is solved is [14]:


-VD (r). V+ o-g (r) (r)= o-g, (r),) () + -Xg vo-,g, (r)g, ((r) + q (r) (1.5)


Because of the major differences in fuel bundle design, void fraction history,

control blade history, enrichment distribution, gadolinium content and accumulated

exposure, the fuel bundle's nuclear characteristics in the core are very different both

radially and axially. Neighboring fuel bundles also have an influence on the

characteristics of the modeled fuel bundle; however, modeling the effects of these

neighboring fuel bundles may be a daunting task because each neighboring fuel bundle in

the core incurs nuclear characteristics that are unique from every other bundle. Therefore

assumptions have to be made in order to be able to achieve an effective and timely

approximation of the fuel bundle's nuclear behavior [5].

TGBLA 6 makes key assumptions in order to accurately approximate a fuel

bundle's nuclear characteristics. Because fuel bundle designs may be varied axially in

bundle geometry, gadolinium content, enrichment, and void concentration, the influence

of axial conditions are considered of primary influence to the fuel bundle's nuclear

behavior. TGLBA 6 completes 2-dimensional lattice physics calculations at different

exposure points for certain axial sections where there exists a known major variation in









fuel bundle geometry. Because in certain defined axial zones the void concentration

changes drastically and because the fuel bundle characteristics are also needed for certain

temperature states, each fuel axial zone is modeled at 0%, 40%, and 70% void

concentration. Though potentially any parameter may be varied by TGBLA 6, the main

variables manipulated in a lattice design are the pellet enrichment, number of gadolinium

rods and gadolinium concentration in each gadolinium rod. As a result of utilizing 2-

dimensional calculations in certain axial zones of the fuel bundle, enrichment distribution

and gadolinium content are only varied in the axial zones represented by the 2-

dimensional lattice physics calculations [5].

Because the influence of neighboring fuel bundle was assumed a secondary

influence on the bundle behavior, TGBLA 6 assumes infinite lattice behavior as a

boundary condition. Assuming infinite lattice behavior results in a good approximation

of the lattice power peaking distribution as well as an accurate generation of group

constants to be later utilized in PANAC11.

Utilizing these design constraints and boundary conditions, TGBLA 6 uses the

solution of the multi-group diffusion theory equation to generate group averaged cross-

sections for 3 energy ranges. Group constants are generated for the fast, epithermal and

thermal energy range to be later used by 3-dimensional core simulator PANAC 11.

PANAC11

After the multi-group cross sections were collapsed and generated by TGBLA6,

Panacl 1 was utilized as the 3-dimensional full core simulator. Panacl 1 is a static, three-

dimensional coupled nuclear-thermal-hydraulic computer program utilized to represent a

BWR core by a coarse-mesh nodal, 1-1/2 group (quasi-two group), static diffusion theory

approximation. The program was utilized explicitly for detailed three-dimensional









calculations of neutron flux, power distributions, and thermal limits at different exposure

steps during reactor core life. The main variable parameters in PANAC 11 were the

control rod positions, refueling patterns, coolant flows, reactor pressures, reactor power

level as well as other operational and design variables [8].

The diffusion equations are solved using the fast energy group. Resonance energy

neutronic effects are included in the model by relating the resonance fluxes to the fast

energy flux. The thermal flux is represented by an asymptotic expansion using a slowing

down source from the epithermal region. A pin power reconstruction model is also

implemented to account for the effect of flux gradients across the nodes on the local

peaking distribution.

Utilized Temperature States, Boron Concentrations and Lattice Types

There was a combination of 4 main types of temperature states and boron

concentrations investigated in the study. These states included the hot uncontrolled state

(HU), cold uncontrolled state (CUO), cold controlled state (CC) and a cold state

containing soluble boron (CU###). The HU state was defined to be the operating

temperature state with no control blade placed next to the fuel lattice. All cold states

were to be defined at a moderator temperature of 160C, and all the cold uncontrolled

states also had no control bladed placed next to the fuel lattice. The CU### condition

was designated as a cold lattice containing a homogenously dispersed soluble boron

solution at a specified boron concentration. The CC condition represented a cold fuel

lattice with a control blade placed in the upper and left side of the lattice.

There were two fuel lattice types examined in the enrichment perturbation portion

of this study. A "C" lattice was defined to be a fuel lattice that exhibited the same

amount of moderator spacing on all four sides of the lattice while a "D" lattice exhibited









slightly more moderator spacing in the vicinity of a control blade location. Therefore the

radial power distributions of the two different lattices are slightly different in lattice

peaking characteristics.

Measurement of SLCS and SDM during the Lattice Development Stage

SLCS and SDM are both global parameters that describe a margin experienced by

the entire nuclear reactor core. Therefore the calculation of these parameters involves

utilizing a 3-dimensional core simulation tool. Fuel bundles are generally designed by

first utilizing a 2-dimensional fuel lattice physics tool to create average collapsed group

cross sections to then be utilized by the 3-dimensional reactor core simulator. An

abundant amount of energy groups are utilized in the lattice physics calculation in order

to properly model the physics of the lattice. Many bundles within the reactor core will

exhibit similar characteristics due to the similar enrichment or gadolinium concentration

within the fuel bundle. Therefore by generating these average group cross sections for

similar fuel bundles, the 3-dimensional core calculation is significantly faster because the

calculation does not involve solving equations at an abundant amount of energies for

many different points within the reactor core [14].

The 3-dimensional simulator only utilizes few averaged group cross sections

(usually 3 averaged groups). The 3-dimensional core simulation tool utilized for this

study, PANAC 11 separated the reactor core into a series of 6 in. cubic nodes. A flux was

then solved in each individual node.

Enhancement to the fuel bundle design therefore involvements modifications to the

fuel design in both the lattice physics development stage and the 3-dimensional core

simulation stage. Since SDM and SLCS global parameters defined for the entire system,









a separate set of parameters must be utilized for characterizing how fuel improvements in

the lattice development stage will affect the full core global parameters.

Maximizing SDM and SLCS involves increasing the difference between the hot

operating condition and the cold shutdown condition. Therefore in the lattice

development stage an enhancement in SLCS and SDM meant and improvement in the

difference between the hot operating k- and cold shutdown k-.

The cold shutdown condition related to SDM is defined to be when the lattice is

controlled by the placing a control blade next to it. The parameter used to represent the

maximized difference between hot operating k- and cold shutdown k- was designated

HUCC, and calculated by the equation:

HUCC = k HotOperahng_ k ColdControlled (2.5)

The cold shutdown condition related to SLCS is when the lattice is controlled by

placing a homogeneously dispersed solution throughout the fuel lattice. The parameter

used to represent the maximized difference between hot operating k- and cold shutdown

k- was designated HUCU###. The symbol ### represents the amount of parts-per-

million of boron in the solution. HUCU### is calculated by the following equation:

HUCU# # #= kHotOperathng kColdBoratedSoluhonat ###ppm (2.6)

Maximizing these parameters in the lattice development stage will maximize the

global parameters that these parameters represent in the 3-dimensional core modeling

stage.

Fuel Bundle Geometry

The fuel bundle utilized in the lattice physics calculations was a typical 10X10

boiling water reactor fuel bundle design with 92 fuel rods and two water rods. Figure 2-1









displays a cross sectional view of the fuel bundle geometry. Since each lattice physics

calculations was completed utilizing a 2-dimensional model, the fuel bundle had to be

sub-sectioned into 2-dimensional axial zones in order to accurately model the sections of

the bundle that experience different void concentrations, decreased average moderator

density, variable gadolinium and enrichment placement, and different lattice geometries.

The axial zones utilized were designated naturally enriched bottom (NAT), power-

shaping zone (PSZ), dominant zone (DOM), plenum zone (PLE), vanished rod location

zone (VAN), natural vanished rod zone (N-V) and natural top zone(N-T). The NAT was

a naturally enriched zone filled with all 92 fuel rods and no gadolinium. This zone

represented the first 6 inches of the bottom of the active core length. Both the PSZ and

DOM were enriched zones containing 92 fuel rods with gadolinium present in certain

locations. The PSZ zone was located on top of the NAT zone and was 48 inches in

length. The DOM zone was located on top of the PSZ and was 30 inches long. Though in

2-D geometry these axial zones were identical, the zones are separated due to the

different inherent thermo-hydraulic and neutronic characteristics experienced in each

zone.

The PLE was a 12 inch zone containing 78 fuel rods and gadolinium rods present in

needed locations. This zone was used to model the interface of the part length rods and

the vanished rod locations began. On top of the PLE was the VAN. The VAN contained

78 fuel rods and 14 vanished rod locations and ranged between 37 and 48 inches in

length. On top of the VAN was the N-V. The N-V contained natural enrichment and has

pellets existing in 78 fuel rod locations. The N-T was also naturally enrichment but only

had pellets in locations where no gadolinium existed in axial portions of that specific rod.












0000000000
OOOOOOOOOO
FOo oor
00000 0000
00 0000
000 0000
00 00000 J
S Co~t~ir Blad


KK


F Full Length Ful
) Part Lengh F I
Tie Feul Rod

w Large Water Rod
TopGe Beanm Getertkne


cI


I9


Oq0 ooo0 oo
O^ o_ o oo 7Po
00 0 ow

0100000 000
0000000000

oo000000000
-4--


1I


I(f
HW-" i


Figure 2-1. A cross sectional view of the modeled fuel bundle.


T-


HA


As displayed in figure 2-2, there existed only 4 possible fuel rod geometries. The

DOM and PSZ had the same fuel rod geometry, and the VAN and N-V had the same fuel

rod geometry. In the N-T the locations marked "E" are used to represent empty fuel

locations in the lattice of where gadolinium rods exist in the N-V. In the N-V, N-T, and

VAN the locations marked "V" are used to represent the vanished rod locations. In the

PLE the locations marked "E" are used to represent the area of the plenum tip of the part

length fuel rods.

Though 4 different possible fuel rod geometries exist only 3 different types of

geometries were modeled in the lattice physics investigations. The DOM, PLE and VAN

region were modeled. The N-T was not investigated because this region contained only


0000000000
0000000000

00000 w000
( 0 00000
000 00000
0000000000
0000000000
o0o0OOOO
00 D00 0 0


- ^


Rod


I


L


c


















natural enrichment and therefore the low power level experienced in this region would




never be most limiting in any cold shutdown condition.


N-T
A 8 0 D E


0,71 071 071 71 071 71 071 071 071 0.71 071

0,71 V 071 V 0.71 E V E V 0,71

0 71 071 E 071 E 071 E 071 E 071

0.71 V 0.71 E 071 WR 0.71 V 0.71

0.71 0.71 E 71 V 0.71 0.71 0.71

0.71 E 071 WR V E 07-1 E 0.71

0.71 V E E 0.71 E V 0 71

0o.1 E 0o71 071 0.71 0.71 E 071 0.71

0.71 V E V 0.71 E V E V 0.71

0.71 071 071 07.7.71 0.71 0.71 071 0.71 071


M-V
A I 0 D E F O H I ,J
071 071 071 071 071 071 071 071 071 071

0.71 V 0.71 V 071071V 0 71 V 071

071 0.71 071 071 071 0071 071 0771 071 071

071 V 0.71 0.71 0.71 'R 071 V 0,71

071 0.71 071 0.71 V 0.71 071 071

0,71 071 071 R V o.71 0.71 0,71 071

071 V 0.701 07.071 0.71 V 071

071 071 0.71 0.71 071 071 0.7 071 071071

071 V 071 V 071 0.71 V 071 V 071

071 0.71 071 0.71 071 0.71 0,71 071 071071


PLE
A B D E F a H I J
1,6I W 3 3.20 3Z9 395 Z9 5 3.L35s 3..9 3.95 2S

30o E .0 E 34O 3.i5 E E ,5
7I ~ 777 4-90 ..40 ^^

3.9S E 3 4 .4 3.95 'R 4-90 E 490

39 380 35 E -90 490 l90

3,95 3 o5 4,3 *A IR E 4- 4

3.gS E g 490 &M E 14,90













, = 4. J a o a J 4

4 4 4 "
*. 40 490 W4 90 4 -' 1 4,90
"E i0E 4S E
220- 3.95 4.90 4.909 490 4 9: 490 4,


PSZ


1W&0 20 3.20 Z.96 3Z953,95 3.951 35 3.95 Z80



*-, ;..40 a


S,9- 30 3.95 4.9 490 4.90 490
S 2a 440 WR L 90 4.9 4.0 44.9



--..- I 0 Sk5



2SO 3.95 4.90 4.90 9 49 4. M 4.90 4.90 J30


Figure 2-2. The geometric setup of the fuel lattice axial zones.


Thermal Limit Design Considerations




Nuclear reactors are designed so that operation will not induce unnecessary risk to




the health and safety of the general public. Therefore thermal limits are imposed on


Hear
I 4.2

tie::'


6.00


V. I 3724


%60- 1


RLE


1

-i
2


4

6
1200 a

7


IS







-I
1

2



4
6



7


VAN
A U O D E


395 V 4.40 3.9 R -V 4.0 V 4.90-
3.B 3.60, 3 j V -A-. 4 90- 4,90 4390

-5 S- 40 W R V M,1 W 4 _So

395 V 0,7 -1 4.90 V 1 4.90

39 4 4043 4 4.90 0 1 401 0
[EV4 -90 B v

23 4.90 490490490490 9040 3


DOA
A a 0 D E F O H I J








3031 0316 071 .5 90 0- 4.930 4,90 4-90

07S710" 071 0710-











07 07 07 071 071 071 0.71 07A 1
S95&04.0 4- -.90 4.9003770











01 071 0710.71 07071071071071 .71 071 071 0.71 071


0,71 071 0.71 071 071 071 071 071 071 071

071 071 071 071071 071 0.71 071 0.71 071


PSZ 4AM00

2

3

4

5
e


7


a
NAT 6.00J
1S


F 0 H 1 J


F O H I j









certain core parameters to ensure that radioactive release during reactor operation or any

type of transient event does not exceed the acceptable limits imposed by the NRC.

Constraining operation to within the thermal limits of the fuel guarantees that during

normal operation and emergency transients the fuel integrity will be maintained. For the

applicability of this project the two main thermal limits monitored were MFLPD and

MFLCPR. These limits are set to limit boiling around the fuel rod locations and limit

fuel rod power density in order to preserve fuel integrity [11].

Linear heat generation rate (LHGR) is the amount of power produced per length of

fuel and is defined as

Average LHGR = Maximum Thermal Power Output (1.6)
S (Number _of _Fuel _Assemblies XFuel Rods _Per _Assmebly XFuel Rod Length )

A maximum average LHGR is specified for the utility in order to limit the plastic

strain or deformation of the cladding. A limit of 1% deformation of zicaloy cladding is

considered a conservative limit below which fuel damage is not expected to occur.

New pellets undergo slight densification during irradiation. This causes the gap

between the fuel and the cladding to increase and thus decrease thermal conductance.

The pellet densification also has a shrinking axial effect. If one of the pellets gets stuck

during this process, a gap is created resulting in more fissions occurring in newly exposed

faces of the pellets increasing heat flux in that area [6]. Therefore LHGR is adjusted for

the possible elevated heat flux and is defined as


LHGRh1int = LHGRdegn 1[- m (1.7)
1718X LT

Where:

LHGRdesign = Maximum LHGR allowable to prevent clad damage









LT = Total active core length

L = Axial position in feet above the bottom of the core


I ax= Maximum power spiking penalty
P max

In the core simulator LHGR is calculated for certain axial nodes. MFLPD is the

maximum fractional limiting power density for the most limiting node and is defined as

LHGR
MFPLD = maxnode (1.8)
LHGRhm ,t

As long as MFLPD is less than one LHGR is not exceeded. However, because

these calculations are based off of the assumptions of the maximum power spike penalty,

and because the designer needs to be certain that MFLPD will never exceed one, the

design basis requirement for MFLPD is 0.909 to ensure enough variation between the

actual calculation and the measured data [8].

Critical power is the bundle power required to produce transition boiling in a

reactor channel. If transition boiling were to manifest in a channel, it may lead to fuel

rod dry out in the channel with the inability of the fuel rod surface to rewet. This

phenomenon leads to a decrease in the ability of the clad to reject heat to the water

through convection and thus heat up the clad to the point of mechanical failure [6]. CPR

is the ratio to determine how close the actual power is to transition boiling and is defined

as:

CP
CPR =- (1.9)
AP

CP = Critical power for transition boiling

AP = Actual power.









CPR must always be below 1.0 for safe operation. MFLCPR is the flow adjusted

ratio of the operating limit CPR for a specific fuel type to the CPR of that bundle and is

defined as:

CPR Limit K
MFLCPR = -(1.10)
CPR

Kf = Flow adjustment factor

Since MFLCPR should never exceed one in any section of the reactor during

operation, and since slight uncertainty exists in knowing the actual power of the reactor

and the critical power for a specified bundle, the design basis for MFLCPR is set to 0.930

to accommodate these uncertainties [8].















CHAPTER 3
MAXIMIZING HOT-COLD BORATED k- DIFFERENCE UTILIZING
ENRICHMENT

TGBLA 6 was utilized to understand SLCS improvement by enhancing lattice

behavior characteristics utilizing enrichment perturbations. C and D lattices types were

examined at 0%, 40% and 70% void fraction. The four system states inspected were HU,

CUO, CU###, and CC. Lattices with a homogeneous enrichment distribution were

examined to determine the effects of lattice average enrichment on the enhancement of

the HUCU### and HUCC. Next, Local and gross enrichment perturbations were also

analyzed to determine the effects of these types of enrichment perturbations on

HUCU### and HUCC.

Since an enormous amount of combinations of temperature states, lattice axial

zones, lattice types and void concentrations could be created, the homogenous

enrichment work was utilized to determine which of these temperature states, lattice axial

zones, lattice types and void concentration were most limiting in order to limit the

amount of cases to investigate therefore only examining the most effective strategies for

enhancement.

Homogeneous Enrichment Distribution

A homogeneously enriched distribution was defined to be a fuel lattice with

constant enrichment throughout the lattice. Therefore homogenous enrichment

perturbations were defined as a change in enrichment to every fuel pin in the lattice by

the exact same amount.












Determining the Most Limiting Lattice Axial Zone and Void Concentration


A variety of tests were completed to determine which lattice parameters were most


limiting to HUCU### and HUCC. Figure 3-1 depicts the effects of lattice type and void


fraction on HUCU###. Lattice type did not significantly affect HUCU###; however, as


void fraction increased HUCU### decreased.


For this case, at 0 GWD/STU the difference in HUCU### was solely related to the


effective moderator density difference at increased void fraction. At higher void


fractions the average moderator density was decreased. Because the average moderator


density was decreased fewer neutrons were thermalized and absorbed by the fuel for


fission, and an increased amount of neutrons were parasitically captured by the fuel [18].


Therefore HUCU### was smaller for higher void fractions.




0 18

0 16

0 14 -

0 12
S--- 0% Void, C Lattice
0 1 --40% Void, C Lattice
70% Void, C Lattice
0% Void D lattice
0 08 --40% Void, D Lattice
70% Void, D lattice
006

004

0 02


0 10 20 30 40 50 60 70
Exposure (GWDISTU)

Figure 3-1. Exposure dependent HUCU660 for the DOM at 3.95% enrichment.


Initially U238 was the main parasitic neutron absorber in the fuel due to increased


void concentration. When U238 absorbed a neutron it became Pu239. Because Pu239 had a


high thermal absorption cross section (oa = 1015b) as well as multiple resonance


absorption peaks, it became another main parasitic neutron absorber. The increases in









build up of parasitic neutron absorbers lead to a decrease in the effectiveness of boron to

capture thermal neutrons [18]. Therefore as the lattice burned, plutonium was built up

thus increasing the content of competing neutrons absorbers and therefore decreasing

HUCU###. Furthermore, in the higher void history condition it took longer for

plutonium production to reach an equilibrium state; therefore at increased void history

conditions HUCU### decreases at a faster rate for a longer amount of time.

Figure 3-2 illustrates the effects of the combination of axial zone and void fraction

on HUCU###. The DOM was the most limiting axial zone because of the maximum

amount of fuel rod inventory present in the lattice and minimum amount of volume to

place borated water. A lattice geometry that allows for more moderator space allows for

more ability to place borated water in the lattice; therefore since the VAN and PLE both

have evacuated regions where more space exists to place a boron volume these axial

zones were not the most limiting in terms of HUCU###.

In order to maximize improvements in HUCU###, enhancements must be made to

the most limiting conditions ofHUCU###. HUCU### was most limiting in the 70%

void fraction and DOM case; however, in the core, on average, the DOM exhibits a 40%

void fraction therefore modeling a DOM at 70% void fraction would not have been an

accurate realistic model to examine SLCS. The realistic model utilized which was most

limiting was determined to be 40% void fraction in the DOM. Therefore since lattice

type did not significantly effect HUCU###, and the DOM 40% void fraction was the

most realistic limiting condition state, the C lattice type, DOM, 40% void fraction lattice

was chosen as the base lattice in which all other perturbations were compared.











025




02



--Dom Zone, 0% Void
0 15 -Dom Zone, 40% Void
-Dom Zone, 70% Void
SPie Zone, 0% Void
Pie Zone, 40% Void
Sc --Pie Zone, 70% Void
-Van Zone, 0% Void
0 1 --Van Zone, 40% Void
Van Zone, 70% Void



005





0 10 20 30 40 50 60 70
Exposure (GWDISTU)

Figure 3-2. Exposure dependent HUCU660 at varied void fraction and axial zone for the
C lattice at an enrichment of 3.95%.

Understanding the Exposure Dependent HUCU### Curve

Understanding the exposure dependence of the HUCU### curve was paramount to


determining the appropriate strategy for enhancing HUCU###. Figure 3-3 depicts the


two major portions of the exposure dependent HUCU### curve. Portion A encompasses


0 GWD/STU to roughly 11-15 GWD/STU depending upon geometry of the axial zone


and void concentration. Portion B encompasses the rest of the HUCU### curve.


When a fissile isotope absorbs a neutron, the isotope may either undergo fission or


parasitic capture. The capture-to-fission ratio is defined by:


O
a' (3.1)
(7f


In a thermal reactor the majority of the neutrons cause fissions at thermal energies in


U235. Therefore for most thermal reactor applications the capture-to-fission ratio is an









explanation of the fission efficiency of the thermal neutrons [13]. As ac increases k-

decreases because more thermal neutrons undergo parasitic capture and are removed

from the system instead of undergoing a fission event and creating more neutrons [18].

During portion A of the exposure dependent HUCU### curve, HUCU### is

decreasing due to increased plutonium production. Pu239 has a capture-fission-ratio of

0.370 (2200 m/s neutron) while U235 has a capture to fission ratio of 0.175 (2200 m/s

neutron) [13]. Therefore with an increased a plutonium acts as a competing neutron

absorber that decreases boron worth thus limiting the effective absorption ability of the

boron.

The capture-to-fission ratio is a function of the system temperature. Doppler

broadening is a phenomenon in which due to the kinetic motion of the target atoms at

elevated temperatures the resonance absorption cross sections broaden while the peak

magnitude of the cross section decreases, and in most cases slightly preserving the area

under the original resonance [3]. Therefore though the effective peak of the cross section

has decreased, the width of the resonance is increased and therefore the resonance affects

a greater range of energy of neutrons; therefore causing a greater interaction rate in that

energy interval and thus leading to more absorption and decreased flux in that energy

interval [15]. At higher temperatures there is more kinetic motion of target particles and

thus more Doppler broadening of the resonance cross sections. With increased parasitic

capture and decrease thermal-to-fast flux ratio, HU k- decreases at a much faster rate

than CU### during Portion A of the HUCU curve because the worth of the plutonium

produced is progressively worth more in the hot operating condition than in the cold

condition.









The thermal-to-fast flux ratio is defined by:


a, = thermal (3.2)
1 fast

As the temperature of the system increases the average moderator density

decreases. Because the average moderator density is decreased the neutron spectrum has

a higher density in the fast region. In the hot operating condition the decreased average

moderator density causes an increase in ai; therefore fewer neutrons are available for

thermal fission events. Since at elevated temperatures there exists greater Doppler

Broadening as well as decreased average moderator density, the decrease in al leads to a

decrease in k- of the system. Therefore the increasing a and decreasing aL in the hot

condition leads to a decrease in the HUCU### curve because the hot k- is decreasing

faster in comparison to the cold k-.

















Exposure (GWD/STU)

Figure 3-3. Exposure dependant HUCU### curve.

During portion B, as plutonium production reaches an equilibrium concentration

the difference between the hot operating condition and cold condition worth becomes











almost constant and therefore during this portion of the curve HUCU### experiences


relatively no exposure dependence no exposure dependence.


Enrichment and Boron Concentration Effects

Figure 3-4 depicts the effects of increasing enrichment on HUCU### at different


boron concentrations. For each 1.0% increase in average enrichment 0.0188 HUCU###


was lost. Since the derivatives of HUCU### were equivalent at different boron


concentrations, the amount of HUCU### gained from a boron concentration increase was


linearly dependent on the boron concentration increase and not also affected by the


average enrichment. Equation 3.3 calculated 1.84 x 10-4 HUCU660 gained for each 1


ppm or boron introduced to the lattice. Therefore 99.6 ppm of boron was required to


compensate for a 1.0% average enrichment increase.


HUCU### gained AHUCU### (3.4)
1 ppm Boron ABoron Concentration pec Encmnt



03


0 25





0 15- 742 PPM
792 PPM
-I-935 PPM
01


0 05



0 04 08 12 16 2 24 28 32 36 4 44 48 52
Enrichment

Figure 3-4. Beginning of cycled HUCU vs. enrichment in the DOM, at 40% void
fraction, for a C lattice.











Power Peaking Distribution

The lattice power peaking distribution was a function of the relative distance of


fissile material from the moderating regions. Moderation capability of certain lattice


regions was a function of the water boundaries of that certain lattice region as well as the


temperature state, boron concentration and geometry of poison utilized within the lattice.


Power Peaking distributions for a homogenous 3.95% enrichment lattice at 5 GWD/STU


are displayed in figure 3-5 for HU, CUO, CC and CU660. Each fuel pin location is


identified by the horizontal and vertical location in which the pin resides. The water rod


locations were marked with a zero in order to distinguish the water rod locations from the


fuel pin locations.

5 ACHJ395Enndcrert, DmZore, CLatbce4MP/oVi d 5 GACQ3 95 Ennchmr t DmZxe, CLattce 4/obVad
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
1 1122 10747 11 1 9 1 251141 1 057402 057 8 0618 0646 43 8Q 113 2 10067 1l B 2 0520538 06241 02t762 Q7 07~ 9 130>x> 12
3 11229 11415 3 Q57 06241 06763 0741 10321 12>x>110
4 10747 0 1022 4 0588902 0741 10385 0 0 1 C919 28 110>x>10
5 1061 0 01069 5 0615807062 1035 0 0 1129710878 105>x> 100
6 10 0 0 10621 6 06496 07605 0 0 11616 10421 1019 100>x>0
7 125 0 0 1076 7 06943079 0 0 116616 103131021 12 095>x> 090
8 1141 11245 8 07099 19129710421 11 2151 1454 0> x-> 085
9 10 1 9 10132 1 OE2 10B78 149 1 1454 1085>x 0
10 11415 1 1 10621 10711245 10 080>x

5 GCL,395Ennchrr DmeZe, CLatoeD/oVad 5 GWoC ,3 9 BEndirr [hmZae, CL atoe 4C/oVd


1 11221 1 OB47 107713 11016 1 1378 1 11859 113 1079 10736 1 0 1 0 11197 11956
2 2 11859 11963
3 11221 07945 07877 11395 3 11033 1 12D9
4 1047 07877 0 0 1104 4 10779 0 0 1C46
5 1072 10713 0 0 1 02 5 10736 15 0 0 1 B31




9 9 11~6 1189
10 11395 11041 s 10 6 1B 11284 10 113 1129 10946 1 0831 10762 1 B09 11097 118981
Figure 3-5. The power peaking distributions at 5 GWD/STU, 3.95% enrichment,
DOM, C lattice, and 40% void fraction vs. temperature state and boron
concentration.









Figure 3-5 suggests that there were significant differences between the HU and

CUO power peaking distributions. In the HU state the outer borders of the lattice

exhibited the greatest amount of power peaking due to the close proximity of large areas

of water to that location and therefore displaying the greatest moderation capabilities.

Due to the moderation of the internal water rod locations, fuel pins located near the

internal water rods exhibited a higher relative power than locations that were located

away from the borders of the lattice and away from the internal water rod locations. The

lack of moderation for the fuel pins located away from the borders of the lattice and away

from the internal water rod locations caused these locations to exhibit the lowest relative

power.

In the CUO state, power was raised in the highly moderated areas. With no voids,

the outer borders of the lattice and the internal water rod locations exhibit a greater

amount of power peaking than the areas of the lattice that were between these locations.

Because of this effect, the importance of fuel pins located away from the borders and

water rod locations were significantly decreased, and if a perturbation were to be made to

a fuel lattice in order to improve inherent HUCUO characteristics these locations would

not play an important role.

There were also significant differences in power peaking distribution for different

poison types. In the CC state a high anisotropy of power peaking was exhibited due to

poison residing at the covers of the lattice. Fuel pins located closest to the control blade

exhibit the greatest power suppression; therefore if a poison introduction was necessary

for power suppression in the HU state, placing that poison away from the greatest power

suppressed pins in the CC condition will achieve the greatest improvement in HUCC.









In the CU660 state the boron caused the power to be suppressed in highly

moderated regions thereby flattening out the power distribution of the lattice. Because

boron was present in the moderator, regions that had greater power peaking due to

increased moderation capability also consequently had greater power suppression from

the boron dispersed within the moderator. Because of the flatter power distribution in the

CU660 as compared with the HU state, placing power suppressors in peaked locations

corresponding to the HU state did not necessarily have as severe an impact on the cold

borated state. However, this leads to a distinct design advantage because it may be

possible switch locations of a distributed poison and have a miniscule effect on HU but a

major effect on CU###. Therefore in order to maximize HUCU###, power suppressors

must be placed in areas where the CU### state exhibits a higher power peak than the

power peak in the HU state.

As enrichment increased the power peaking distribution in the lattice became more

skewed. Figure 3-6 displays that as average enrichment was increased in the CU### state

the power peaked more in the border regions and internal water rod locations of the

lattice. Figure 3-7 displays that as average enrichment was increased in the HU state the

power also peaked more in the border regions and internal water rod locations of the

lattice. Therefore there exist fewer locations in which moving a distributed poison will

not greatly affect the HU state while greatly affecting the CU### state.

Unfortunately this demonstrates that in a power up rate or increased exposure

design, it will be harder for the designer to create a design that improves SLCS and

decreases the power peak in the HU state. Figure 3-6. The power peaking distributions at

5 GWD/STU, CU660, DOM, C lattice, and 40% void fraction vs. enrichment.














5 GWD,CU660,0 71 Enrichment, Dom Zone, C Lattice 40% Vad 5 GWD,CU0,3 95 Ennchment, Dm Zone, C Lattice 40% Vad
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
1 1 1923 1095 10595 10471 1 0462 10503 10557 1067 10996 1 1929 1 1 1221 10847 10772 1 0853 1 1016 1 1378
2 1 095 1 0986 2
3 10595 1 0662 3 1 1221 07945 07877 11395
4 10471 0 0 10553 4 1 0847 07877 0 0 104
5 10462 1 0681 0 0 10501 5 1 0772 10713 0 0 10882
6 10503 0 0 10683 10461 6 1 0853 0 0 10717 10806
7 10557 0 0 10472 7 1 1016 0 0 07893 1 0885
8 1 067 10597 8 1 1378 07893 07966 1 1264
9 10996 1 0953 9
10 1 1929 10986 10662 10553 1 0501 10461 10472 10597 10953 1 1923 10 1 1395 1 104 10882 1 0806 10885 1 1264

5 GWD CU660,3 2 Ennchment, Dom Zone, C Lattice 40%Vad 5 GWD, CU660, 4 9 Enrichment, Dom Zone, C Lattice, 40% Vad
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
1 1168 10965 10704 1 0665 10733 10846 1 1 1166 10862 10816 1 0891 1 1016 11303
2 1 168 1 1775 2
3 10965 11101 3 1 1166 07969 07943 11319
4 1 0704 0 0 10859 4 1 0862 07943 0 01038
5 10665 1 0768 0 0 10751 5 1 0816 1095 0 0 10919
6 10733 0 0 10772 1 0686 6 1 0891 0 0 1 09541 0848
7 10846 0 0 10728 7 1 1016 0 0 07959 10898
8 1 1093 1 0993 8 11303 07959 0799 1 1206
9 11771 11712 9
10 1 1775 1 1101 10859 1 0751 10686 1 0728 10993 1 1712 10 1 1319 1 1038 10919 1 0848 10898 1 1206

Figure 3-6. The power peaking distribution at 5 GWD/STU, CU660, DOM, C lattice,

and 40% void fraction versus enrichment.


5 HU,0 71 Enchment, Dcm Zone, C Lathce 40% 0 d 5 GVOHU,395 Ennchmet, Dom Zne, C Lattice 40%oVad
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 K
11219 10702 10443 10365 10418 10565 10836 1134 W 1 11229 10747 1061 1069 10925 1141 1130 < x

2 11219 10126 11328 2 10067 102 130 > x > 12C

3 10703 10827 3 11229 11415 120 > x > 1 1

4 10443 0 0 10557 4 10747 0 0 1932 110 > x > 1 0

5 10365 10071 0 0 10412 5 1061 0 0 1699 105 > x > 1 OC

6 10418 0 0 172 1036 6 1069 0 0 1 621 100 > x > 0 9
7 10565 0 0 10438 7 10925 0 0 1076 0 95 > x > 0 9C

8 108336 8 1141 11245 0 90 > x > 0

9 1134 10126 11215 9 1028 1008 0 85 > x > 0 8C

10 11328 10827 10557 10412 1036 10438 10698 11215 10 11415 10932 10699 10621 1076 11245 080 x


5 GDHU,32 Enndimet, DomZcne, C Lathce 40/ Vad 5 GAC HU, 49 Erndmmet, DomZcne, C Lance, 40%/Vad

1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
1 11116 10668 10537 10614 1 120 9 1 11346 10837 10695 10778 11017 11527

2 1i 1 1013 2 1 004 1 16
3 1 1116 1 128 3 1 1346 0798 1 1533

4 10668 0 0 10845 4 10837 07826 0 0 11026

5 10537 0 0 10621 5 10 5 10045 0 0 10789
6 1 0614 0 0 1 0546 6 10778 0 0 1 00471 0709




9 1 013 1013 9 1 6 1002
10 112 10845 10621 10546 107811 10 11533 11026 10789 107 10853 11365

Figure 3-7. The power peaking distributions at 5 GWD/STU, HU, DOM, C lattice, and

40% void fraction vs. enrichment.









Heterogeneous Enrichment Distribution

After determining the effects of average enrichment on the behavior of HUCU###,

heterogeneous enrichment perturbations were then examined in order to determine if

enrichment changes to individual fuel pins could affect HUCU###. The two major types

of enrichment perturbations investigated were localized enrichment perturbations and

gross enrichment perturbations. Localized enrichment perturbations were considered to

be small sets of fuel pins that were either increased or decreased in enrichment by a

certain amount holding the rest of the fuel lattice at constant enrichment. Gross

Enrichment perturbations were considered to be a large lump of fuel pins that were either

increased or decreased in enrichment by a certain amount holding the average enrichment

of the entire lattice constant.

Localized Enrichment Perturbation

Localized enrichment perturbation patterns were generated based on the power

peaking distribution map. The fuel pin locations were set into groups based on locations

exhibiting similar power peaking in the homogeneously enriched lattice calculation.

Since lattice power peaking was determined to be dependent on enrichment, 3.95%

enrichment was chosen as the distribution for which the determination of the pattern type

was made.

Figure 3-8 displays the distribution of perturbations made to the lattice. Each

group of numbers represents a group of fuel pins that were either increased or decreased

by 1.0% enrichment as the rest of the lattice was kept at a constant enrichment.

Localized enrichment perturbations had no effect on HUCU### as depicted in figure 3-9.

The minute difference of 0.00373 HUCU### in figure 3-9 was considered only a function








40



of the 0.2% average enrichment difference exhibited between each pattern utilized.


Therefore localized enrichment perturbation did not greatly affect HUCU### behavior.



1 2 3 4 5 6 7 8 9 10


Figure 3-8. Localized enrichment perturbation map.
Figure 3-8. Localized enrichment perturbation map.


0.078689


0.074957


- Map 1, Pattem 1
- Map 1, Pattem 2
Map 1, Pattem 3
Map 1, Pattem 4
- Map 1, Pattem 5
- Map 1, Pattem 6
- Map 1, Pattem 7
- Map 1, Pattem 8
Map 1, Pattem 9


Maximum Change in Hot Uncontolled k-infinity Due To Cold Uncontrolled k-infinity =
002 0.003732 Wiich is Soley a Function of Average Enrichment Change (0.2%).



0 10 20 30 40 50 60 70
Exposure (GWDISTU)



Figure 3-9. Exposure dependent HUCU660 for different localized enrichment
perturbation patterns.


01


008


' 006











Gross Enrichment Perturbation

Gross enrichment perturbations were next examined in order to determine how

these type of lattice perturbations would affect shutdown behavior. Figure 3-10 displays

an example pattern of gross enrichment perturbations and table 3-1 lists the enrichment

perturbations made to that example pattern.

Table 3-1. Gross enrichment perturbation scheme for figure 3-10.

Enrichment Enrichment
Pattern Perturbation (1-2) Pattern Perturbation (1-2)
1 1.6%-4.9% 7 4.9%-1.6%
2 2.4%-4.9% 8 4.9%-2.4%
3 3.2%-4.9% 9 4.9%-3.2%
4 4.4%-4.9% 10 4.9%-4.4%
5 3.2%-4.4% 11 4.4%-3.2%
6 3.6%-4.4% 12 4.4%-3.6%


1 2 3 4 5 6 7 8 9 10
1
2
3
4 W W
5 W W
6 W W
7W W
8
9
10
Figure 3-10. An example of a gross enrichment perturbation map.

Gross enrichment perturbation demonstrated no effect on HUCU###. Though

HUCU### was not a function of localized and gross enrichment perturbation, HUCC was

highly dependent upon these perturbations. Figure 3-11 and 3-12 display the difference

in effect of gross lattice perturbation skewing. Notice in figure 3-11 no effect was

noticed on HUCU###; however, in figure 3-12 HUCC was highly dependent upon

enrichment distribution.
























- Map 6, Pattern 2
Map 6, Pattern 8


0 10 20 30 40 50 60 70
Exposure (GWDISTU)

Figure 3-11. Exposure dependent HUCU660 at 40% void fraction, in the DOM, with a C
lattice.


025




02-


-Map 6, Pattern 2
-Map 6, Pattern 8


005


0 10 20 30 40 50 60 70
Exposure (GWDISTU)

Figure 3-12. Exposure dependent HUCC at 40% void fraction, in the DOM, with a C
lattice.


Placing a higher enrichment closer to the control blade location allowed for greater


power suppression and thus enhanced HUCC due to the increased control the blade


exhibited over the maximum power producing section of the lattice. Therefore though






43


enrichment perturbation was not limiting in HUCU###, distorting the enrichment

distribution had an effect on HUCC. Though SLCS does not depend on local or gross

enrichment perturbation, SDM is sensitive to this type of perturbation. However, the

designer may only enhance HUCU### by perturbing average enrichment of the entire

lattice therefore as long as the average of the enrichment of the lattice satisfies SLCS

requirements the enrichment may be perturbed to meet SDM without violating SLCS.














CHAPTER 4
MAXIMIZING HOT-COLD BORATED ko, DIFFERENCE UTILIZING
GADOLINIUM

There were four different isolated studies examined for the purpose of determining

the optimum strategies for utilizing gadolinium to enhance HUCU###. Gadolinium rods

were examined in a variety of clumped geometries in order to determine the lumped

spatial self-shielding effects. After the effects of self-shielding were determined, the

effects of increasing the amount of gadolinium rods on HUCU### were investigated.

Gadolinium concentration was next analyzed. Finally, gadolinium rod placement was

examined to determine the optimum gadolinium locations for enhancing HUCU###

without diminishing HUCC.

Spatial Self-Shielding Effects of Gadolinium Rods on HUCU###

Gadolinium rods were clumped in a variety of geometries to determine the effects

of lumped spatial self-shielding on HUCU###. Four samples of examined clumped

gadolinium rod geometries are displayed in figure 4-1. Clumping the gadolinium rods

decreased the BOL gadolinium worth because the gadolinium rods were effectively

spatially self-shielding each other from the impinging neutron flux. The self-shielding of

the gadolinium decreased the effective surface area utilized for neutron absorption [15].

Because the thermal-to-fast flux ratio was much higher in the cold state than in the hot

state more neutrons were likely to be thermally absorbed in the gadolinium in the cold

condition; therefore decreasing the effective surface area for neutron absorption decreases









the effectiveness of the power suppression from the gadolinium rods and thus decreasing

HUCU###.

Pattern 10 Pattern 11
12345678910 12345678910




6 WW ~ WW
7 W_ 7 wWW
8 _8 I--

10 9 10


Pattern 12 Pattern 13
1 2 34 5 6 7 8 910 1 2 3 45 67 8 910
1 1
2 -2
3 3 -- m-
4 WW 4 WW
5 WW 5 ww
6 WW V 6 vW w
7 WW 7 WW
8-- 8-m m
3 9
10 10


Figure 4-1. Four sample clumped geometries.

Figure 4-2 depicts the effects of gadolinium spatial self-shielding on gadolinium

worth. Highlighted in red are the patterns corresponding to those displayed in figure 4-1.

As gadolinium clumping increased, gadolinium worth decreased. Face adjacent

clumping of all four sides of a gadolinium rod resulted in a 38% decrease in gadolinium

worth, and face adjacent clumping of two sides resulted in a 19% decrease in gadolinium

worth. Diagonal face adjacent clumping lead to a 7% decrease in gadolinium worth.






















-0 05


| -0 15
E



S-0 2


Pattern

--HU Gad Worth -U-CUO Gad Worth CU660 Gad Worth CU 935 Gad Worth -*-CC Gad Worth

Figure 4-2. Corresponding 0 GWD/STU gadolinium worth for the patterns displayed in

figure 4-1.






0 12



01



008



S006
U


0 10 20 30 40 50 60 70
Exposure (GWDISTU)
-Pattern 10 -Pattern 11 Pattern 12 Pattern 13

Figure 4-3. Corresponding exposure dependent gadolinium clumping effects on

HUCU660 for the patterns displayed in figure 4-1.









Because clumping a group of gadolinium rods decreased the effective surface area

for absorption, the exposure time required to bum out the gadolinium increased. In figure

4-3 as clumping increased the exposure point in which gadolinium burns out also

increased. Also depicted in figure 4-3 is the decrease in HUCU### as a function of

increase clumping. Therefore in order to design an optimum lattice to enhance

HUCU### gadolinium rods must be spaced as far apart as reasonably achievable and face

adjacent and diagonal adjacent clumping must be eliminated.

The Effects of Increasing the Amount of Gadolinium Rods on HUCU###

In a power up-rate and an increased exposure cycle, extra positive reactivity must

be installed in the fuel bundle. Placing extra positive reactivity will cause a greater

skewing of the lattice power peaking as well as violating beginning of cycle critical

eigenvalue requirements. In order to decrease the beginning of cycle eigenvalue to the

critical requirements and decrease power peaking to improve lattice efficiency and meet

thermal margins, gadolinium must be placed in the fuel bundle. As more positive

reactivity is installed, more gadolinium rods at higher concentrations are needed.

The effects of increasing the amount of gadolinium rods in the lattice on

gadolinium worth are demonstrated in figure 4-4. The gadolinium rod placement

geometry was held constant while 8 to 18 rods were placed in the lattice. As the amount

of gadolinium rods placed in the lattice increased, the degree of gadolinium clumping

decreased due to the size limitations of the lattice. In the increase of 15 gadolinium rods

to 16 gadolinium rods, a clumped geometry was utilized that resulted in a decrease in

gadolinium worth. Each gadolinium rod insertion for the hot condition was worth -

0.0103 Ak/k while each gadolinium rod insertion in the CU660 case was worth -0.0095











Ak/k leading to 0.5 mAk/k difference in gadolinium worth between the hot and cold


lattice states.





0-


-0 05


2 -0 1


0
5 -0 15


o -02 -


-0 25


-0 3
8 9 10 11 12 13 14 15 16 17 18
Number of Rods

-4-HU Gad Worth -U-CUO Gad Worth CU660 Gad Worth CU935 Gad Worth -*-CC Gad Worth

Figure 4-4. The effects of increased number of gadolinium rods on the gadolinium worth
at 0 GWD/STU.

Increasing the amount of gadolinium rods decreased hot and cold k.o by increasing

the amount of neutrons removed from the system by absorption. HUCU### also

decreased as the number of gadolinium rods increased. Figure 4-5 displays BOL

decrease in HUCU660 as a function of increasing the amount of gadolinium rods placed

in the lattice.


The thermal-to-fast flux ratio is higher in the cold state than in the hot state due to


the cold states increased moderator density, and the thermal-to-fast flux ratio is lower in

higher boron concentration due to decreased thermal neutron availability after boron


capture. Gadolinium is dominantly a thermal neutron absorber therefore in the increased











thermal-to-fast flux ratio gadolinium was a more effective absorber. The decreased

thermal-to-fast flux in the hot state as compared to the cold borated states results in a

decrease in HUCU### as each gadolinium rod was inserted because the gadolinium was

worth more per rod insertion in the cold state than in the hot state.

Gadolinium rod worth was also a function of the boron concentration utilized in the

cold condition. At 0 GWD/STU HUCU660 changes -0.0017 per rod insertion (10-13

ppm boron equivalence) while HUCU935 changes -0.0025 per rod insertion (16-20 ppm

boron equivalence). This demonstrates that when a utility decides to go to a power up-

rate or increased exposure cycle, the increased amount of gadolinium needed to offset the

increased installed reactivity will result in a decrease in the HUCU### parameter on the

lattice level resulting in a decrease in SLCS margin on the core wide level.




012



01



0 08 -

y= -0.0025x+ 0.1197
S-- HUCU660
0 06 -


0 04 -
y = -0.0017x + 0.0703


002



0 2 4 6 8 10 12 14 16 18 20
Number of Gadolinium Rods

Figure 4-5. The effects of the number of gadolinium rods inserted on HUCU at 0
GWD/STU.











The Effects of Increasing the Gadolinium Concentration on HUCU###

The effects of increasing gadolinium concentration of a given gadolinium


configuration on HUCU### was next examined to determine if gadolinium concentration


was a design constraint for HUCU###. Increasing the gadolinium concentration of the


lattice had similar results to increasing the amount of rods in the lattice. Figure 4-6


displays the increase in gadolinium worth as a function of increasing concentration. For


CU660 at 0 GWD/STU a 1% increase in gadolinium concentration for 14 gadolinium


rods is worth -0.002343 Ak/k. For HU at 0 GWD/STU a 1% increase in gadolinium


concentration for 14 gadolinium rods is worth -0.004587 Ak/k.




0 -


-005 -


-0 1 -


-015-


-0 2 -


-0 25


-0 3
0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
Gadolinium Concentration
-*-HU Gad Worth ---CUO Gad Worth CU660 Gad Worth CU935 Gad Worth -*-CC Gad Worth

Figure 4-6. The effect of increasing the gadolinium concentration for 14 gadolinium rods
on gadolinium worth at 0 GWD/STU.

The HU state exhibited 2 times greater worth in increasing 1% in gadolinium worth


than the CU state; therefore increasing the gadolinium concentration will decrease


HUCU###. Figure 4-7 exhibits the decrease in HUCU### as the gadolinium







51


concentration is increased. For a 1% Change In Concentration for 14 Gadolinium Rods


at 0 GWD/STU, HU660 changes -0.004504 (33 ppm boron equivalent) and HU935


changes -0.004906 (36 ppm boron equivalent).





0 12


01


0 08 -
y = -0.4906x + 0.1198

HUCU660
( 0 6 -
I--- HUCU935

0 04 -
y = -0.4504x + 0.0795

002


0
0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
Gadolinium Concentration

Figure 4-7. The effect of increasing the gadolinium concentration of 14 gadolinium rods
on HUCU at 0 GWD/STU

The Importance of Gadolinium Rod Location

The lattice power distribution is never uniform. The effectiveness of gadolinium to


suppress power while increasing HUCU### was highly dependent upon the location in


the lattice in which the gadolinium was placed. Since many parameters contributed to the


power distribution within the lattice, determining an optimum location for gadolinium

placement resulted from satisfying all the parameters that were most limiting. The power


peaking distributions for the HU, CU### and CC states differed therefore determining an


optimum location for placing gadolinium involved placement in areas that maximized










improvement to the most limiting state without violating the parameter requirements of

the other states.

Two gadolinium rods were placed in series of different locations throughout the

lattice to determine the areas in which HUCU### and HUCC could be maximized (two

gadoliniums rods were used in order to preserve mirror symmetry of the lattice design).

Figure 4-8 presents the locations examined and corresponding case numbers of the

gadolinium location tests.


1 2 3 4 5 6 7 8 9 10

1

2 11

3 13

4 W W

5 W W

6 W W 18 20

7 W W19

8 11 18 21 23

9 13 20 23

10

Figure 4-8. Gadolinium rod placement diagram.

The effectiveness of the gadolinium placement was a function of the power

distribution. The power distribution was a function of the moderation ability and poison

geometry. In the CC state as gadolinium rods were placed further from the edges of the

control blade the worth of the gadolinium increased. This was due to the decreased

competition for neutrons between the gadolinium and the control blade. If the









gadolinium was placed to close to the control blade then effectively the gadolinium and

blade were spatially self-shielding each other and therefore decreased both of the

poisons' effective worth. In the CU### case, the difference in worth of a certain

gadolinium rod location was not a function of boron poison geometry (assuming no

clumping) because the boron poison geometry was uniform; however, the difference in

worth of certain gadolinium rod locations was related to the moderation capability of the

fuel lattice. Areas of the lattice exhibiting more moderation created more thermal

neutrons, leading to a higher power peaking. Gadolinium was worth more in these areas

of increased moderation capability due to the increased amount of thermal neutrons

available for absorption.

Figure 4-9 displays the difference in gadolinium worth as a function of gadolinium

location corresponding to the patterns in figure 4-8. Certain pattern changes caused

opposing worth differences in different temperature and boron geometry states. In the

change from pattern 4 to pattern 5 and in the change from pattern 15 to pattern 16, CC

gadolinium worth increased while CU660 and HU gadolinium worth decreased. In the

change from pattern 10 to pattern 11, CC gadolinium worth greatly decreased while

CU660 worth slightly decreased and HU worth slightly increased.

Placing a gadolinium rod in a higher power peaked area resulted in up to a 5%

increase in gadolinium worth. In the CC case, increasing the distance of the gadolinium

rod from the center of the control blade increased gadolinium worth by 0.00525 until the

water rods in the center of the lattice were reached. Once the water rods were reached in

the CC lattice (on the lower right diagonal half of the lattice), the power distribution and

gadolinium rod worth become independent of the effects of the control blade.







54



Altering a current lattice design to improve SDM and SLCS while maintaining HU


must include shifting gadolinium locations where both the CU### and CC states have the


most significant worth improvement while HU only has slight gadolinium worth increase.


Improving CC gadolinium worth involves placing the gadolinium away from the control


blade so that the gadolinium does not compete for neutrons with the boron in the control


blade for. Enhancing the CU### worth involves spacing out the gadolinium so that


spatial self-shielding does not occur and placing the gadolinium in the areas of highest


power peaking exhibited by the cold power shape (areas of greatest moderation


capabilities). Enhancing the HU worth also involves spacing out gadolinium and placing


them in areas of highest power peaking corresponding to the HU power shape. The


optimum gadolinium pattern for any given amount of gadolinium rods is the design that


meets all three of these criteria.





0

-0 005

-0 01

-0 015

S-002

-0 025

3 -003

-0 035

-0 04

-0 045

-0 05
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Pattern

-*-HU Gad Worth ---CUO Gad Worth CU660 Gad Worth CC Gad Worth

Figure 4-9. Gadolinium worth versus location for 0 GWD/STU, 7% gadolinium
concentration,









Fuel Lattice Design Conclusions

Certain parameters in the 2-dimensional fuel lattice design have a significant

contribution to HUCU###. Lattice enrichment may be utilized to enhance HUCU###.

While local lattice enrichment perturbations do not contribute to HUCU###, decreasing

the lattice average enrichment decreases the power distribution skewing of the lattice and

therefore increases HUCU###. However, with the demand for increased cycle lengths at

higher powers, unfortunately higher average enrichments are needed to meet these

requirements. Therefore with the needed increased average enrichment of the bundles

will result in an increase in power skewing of the lattice thus decreasing HUCU###.

The addition of competing thermal neutron poisons decreases HUCU###. Fuel

designs with a tighter pitch between fuel rods lead to an increased production of

plutonium. Plutonium is a competing thermal neutron poison. Therefore creating smaller

fuel rods with a tighter pitch may increase the heat transfer of the fuel bundle, but also wa

greater amount of plutonium is generated and therefore HUCU### is compromised. Also

utilization of mixed oxide fuels (MOX) introduces an increased plutonium inventory in

the core therefore further decreasing HUCU###.

Gadolinium is also a thermal neutron poison. Enough positive reactivity must be

installed into the reactor core at the beginning of cycle in order to meet the cycle length

requirements. Gadolinium must installed in each of the fuel bundles in order to make

sure that with the installed reactivity the reactor core is critical throughout operation.

Therefore though gadolinium competes for thermal neutrons thereby decreasing

HUCU###, it is a necessary component of reactor operation.

In order to enhance HUCU### while maintaining a cycle operation goal, only

certain fuel lattice parameters may be varied. Cycle length and power level is dependent









upon installed reactivity; therefore average enrichment of the fuel is a fixed parameter if

the number of fresh bundles utilized in the design is fixed. Plutonium production is

related to power level, fuel lattice pitch, and isotope content of the fuel. In most cases all

of those are fixed.

The only design parameter with room for enhancement is gadolinium. If

gadolinium is utilized effectively in certain locations of the fuel bundle, maximized

differences between the HU and CU### may be created; therefore HUCU### is

improved leading to an improvement in SLCS on the full core level. However, in

increased average enrichment cores more gadolinium rods are needed to meet critical

eigenvalue requirements; therefore resulting in fewer locations to manipulate gadolinium

rod placement for improvement in HUCU###. Therefore if gadolinium has already been

placed in the areas of maximized HUCU### enhancement further techniques must be

utilized on the full core level to enhance HUCU###.














CHAPTER 5
FULL CORE SLCS MODELING

The reference base reactor core analyzed was a generic BWR/3. The reactor core

was quarter core symmetric meaning that only a quarter of the full core had to be

modeled to accurately represent characteristics of the full core. Figure 5-1 was the base

reference core in which all perturbations were compared. Two different average bundle

enrichments of fresh fuel were loaded into the core for the investigated cycle. The high

enrichment bundles (fuel type 19) were 4.18% enriched, and the low enriched bundles

(fuel type 20) were 3.89% enriched.

Three major types of perturbations utilized to enhance SLC S were investigated on

the full core level. The perturbations were selected based on the knowledge generated

from the lattice physics analysis, and fell into two distinct characteristic types. The first

type involved making a perturbation to the entire bundle. Gadolinium rod placement

perturbations to the entire bundle were examined in order to determine the maximum

achievable enhancement to SLCS. The second type involved perturbing the axial power

shape. Based on the fact that average enrichment was also a dominant parameter in

enhancing HUCU### in the lattice physics calculations, axial power shaping techniques

utilizing enrichment differencing in certain axial zones was next examined to determine

the maximum achievable enhancement to SLCS by perturbing the cold axial power

shape. Gadolinium insertion in certain axial zones was also examined to determine if this

method was also effective in enhancing SLCS by perturbing the axial power shape

utilizing a poison.







58





33899 33557 32011 31230 29071
















7 Bundle Exposure6 6 6 6
33219 30031 26946 25807 26128 25187
6 6 6 7 7 7
33611 32365 30011 30073

33451 28541 29470 27398 27574

33992 33626 25332 25539

33476 33533
6 6Twice
34030 28585 25349

32422 29444 25517
6 7 6
29955 27372
7 7
33233 30085 27593

33890 30081
7 6
33527 26959
6 6
31928 25804
6 7
31253 26087
6 7
29057 25179
6 7






Bundle Exposure



Bundle Type Twice
Burned
Bundle


Figure 5-1. Reference base core fuel bundle loading map.

In each perturbation case the critical eigenvalue, thermal limits, and SDM were

monitored in order to determine if the enhancement to SLCS would violate the

requirements of these margins. Calculated eigenvalue was monitored for each case to


determine if calculated eigenvalue varied more than 0.001 Ak from the base case critical









eigenvalue. At BOC rods patterns may always be adjusted in order to make the reactor

critical and have critical eigenvalue deviate less than 0.001 Ak; however, at EOC when

all the rods are pulled out of the core and no other form of positive reactivity may exist in

the core to supply reactivity for criticality any decrease in critical eigenvalue as compared

from the base case resulted in loss of cycle exposure and decreases of cycle energy.

MFLPD was observed to make sure no perturbation resulted in a MFLPD greater than

0.909 as the BWR design basis requires. MFLCPR was also monitored to be certain that

no perturbation caused a MFLCPR to become greater than the design basis requirements

of 0.930.

A SLCS enhancement that leads to decreased cycle energy and results in loss of

cycle exposure was unacceptable due to the $ 1,000,000 at day cost involved in shutting

the reactor down early. Furthermore, a core that does not meet thermal limits may not be

licensed; therefore though SLCS may be improved through a certain modification, if that

modification leads to unacceptable thermal margin, the core will not be licensed to

operate. The optimum enhancement for SLCS involves an enhancement that meets

thermal limits and does not deplete EOC calculated eigenvalue.

Enhancing SLCS by Perturbing the Location of Gadolinium Rods

A gadolinium placement modification to certain lattice axial zones was made in

each fresh fuel type separately. The lattice physics calculations ensured that HUCU###

improved with enhanced gadolinium location loading; therefore applying the technique to

each fuel type individually determined the limiting effects of core radial and axial power

weighting incurred on the enhancement of SLCS. The gadolinium rods that were

interchanged were chosen based on the fact that the cold power peaking map displayed an










increased worth for the new rod locations while the hot power peaking displayed a lesser

change in worth. Figure 5-2 displays the base DOM gadolinium geometry and the areas

circled in red correspond to where the gadolinium pins were interchanged with normal

fuel pins in the perturbed cases. These perturbations were made to each axial zone

individually and then to the entire bundle for each fresh bundle type.

A B C D E F G H I J
1 1.60 2.80 3.20 3.95 3.95 3.95 3.95 3.95 3.95 2.80

2 2.80 2.80 3.20 3.95 3.60 3.95 3.95 3.95
S 0 -10
3 3.20 3.20 4 4.40 .0 4.40 0, 4.40 4.90
SOu __ IO uu IF. CI0 00
4 3.95 3.95 4.40 '3.95 WR 4.90 4.90 4.90 %U235
4 90 Enrichment
5 3.95 3.60 6 0 3.95 4.90 4.90 4.90 4.90

6 3.95 5 4.40 WR 4.9 4.90 4.90 0 4.90
800 -00 .
7 3.95 3.95 0 4.90 4.90 4.90 4.90 % Gadolinium
":'04
8 3.95 4.40 4.90 4.90 4.90 4.90 4.90
4 ------40 --40 0--40
9 3.95 3.95 4.90 4.90 : 4.90 c 4.90 4.90
3 8.0oo --oo
10 2.80 3.95 4.90 4.90 4.90 4.90 4.90 4.90 4.90 3.20


Figure 5-2. The perturbation diagram for the gadolinium rod perturbation cases.

SLCS margin was improved by the greatest amount when every axial zone

containing gadolinium was perturbed. Figure 5-3 displays the maximum SLCS

enhancement exhibited by each bundle type perturbation. Case 1 and Case 2 represent

perturbations made to every axial zone in the bundle containing gadolinium. Case 1

represents when the perturbation was only made to the high enrichment bundles, and

Case 2 represents when the perturbation was only made to the low enrichment bundles.

Case 1 exhibited a 0.0095 improvement in BOC SLCS margin while Case 2

exhibited a 0.0341 increase in BOC SLCS margin. The low enrichment bundles

represented 69% of the total loaded batch fraction and the majority of these bundles











resided in the high power peak locations in the interior core region; therefore any


perturbations made to these bundles demonstrated a more pronounced enhancement than


perturbations made to the high enrichment bundles.




004

0 035

0 03

0 025

002,

0015

001

0005

0
0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
-*-Base ----Case 1 -- Case 2 -*-Most Limiting Base SLCS


Figure 5-3. Exposure dependent SLCS for the reference base case, the case in which the
perturbation was made to only all of the fresh low enrichment bundles (case
2), and the case in which the perturbation was made to only all of the fresh
high enrichment bundles (case 1).

Though SLCS margin was enhanced, other limiting factors were greatly affected.


Table 5-1 displays the effects of the enrichment perturbation on eigenvalue, and


highlighted in red are the points in which eigenvalue deviated more than 0.001 Ak from


the critical eigenvalue. All exposure points ending in an A represented the exposure step


in which the control blade was shifted into the next pattern configuration. Due to the


slight increase in gadolinium utilization caused by the perturbation, BOC eigenvalue


decreased; and due to the saved positive reactivity from BOC, mid-cycle eigenvalue


increased. In order to increase BOC eigenvalue and decrease mid-cycle eigenvalue the










control blade patterns were manipulated to offset this reactivity imbalance. Table 5-2

displays the exposure dependent MFLPD. Case 2 improved most limiting MFLPD below

the most limiting base case MFLPD after the rod pattern adjustment. Table 5-3 displays

the exposure dependent MFLCPR. The most limiting MFLCPR in case 2 was also

improved below the base case after the rod pattern adjustment. Therefore after the

control blade adjustments were made both cases could meet critical eigenvalue

requirements, but only case 2 could also meet MFLPD requirements as well.

Table 5-1. Critical eigenvalue at specified exposure points for the gadolinium rod
location perturbation cases that exhibited greatest enhancement in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Critical Eigenvalue


Base
1.0136
1.0142
1.0141
1.0128
1.0133
1.0122
1.0128
1.0118
1.0126
1.0119
1.013
1.012
1.0134
1.0123
1.0134
1.0135
1.0148
1.0147
1.0149
1.0163
1.0157
1.0159
1.0169
1.0159
1.0164
1.0163


Case 1
1.0123
1.0129
1.0129
1.0117
1.0123
1.0112
1.0119
1.0111
1.012
1.0113
1.0126
1.0117
1.0133
1.0122
1.0133
1.0135
1.0149
1.0148
1.0153
1.0168
1.0164
1.0168
1.0178
1.0168
1.0174
1.0173


Case 1 Fix
1.013
1.0139
1.0139
1.0125
1.0131
1.0112
1.0119
1.0111
1.0121
1.0114
1.0126
1.0118
1.0134
1.0123
1.0134
1.0136
1.0149
1.0148
1.0152
1.0167
1.0162
1.0165
1.0175
1.0165
1.0171
1.017


Case 2
1.0104
1.011
1.0111
1.0101
1.0108
1.0095
1.0104
1.0098
1.0108
1.0103
1.0116
1.0108
1.0126
1.0115
1.0129
1.0134
1.0153
1.0152
1.0163
1.0182
1.0179
1.0186
1.0195
1.0188
1.0195
1.0193


Case 2 Fix
1.0129
1.0146
1.0138
1.0124
1.0132
1.0126
1.0136
1.012
1.0134
1.012
1.0135
1.0125
1.0135
1.0123
1.0137
1.0138
1.0152
1.0152
1.0155
1.0169
1.0164
1.0167
1.0176
1.0167
1.0171
1.0169










Table 5-2. Exposure dependent MFLPD for the gadolinium rod location perturbation
cases that exhibited greatest enhancement in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Base
0.846
0.845
0.806
0.807
0.781
0.742
0.714
0.717
0.703
0.838
0.849
0.847
0.889
0.847
0.852
0.784
0.683
0.72
0.704
0.704
0.704
0.854
0.734
0.841
0.746
0.763


Case 1
0.829
0.85
0.811
0.814
0.789
0.753
0.725
0.726
0.711
0.842
0.857
0.865
0.917
0.865
0.863
0.786
0.686
0.724
0.713
0.709
0.708
0.855
0.742
0.84
0.748
0.766


MFLPD
Case 1 Fix
0.815
0.832
0.796
0.803
0.778
0.754
0.724
0.726
0.71
0.84
0.853
0.859
0.912
0.859
0.863
0.791
0.685
0.729
0.711
0.706
0.705
0.851
0.738
0.834
0.741
0.759


Utilizing any type of gadolinium insertion will only enhance BOC SLCS. Though

BOC SLCS was enhanced in case 2 by gadolinium location improvement, once the

gadolinium burned out (-11,000 MWD/STU for this specific case) the perturbed case

became limiting again in SLCS. Therefore if SLCS is enhanced utilizing this method, the

designer must consider if the gadolinium burn out point is acceptable as well. If at the

gadolinium burn out point SLCS is not acceptable in magnitude, another method must be


utilized to enhance SLCS.


Case 2
0.866
0.858
0.82
0.821
0.795
0.753
0.729
0.734
0.72
0.858
0.878
0.884
0.923
0.892
0.887
0.806
0.699
0.717
0.718
0.73
0.736
0.878
0.762
0.871
0.789
0.805


Case 2 Fix
0.821
0.799
0.779
0.786
0.763
0.712
0.712
0.702
0.682
0.822
0.824
0.825
0.886
0.841
0.878
0.836
0.723
0.755
0.706
0.693
0.691
0.852
0.737
0.837
0.736
0.753










Table 5-3. Exposure dependent MFLCPR for the gadolinium rod
cases that exhibited greatest enhancement in SLCS.


location perturbation


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


MFLCPR


Base
0.711
0.721
0.726
0.737
0.741
0.73
0.736
0.729
0.734
0.773
0.775
0.731
0.732
0.745
0.746
0.754
0.767
0.757
0.79
0.824
0.825
0.829
0.825
0.822
0.822
0.808


Case 1
0.723
0.729
0.733
0.744
0.748
0.737
0.743
0.734
0.738
0.77
0.774
0.733
0.732
0.744
0.743
0.752
0.766
0.756
0.793
0.828
0.83
0.836
0.833
0.83
0.818
0.803


Case 1 Fix
0.727
0.727
0.733
0.741
0.744
0.737
0.743
0.734
0.738
0.77
0.774
0.734
0.732
0.744
0.743
0.751
0.764
0.754
0.792
0.828
0.829
0.834
0.831
0.828
0.814
0.8


Case 2
0.707
0.713
0.717
0.731
0.734
0.717
0.725
0.72
0.726
0.778
0.78
0.734
0.731
0.744
0.748
0.758
0.773
0.766
0.789
0.838
0.836
0.841
0.827
0.836
0.84
0.825


Case 2 Fix
0.714
0.715
0.72
0.721
0.727
0.744
0.754
0.736
0.744
0.767
0.769
0.735
0.737
0.75
0.749
0.752
0.763
0.755
0.785
0.817
0.818
0.821
0.817
0.815
0.816
0.802


Enhancing SLCS at the cost of SDM was not an acceptable option if SDM was

already a limiting constraint from the original design. Figure 5-4 demonstrates that the

modifications made to the low enriched fresh bundles did not diminish SDM below the

most limiting value of the base case. For this perturbation, BOC SDM was improved

0.0190 at the cost of decreasing EOC SDM; however, the decrease in EOC SDM did not

fall below the most limiting SDM value, and therefore the perturbation yielded


acceptable SDM consequence.














0 04

0 035

0 03

0 025

002

0015

001

0 005

0
0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
-*-Base --Case 1 -- Case 2 ---Most Limiting Base SDM

Figure 5-4. Exposure dependent SDM for the gadolinium rod location perturbation cases.

Utilizing a gadolinium location perturbation may be utilized to enhance BOC


SLCS. However, the magnitude of the improvement is limited by the ability to separate


the gadolinium and move it into areas of greater effective worth. As the amount of


gadolinium rods in the lattice increases, the ability to move gadolinium rods to more


effective locations decreases. As fuel bundle designs move to higher average


enrichments more gadolinium rods are needed in the fuel bundles to counteract the


increased installed reactivity. More gadolinium rods in the fuel bundle causes this


technique to be less effective due to the inability to move the gadolinium to locations of


greater effective worth due to the space constraints of the fuel lattice.


Axial Power Shape Characteristics

The base case most limiting power peak bundle location was bundle (15, 11).


Because lattice physics work determined that the DOM was the most limiting geometry


for HUCU###, decreasing the cold power peak in the DOM decreases SLCS.











The cold power shape and the hot power shape were not the same. Figure 5-5 is

the most limiting radial power peaking base case axial power distribution relative to its

radial power peaking. Superimposed in blue over figure 5-5 is an example of the cold

power shape. Though not to scale, the superimposition displays the difference in where

the power peaking resides in the two conditions. The BOC cold power shape was a

cosine shape peaked in the DOM, and the hot power shape was a modified Bessel

function peaked in the PSZ. Therefore in the axial power shape perturbations this

characteristic was utilized to maximize SLCS.


160 o,---






112 2 04 0 0 1 1 1 1
96
-4-Base (15,11)
0 ___________________ _______ -(15,10)
o -(14411)
(15, 12)
Calculated Hot Power Shape







0 02 04 06 08 1 12 14 16 18
Relative Power Peak

Figure 5-5. The base case hot axial power shape with superimposed cold axial power
shape.

Enhancing SLCS through Axial Power Shaping Utilizing Enrichment Differencing

The position of the cold axial power peak is the axial portion of the fuel bundle

exhibiting the least amount of power suppression in the cold condition; therefore the cold

axial power peak region is also the most limiting HUCU### region. As previously

displayed in figure 3-2, the DOM was the most limiting region for HUCU### due to the

decreased availability of borated water locations in the lattice. The VAN exhibited the










greatest HUCU### due to the increased availability to place borated water in the lattice

as a result of the vanished rod locations. Since the region of the cold axial power peak

was the most limiting in HUCU###, shifting the cold axial power peak out of the DOM

and into the VAN should increase SLCS.

The cold axial power peak may be decreased utilizing enrichment differencing in

the DOM and VAN. Figure 5-6 illustrates the goal of enrichment differencing. By

increasing the enrichment in the VAN and decreasing the enrichment in the DOM, the

DOM axial power peak is decreased thereby increasing SLCS.




N-T
N-V








PLE
--Decreased Power Peak in DOM Zone

Base Case Power Peak








Base Case

Perturbed Case Utilizing Axial Enrichment Differencing
NAT
Figure 5-6. Cold axial power shape perturbation diagram.

Axial power shaping, utilizing enrichment differencing, enhanced SLCS at BOC

and at the gadolinium burn out exposure point. Figure 5-7 displays the SLCS











enhancement utilizing axial power shape perturbation by enrichment differencing as well


as SLCS enhancement utilizing gadolinium placement perturbations. Case 11 was the


case that utilized enrichment differencing. Gadolinium perturbations enhanced SLCS


only at BOC; however, axial power shape perturbations utilizing enrichment differencing


in the DOM and VAN enhanced SLCS at both BOC and the gadolinium burn out


exposure point. The BOC SLCS margin enhancement utilizing enrichment differencing


was 0.035 and the gadolinium bum out exposure point enhancement was 0.0142.


Therefore if improvement in SLCS is necessary in both BOC and gadolinium burn out


exposure point the enrichment differencing method is the preferred method.




004

0 035

003

0025

0 02

0015

001

0005

0
0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
-*-Base ---Case 2 -A-Case 11 -*-Most Limiting Base SLCS

Figure 5-7. Exposure dependent SLCS for the gadolinium location perturbation case
(case 2) and axial power shape perturbation utilizing enrichment differencing
case (case 11) that exhibited the greatest enhancement in SLCS.

In the gadolinium perturbation case, the fresh bundles reached a maximum peak


power point once the gadolinium burned out. Once the gadolinium had burned out, the


main parameters affecting the cold power shape were the enrichment distribution and











axial leakage of the bundle. Since the axial leakage of the fuel bundle was a function of

axial height (a fixed parameter) and controlled utilizing top and bottom natural zones,

axial enrichment distribution was the main mode for altering the power shape at the

gadolinium burn out exposure point. Increasing the enrichment distribution in the VAN

and decreasing the enrichment in the DOM resulted in a decreased DOM cold power

peak at the gadolinium burn out exposure point due to the decreased availability of

enrichment in the DOM.

Figure 5-8 presents the SLCS enhancement as a function of enrichment difference

between the DOM and VAN. The maximum amount of SLCS margin enhancement for

BOC and the gadolinium burn out exposure point occurs at 0.30% enrichment difference

between the DOM and VAN.




004


0 035


003


0025


L 002


0015


001


0005 --Base (001) ---Case 17 (0 35) Case 11 (0 31) Case 18 (0 20)
---Case 19 (0 10) -- Case 20 (0 04) --Most Limiting Base SLCS

0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
Figure 5-8. SLCS enhancement utilizing different magnitudes of enrichment differencing
between the DOM and VAN.







70


An optimum enrichment difference arises from the fact that reactivity worth is a


flux weighted. As the enrichment was increased in the VAN, the flux increased in the


VAN; and as the enrichment decreases in the DOM, the flux decreases in the DOM.


Therefore the 0.30 enrichment difference represented the optimum decrease in flux


weighting of the DOM and increase in flux weighting of the VAN that resulted in the


greatest average cold borated negative reactivity insertion.


Axial power shaping utilizing enrichment differencing alters the hot power shape


and mode in which the core burns. The BOC hot axial power profile utilizing enrichment


differencing is displayed in Figure 5-9. By increasing the VAN zone enrichment while


decreasing the DOM zone enrichment, the DOM and PSZ zones exhibited a decrease in


power peak while the VAN zone experiences and increase in power peaking.


144 'r- Increased Ftw Peak In VAN Zone

128- -

112 ------




80^-- -------------------------------









16
64-


Decreaeed RFtwer Feak In DOMand PSZZones



32-

oI


Figure


-- se (15,11)
-U--(15,10)
(14,11)
(15,12)


0 02 04 06 08 1 12 14 16 18
Relative PaerPeak

-e 5-9. The hot axial power shape for maximum SLCS enhancement utilizing
enrichment differencing.









The change in axial power shape slightly altered the calculated eigenvalue and

thermal margins. Table 5-4 demonstrates that critical eigenvalue of the enrichment

perturbations case did not vary more than 0.001 k from the base case critical eigenvalue;

therefore utilizing this method did not warrant a rod pattern adjustment. The final

calculated eigenvalue for the enrichment differencing case was 0.0007 k less than the

critical base case eigenvalue; however, the increased mid cycle energy created could have

been suppressed by utilizing a rod pattern adjustment if determined necessary.

Table 5-4. Critical eigenvalue at specified exposure points for the gadolinium rod
location perturbation case and enrichment differencing case that exhibited
greatest enhancement in SLCS.
Exposure Critical Eigenvalue
(MWD/STU) Base Case 2 Fix Case 11
0 1.0136 1.0129 1.0142
181 1.0142 1.0146 1.0143
907 1.0141 1.0138 1.0143
1814 1.0128 1.0124 1.0128
2722 1.0133 1.0132 1.0133
2722A 1.0122 1.0126 1.013
3629 1.0128 1.0136 1.0136
4536 1.0118 1.012 1.0125
5443 1.0126 1.0134 1.0134
5443A 1.0119 1.012 1.0121
6350 1.013 1.0135 1.0133
7258 1.012 1.0125 1.0127
8165 1.0134 1.0135 1.0142
8165A 1.0123 1.0123 1.013
9072 1.0134 1.0137 1.0141
9979 1.0135 1.0138 1.014
10886 1.0148 1.0152 1.0153
10886A 1.0147 1.0152 1.0152
11794 1.0149 1.0155 1.015
12570 1.0163 1.0169 1.0162
12701 1.0157 1.0164 1.0156
13608 1.0159 1.0167 1.0153
13608A 1.0169 1.0176 1.0163
14061 1.0159 1.0167 1.0151
14334 1.0164 1.0171 1.0157
14570 1.0163 1.0169 1.0156









MFLPD for the axial power shaping utilizing enrichment differencing case also

was under the acceptable limit for all exposure points through out the cycle. Table 5-5

displays that most limiting MFLPD decreased by 0.026 from the base case. Therefore

utilizing this technique improves the MFLPD of the cycle.

Table 5-5. Exposure dependent MFLPD for the gadolinium rod location perturbation
case and the enrichment differencing case that exhibited greatest enhancement
in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Base
0.846
0.845
0.806
0.807
0.781
0.742
0.714
0.717
0.703
0.838
0.849
0.847
0.889
0.847
0.852
0.784
0.683
0.72
0.704
0.704
0.704
0.854
0.734
0.841
0.746
0.763


MFLPD
Case 2 Fix
0.821
0.799
0.779
0.786
0.763
0.712
0.712
0.702
0.682
0.822
0.824
0.825
0.886
0.841
0.878
0.836
0.723
0.755
0.706
0.693
0.691
0.852
0.737
0.837
0.736
0.753


Case 11
0.825
0.834
0.792
0.799
0.771
0.718
0.7
0.699
0.681
0.827
0.83
0.819
0.863
0.818
0.839
0.793
0.69
0.735
0.707
0.692
0.691
0.829
0.72
0.805
0.718
0.728


Table 5-6 shows no significant difference realized in MFLCPR as compared with

the base case. Therefore utilizing this technique does not deplete the cores to meet any


thermal margin requirements.










Table 5-6. Exposure dependent MFLCPR for the gadolinium rod location perturbation
cases that exhibited greatest enhancement in SLCS.


Exposl
(MWD/5


ure MFLCPR
STU) Base Case 2 Fix Case 11
0.711 0.714 0.724


0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


0.721
0.726
0.737
0.741
0.73
0.736
0.729
0.734
0.773
0.775
0.731
0.732
0.745
0.746
0.754
0.767
0.757
0.79
0.824
0.825
0.829
0.825
0.822
0.822
0.808


0.715
0.72
0.721
0.727
0.744
0.754
0.736
0.744
0.767
0.769
0.735
0.737
0.75
0.749
0.752
0.763
0.755
0.785
0.817
0.818
0.821
0.817
0.815
0.816
0.802


The effect of axial enrichment differencing had a similar effect on SDM as the

lattice geometric placement perturbation. Figure 5-10 displays the exposure dependence

effects of these perturbations on SDM at different exposure points. SDM for both

perturbations did not fall below the most limiting base case value; therefore both

enhancements may be utilized to enhance SLCS ifBOC SDM were to be in a limiting

condition. However, only axial power shaping utilizing enrichment differencing also

improves the gadolinium burn out exposure point limiting condition.


0.719
0.725
0.734
0.741
0.742
0.75
0.737
0.744
0.763
0.767
0.741
0.741
0.753
0.751
0.756
0.763
0.753
0.787
0.821
0.821
0.825
0.82
0.818
0.821
0.807














004

0 035

0 03

0 025

S002

0015

001

0005

0
0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
--Base ---Case 2 -A-Case 11 -+-Most Limiting Base SDM

Figure 5-10. Exposure dependent SDM for the gadolinium rod location perturbation case
and axial power shaping utilizing enrichment differencing case.

Enhancing SLCS by Means of Axial Power Shaping Utilizing Gadolinium Insertion

Enrichment differencing was not the only method for perturbing the axial power


shape in order to decrease the power peak in the DOM. Axial power shaping from the


utilization of an additional gadolinium pellets in certain axial zones was also analyzed in


order to determine if the negative reactivity insertion from adding additional gadolinium


was more favorable than shifting the axial enrichment distribution to create similar types


of perturbations. The concept of utilizing the negative reactivity of a gadolinium rod


insertion to decrease the power peak in the most limiting zone was similar to the concept


of decreasing enrichment in the most limiting axial zone. In both cases a negative


reactivity insertion in the limiting axial zone caused the power to peak to decrease in that


zone in which the gadolinium was inserted thus decreasing the worth of that axial zone to


SLCS.











As displayed in figure 4-5 increasing the amount of rods in a lattice decreased

HUCU###; however, ko of the lattice also decreased therefore leading to an improvement

of SLCS due to the decreased cold ko. Figure 5-11 displays the gain in BOC SLCS

utilizing a gadolinium rod insertion as compared with the other types of perturbations

examined. Case 26 represented a gadolinium insertion made in the PSZ, and Case 27

represented a gadolinium rod insertion made in the DOM. Because the cold power shape

peaks in the DOM, inserting a gadolinium rod into the PSZ does not have as drastic of an

effect on SLCS as placing a gadolinium rod in the DOM.




004

0 035

0 03

0025



0015

001

0005


0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
--Base ----Case 2 -A-Case 11 --Case 26 Case 27 --- Most Limiting Base SLCS

Figure 5-11. Exposure dependent SLCS for the gadolinium location perturbation case
(case 2), axial power shape perturbation utilizing enrichment differencing case
(case 11), inserting a gadolinium rod in the PSZ (case 26) and inserting a
gadolinium rod in DOM (case27) that exhibited the greatest enhancement in
SLCS.

The DOM gadolinium rod insertion yielded the greatest increase in BOC SLCS as

compared with the other perturbations. Inserting a gadolinium rod in the DOM enhanced

SLCS by 0.422 while inserting a gadolinium in the PSZ only enhanced SLCS by 0.189.










Table 5-7. Critical eigenvalue at specified exposure points for the gadolinium insertion
into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that
exhibited greatest enhancement in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Critical Eigenvalue


Base
1.0136
1.0142
1.0141
1.0128
1.0133
1.0122
1.0128
1.0118
1.0126
1.0119
1.013
1.012
1.0134
1.0123
1.0134
1.0135
1.0148
1.0147
1.0149
1.0163
1.0157
1.0159
1.0169
1.0159
1.0164
1.0163


Case 26
1.0127
1.0129
1.0131
1.012
1.013
1.0121
1.0129
1.0121
1.013
1.0122
1.0133
1.0124
1.0139
1.0128
1.0137
1.0135
1.0146
1.0145
1.0143
1.0155
1.0148
1.0147
1.0156
1.0146
1.0151
1.015


Case 26f
1.0133
1.0136
1.0138
1.012
1.013
1.0121
1.0129
1.0121
1.013
1.0122
1.0133
1.0125
1.014
1.0128
1.0137
1.0135
1.0145
1.0145
1.0142
1.0161
1.0154
1.0151
1.0162
1.0149
1.0148
1.0146


Case 27
1.0116
1.012
1.0121
1.0111
1.0121
1.0112
1.0122
1.0116
1.0126
1.0118
1.0129
1.0119
1.0133
1.0122
1.0132
1.0134
1.0149
1.0148
1.0151
1.0165
1.0159
1.016
1.017
1.016
1.0167
1.0165


Placing negative reactivity into one region of the bundle without introducing

positive reactivity into some other region will cause a decrease in BOC eigenvalue.

Therefore a rod pattern change was utilized in this method in order to achieve acceptable

BOC eigenvalue requirements. Table 5-7 displays the calculated eigenvalue as compared

with the base case for inserting a gadolinium rod in the DOM zone and in the PSZ before.

Inserting a gadolinium rod in the PSZ greatly reduced the BOC reactivity of the

bundle; therefore the rod patterns had to be adjusted to meet the BOC condition. The


Case 27f
1.013
1.0139
1.0141
1.0121
1.0131
1.0119
1.013
1.0117
1.0128
1.0119
1.0131
1.0121
1.0136
1.0124
1.0134
1.0135
1.0147
1.0147
1.0147
1.016
1.0154
1.0154
1.0164
1.0153
1.016
1.0158











decrease in integrated power realized in the bundle due to the gadolinium insertion in the

PSZ caused the core to fall 0.0017 k short of EOC critical eigenvalue requirements.

However, inserting a gadolinium rod in the DOM zone caused only a 0.0005 k decrease

in EOC critical eigenvalue.

Distorting the hot axial power shape altered the thermal margins of the core.

Figure 5-12 and figure 5-13 displays the distorted hot axial power shape caused from

inserting a gadolinium rod in the PSZ and in the DOM.


160





128 "


112

S96 __
-+-Base (15,11)
0
S8_ --(15,10)
o (14,11)
1, (15,12)
S64


48


32----


16 -_



0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Relative Power Peak

Figure 5-12. The hot axial power shape of the most power peaked fuel bundle caused by
inserting a gadolinium rod into the PSZ.

When inserting a gadolinium rod into the PSZ, the decreased power peak in the

PSZ causes the hot axial power shape to flatten. When inserting a gadolinium rod into







78


the DOM, the extreme decreased power peak in the DOM leads to an increased relative

power peak in the PSZ.


160 ---


14 4 -------


128 "


112


---96 /
o= --Base (15,11)
S80Y --(15,10)
0o (14,11)
(15,12)
64 ---


4 8 ----------------_-----


3 2 ----------------------_





0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Relative Power Peak

Figure 5-13. The hot axial power shape of the most power peaked fuel bundle caused by
inserting a gadolinium rod into the DOM.

Table 5-8 displays MFLPD for the PSZ and DOM gadolinium insertions as

compared to the base case. Inserting a gadolinium rod in the PSZ decreased BOC

MFLPD by 0.083, and also decreased most limiting MFLPD by 0.033. However,

inserting a gadolinium rod in the DOM yielded no significant enhancement in MFLPD.

In both cases there was no significant alteration in MFLCPR as displayed in table 5-9.

Therefore inserting an extra gadolinium rod into a certain axial zone of the fuel bundle

does not hinder the ability to meet thermal margins.









Table 5-8. MFLPD at specified exposure points for the gadolinium insertion into the
DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited
greatest enhancement in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Base
0.846
0.845
0.806
0.807
0.781
0.742
0.714
0.717
0.703
0.838
0.849
0.847
0.889
0.847
0.852
0.784
0.683
0.72
0.704
0.704
0.704
0.854
0.734
0.841
0.746
0.763


Case 26
0.771
0.772
0.748
0.768
0.765
0.707
0.710
0.722
0.698
0.827
0.827
0.816
0.861
0.807
0.841
0.812
0.710
0.757
0.716
0.677
0.677
0.838
0.703
0.821
0.719
0.736


MFLPD
Case 26f
0.763
0.762
0.738
0.768
0.765
0.706
0.709
0.722
0.697
0.825
0.824
0.813
0.856
0.804
0.839
0.815
0.713
0.761
0.722
0.684
0.684
0.866
0.720
0.859
0.710
0.727


Case 27
0.873
0.875
0.832
0.829
0.790
0.746
0.716
0.718
0.703
0.840
0.855
0.853
0.894
0.854
0.849
0.776
0.687
0.715
0.712
0.714
0.709
0.841
0.749
0.822
0.744
0.755


Case 27f
0.847
0.839
0.795
0.811
0.776
0.738
0.707
0.718
0.700
0.836
0.845
0.840
0.882
0.840
0.850
0.789
0.688
0.727
0.711
0.705
0.702
0.831
0.741
0.807
0.737
0.749


Therefore when a


designer chooses to utilize this method for enhancing SLCS, the


designer must decide which parameters are most necessary for achieving the required

result. If the designer is experiencing limiting MFLPD and willing to compromise cycle

energy to meet this requirement, then placing a gadolinium rod in the PSZ is the better

choice. If the designer does not have limiting MFLPD, then inserting a gadolinium rod in

the DOM zone is the better choice. Therefore the choice of one method or the other

depends on the thermal margins and the critical eigenvalue requirements.










Table 5-9. MFLCPR at specified exposure points for the gadolinium insertion into the
DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited
greatest enhancement in SLCS.


Exposure
(MWD/STU)
0
181
907
1814
2722
2722A
3629
4536
5443
5443A
6350
7258
8165
8165A
9072
9979
10886
10886A
11794
12570
12701
13608
13608A
14061
14334
14570


Base
0.711
0.721
0.726
0.737
0.741
0.73
0.736
0.729
0.734
0.773
0.775
0.731
0.732
0.745
0.746
0.754
0.767
0.757
0.79
0.824
0.825
0.829
0.825
0.822
0.822
0.808


Case 26
0.712
0.708
0.717
0.728
0.737
0.724
0.733
0.729
0.738
0.772
0.775
0.741
0.741
0.752
0.748
0.751
0.758
0.748
0.786
0.819
0.819
0.824
0.818
0.817
0.81
0.796


MFLCPR
Case 26f
0.715
0.712
0.721
0.728
0.737
0.724
0.733
0.729
0.738
0.772
0.775
0.742
0.742
0.752
0.748
0.75
0.757
0.747
0.786
0.818
0.818
0.822
0.815
0.814
0.805
0.791


SDM was neither greatly enhanced nor greatly decreased utilizing gadolinium

insertion. Figure 5-14 displays SDM as a function of exposure for the base case and all

three types of perturbations. Therefore if a designer was limited in EOC SDM then

inserting a gadolinium rod should be utilized in order to enhance BOC SDM.

The decision to enhance SLCS margin by inserting a gadolinium rod into a certain

axial zone of a fuel bundle is dependent upon the preexisting limiting conditions of the

fuel design. If the fuel designer decides that maximizing BOC SLCS without concern for


Case 27
0.694
0.683
0.692
0.711
0.724
0.708
0.723
0.723
0.732
0.771
0.774
0.73
0.73
0.742
0.745
0.754
0.768
0.758
0.79
0.824
0.825
0.829
0.825
0.822
0.824
0.809


Case 27f
0.692
0.683
0.695
0.709
0.723
0.717
0.732
0.724
0.734
0.77
0.773
0.733
0.733
0.744
0.744
0.751
0.763
0.753
0.789
0.823
0.823
0.826
0.822
0.819
0.815
0.8











SLCS at the gadolinium bum out exposure point is the most limiting design


characteristic, then inserting a gadolinium rod into the DOM will suffice as a solution to


enhancing BOC SLCS margin. If the designer cannot afford loss in EOC SDM, and only


needs a minimal improvement in thermal margins as well as minimally enhanced BOC


SLCS, then adding a gadolinium in PSZ at the cost of cycle energy may be an adequate


solution to enhancing BOC SLCS.




004

0 035

003

0 025

002

0015

001 -

0005

0
0 2000 4000 6000 8000 10000 12000 14000 16000
MWDISTU
--Base -E-Case 2 --&Case 11 ---Case 26 -- Case 27 -I-Most Limiting Base SDM

Figure 5-14. Exposure dependent SDM for the gadolinium rod location perturbation
case, axial power shaping utilizing enrichment differencing case and the axial
power shaping utilizing gadolinium placement case.

Enhancing BOC SLCS margin utilizing gadolinium perturbations may cause the


gadolinium burn out exposure point to become the most limiting in SLCS. If SLCS at the


gadolinium burn out exposure point is of acceptable magnitude, then the designer has


utilized an acceptable technique for enhancing SLCS. If, however, SLCS at the


gadolinium burn out exposure point is of unacceptable magnitude, then the gadolinium


insertion techniques are not feasible methods for improving SLCS. Therefore the






82


decision to utilize this method is solely dependent upon the limitations of the gadolinium

burn out exposure point.














CHAPTER 6
CONCLUSIONS

SLCS is a core wide phenomenon that is dependent upon the HUCU###

characteristics of each fuel bundle. Introducing fresh bundles into the core with

inherently enhanced HUCU### characteristics will improve SLCS. HUCU### is

improved by manipulating design parameters on the lattice design level as well as in the

full core design. Therefore understanding the most limiting design parameters in both

design aspects and the capability of those parameters to increase HUCU### is paramount

to improving SLCS.

The ability of a certain type of fuel lattice design perturbation to enhance SLCS

was determined by the limiting characteristics of that lattice perturbation. HUCU###

was highly dependent upon average enrichment. As average enrichment of the fuel

bundle was increased HUCU### decreased thereby decreasing SLCS on the full core

level. Increasing enrichment has a greater impact per percent increase of reactivity in the

cold, collapsed void lattice state then in the hot Doppler broadened voided operating

state. Localized enrichment perturbations did not affect HUCU### therefore when a

designer creates a lattice with SLCS in mind they need only be concerned with the

average enrichment of the lattice and not how the local enrichment is schemed.

Gadolinium rods also had a significant impact on HUCU### lattice behavior and

therefore significantly impacted SLCS. Gadolinium geometries that were clumped and

incurred significant spatial self-shielding decreased HUCU### while gadolinium

geometries that were spread out limiting the self-shielding exhibited an increased









HUCU###. Increasing the amount of gadolinium rods in the fuel lattice decreased

relative HUCU###; however, inserting the gadolinium also reduced k- thereby actually

improving SLCS by decreasing the worth of the bundle to the entire core. Therefore

increasing the amount of gadolinium rods in the bundle had a diminishing return.

Increasing the gadolinium concentration also decreased the relative HUCU###; however,

increasing the gadolinium concentration also reduced k- thereby also improving SLCS

for the whole core. Optimum locations for gadolinium rod placement exist for certain

amounts of gadolinium rods. These optimum placement locations are realized by

understanding the difference in power peaking between the hot and cold homogenously

enriched power shapes (the power shape realized explicitly from geometry of the fuel

bundle and flux level) and placing gadolinium rods in areas where the difference in

power peak between the two states is the greatest.

After understanding the 2-dimensional lattice physics calculations, perturbations

were made to fuel bundles in the full core simulator in order to determine effects on full

core criticality and thermal limits. Perturbing the placement of the gadolinium rods in

order to maximize gadolinium worth utilized in the cold borated condition improved

BOC SLCS at the expense of decreased BOC critical eigenvalue. Therefore after

perturbing gadolinium locations to maximize negative reactivity, the control blade

patterns must be adjusted in order to introduce enough positive reactivity in the hot

condition to meet the critical eigenvalue requirements. Perturbing the axial enrichment

distribution in order to decrease the power peaking in the axial zone most limiting to

HUCU### decreased that axial zones flux importance to the SLCS calculation and

thereby improved both the BOC and the gadolinium burn out exposure point SLCS.









However, utilizing this method causes an increased complexity in manufacturing of the

bundle and therefore leading to an increased production cost. Inserting an extra

gadolinium rod into a certain axial zone in order to also perturb the axial power shape

improved BOC SLCS without decreasing EOC SDM. However, utilizing gadolinium

perturbations only helped improve the BOC SLCS and did not enhance the gadolinium

burn out exposure point SLCS.

SLCS may always be improved by increasing the boron concentration or boron

enrichment in the SLCS tank. However, if the utility is limited by time, cost or

aggravation then utilizing an acceptable design technique in order to enhance SLCS

margin is solely dependent upon the limiting characteristics of the core behavior and the

acceptable sacrifice in margin of those parameters.

Unfortunately, not all core situations will have a possible remedy for SLCS. The

greatest increase in BOC SLCS utilizing any of the mentioned techniques was roughly

0.5% and the gadolinium burnout point maximum improvement was 0.14%. Therefore

utilities exhibiting marginal SLCS fuel design difficulties that wish to have power output

increases in their following cycles, increasing the average enrichment and gadolinium

content in their core, will need to understand the limitations of the inherent fuel design.

Utilities must then realize that an increase in boron concentration of their SLCS tank or

utilizing enriched boron is needed if they wish to accommodate SLCS while not incurring

the extra cost per cycle of loading extra bundles to flatten the power distribution and

reduce SLCS.














CHAPTER 7
FUTURE WORK

The purpose of this study was to conduct a sensitivity analysis in order to

determine limiting fuel design characteristics for SLCS. The methodology developed by

Yasushi Hirano, Kazuki Hida, Koichi Sakurada and Munenari Yamamoto utilized a fixed

gadolinium pattern and then generated an optimal enrichment distributions for a 2-

dimensional BWR fuel lattice [10]. Since this study proved that radial enrichment

distribution was not a factor in SLCS and that SLCS was only limited by average

enrichment, the possibility exists to expand on the enrichment distribution tool and

develop a tool that determines an optimum SLCS gadolinium placement for a given

lattice average enrichment. The tool would basically compare homogenously enriched

hot and cold lattice power distributions and determine an optimum gadolinium scheme

based on the maximum difference in the two power distributions. Because the placement

of the gadolinium for SLCS is basically decoupled from the enrichment distribution, this

problem does not become over-constrained, and therefore it is possible to obtain an

optimum gadolinium configuration for SLCS while creating an optimum enrichment

distribution for thermal limit and fuel efficiency requirements.

The optimum enrichment distribution methodology was also a 2-dimensional

methodology. This study concluded that axial enrichment and gadolinium perturbations

may be utilized to improve SLCS. In modem core design strategy 2-dimensional lattice

calculations are completed and then the group constants from the 2-dimensional codes are

utilized by the full core simulators because of computational time constraints and









memory requirements of the processor. With computers getting faster and distributed

parallel computing schemes becoming more optimized, core design may reach a point

where full 3-dimensional bundles are modeled assuming an infinite bundle approximation

(or a more brilliant scheme) to get group constants for the full core simulator. When this

technology is available, utilizing the design criteria from this study for the SLCS portion,

a full bundle axial and radial enrichment and gadolinium configuration optimization

methodology may be devised that creates the optimum fuel bundle for SLCS, SDM,

thermal margin and fuel utilization. This will create an automated core design

environment thus freeing the designer's time to allow for examination of other pressing

issues in the design strategy.
















LIST OF REFERENCES


1. Aoyama, Mooto, Sadao Uchikawa and Renzo Takeda, "Reactivity Control Method
for Extended Burnup of Boiling Water Reactor Fuel Bundles," Journal of Nuclear
Science and Technology, 26, pp.403-410, April 1989.

2. Cochran, Robert and Nicholas Tsoulfanidis, The Nuclear Fuel Cycle: Analysis and
Management, American Nuclear Society, La Grange Park, Illinois, 1999.

3. Dresner, Lawerance, Resonance Absorption in Nuclear Reactors, Pergamon Press,
New York, New York, 1976.

4. Duderstadt, James and Louis Hamilton, Nuclear Reactor Analysis, John Wiley &
Sons, Inc., New York, New York, 1976

5. General Electric Company, "TGBLA06A; General Electric Lattice Physics
Method," DRF A00-05526, October, 1994. (Proprietary Information)

6. General Physics Corporation, "BWR Generic Fundamentals: Chapter 9 Core
Thermal Limits," Columbia, Maryland 1993. (Proprietary Information)

7. Glasstone, Samuel and Walter H. Jordan, Nuclear Power and its Environmental
Effects, American Nuclear Society, La Grange Park, Illinois, 1980.

8. Global Nuclear Fuels, "PANAC11 User's Manual," UM-0021 Rev. 1, February
2001. (Proprietary Information)

9. Hida, Kazuki and Ritsuo Yoshioka, "Optimal Axial Enrichment Distribution of the
Boiling Water Reactor Fuel Under the Haling Strategy," Nuclear Technology, 80,
pp. 423-430, March 1988.

10. Hirano, Yasushi, Kazuki Hida, Koichi Sakurada, and Munenari Yamamoto,
"Optimization of Fuel Rod Enrichment Distribution to Minimize Rod Power
Peaking throughout Life with BWR Fuel Assembly," Journal of Nuclear Science
and Technology, 34, pp. 5-12, January 1997.

11. Kazimi, Mujid and Neil Todreas, Nuclear Systems 1: Thermal Hydraulic
Fundamentals, Taylor and Francis, Bristol, PA, 1993

12. Lahey, R.T. and F.J. Moody, The Thermal Hydraulics of a Boiling Water Nuclear
Reactor, American Nuclear Society, La Grange Park, Illinois, 1979.




Full Text

PAGE 1

OPTIMUM BOILING WATER REACTOR FUEL DESIGN STRATEGIES TO ENHANCE REACTOR SHUTDOWN BY THE STANDBY LIQUID CONTROL SYSTEM By MICHAEL LORNE FENSIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2004

PAGE 2

Copyright 2004 by Michael Lorne Fensin

PAGE 3

This document is dedicated to the memory of my late grandmothers Bernice Anker and Edna Fensin.

PAGE 4

ACKNOWLEDGMENTS I would like to acknowledge Dr. Samim Anghaie for chairing my committee, supplying a connection to Global Nuclear Fuels of America, and providing excellent tutoring and advice as my graduate advisor. I would also like to thank Dr. Bob Coldwell Dr. Edward Dugan, Dr. Alireza Haghighat, Dr. David Hintenlang, Dr. Travis Knight, Dr. Alan Jacobs, Dr. Tim Olson, Dr. Benard Mair, Proffessor Jim Tulenko and Dr. William Vernetson for providing me countless hours of instruction in all areas of nuclear engineering and mathematical computation during my graduate studies. I would like to acknowledge Global Nuclear Fuels of America for the sponsorship of its computer codes, time and efforts. From Global Nuclear Fuels of America I would specifically like to thank Dr. Mehdi Asgari, Kenneth Gardner, Roland Jackson, J.D. Kavaal, Thomas Marcille, V.W. Mills, Dr. Brian Moore and Tony Reese for supplying intriguing knowledge and guidance during the course of the study. I want to thank my family for being a constant source of support and pushing me to completion. Without their support none of this would have been possible. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT.......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 The Boiling Water Reactor System..............................................................................4 Boiling Water Reactor History.....................................................................................8 History of Fuel Bundle Development.........................................................................10 The SLCS Event.........................................................................................................12 Project Scope..............................................................................................................13 2 MODEL AND METHODOLOGIES.........................................................................15 Standby Liquid Control System and Shutdown Margin.............................................15 Modeling Tools...........................................................................................................17 TGBLA 6.............................................................................................................17 PANAC11............................................................................................................18 Utilized Temperature States, Boron Concentrations and Lattice Types.....................19 Measurement of SLCS and SDM during the Lattice Development Stage.................20 Fuel Bundle Geometry................................................................................................21 Thermal Limit Design Considerations........................................................................24 3 MAXIMIZING HOT-COLD BORATED k DIFFERENCE UTILIZING ENRICHMENT..........................................................................................................28 Homogeneous Enrichment Distribution.....................................................................28 Determining the Most Limiting Lattice Axial Zone and Void Concentration....29 Understanding the Exposure Dependent HUCU### Curve................................31 Enrichment and Boron Concentration Effects.....................................................34 Power Peaking Distribution.................................................................................35 v

PAGE 6

Heterogeneous Enrichment Distribution....................................................................39 Localized Enrichment Perturbation.....................................................................39 Gross Enrichment Perturbation...........................................................................41 4 MAXIMIZING HOT-COLD BORATED k DIFFERENCE UTILIZING GADOLINIUM..........................................................................................................44 Spatial Self-Shielding Effects of Gadolinium Rods on HUCU###............................44 The Effects of Increasing the Amount of Gadolinium Rods on HUCU###...............47 The Effects of Increasing the Gadolinium Concentration on HUCU###...................50 The Importance of Gadolinium Rod Location............................................................51 Fuel Lattice Design Conclusions................................................................................55 5 FULL CORE SLCS MODELING..............................................................................57 Enhancing SLCS by Perturbing the Location of Gadolinium Rods...........................59 Axial Power Shape Characteristics.............................................................................65 Enhancing SLCS through Axial Power Shaping Utilizing Enrichment Differencing...........................................................................................................66 Enhancing SLCS by Means of Axial Power Shaping Utilizing Gadolinium Insertion.................................................................................................................74 6 CONCLUSIONS........................................................................................................83 7 FUTURE WORK........................................................................................................86 LIST OF REFERENCES...................................................................................................88 BIOGRAPHICAL SKETCH.............................................................................................90 vi

PAGE 7

LIST OF TABLES Table page 3-1 Gross enrichment perturbation scheme for figure 3-10............................................41 5-1 Critical eigenvalue at specified exposure points for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS............................62 5-2 Exposure dependent MFLPD for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS..........................................................63 5-3 Exposure dependent MFLCPR for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS..........................................................64 5-4 Critical eigenvalue at specified exposure points for the gadolinium rod location perturbation case and enrichment differencing case that exhibited greatest enhancement in SLCS..............................................................................................71 5-5 Exposure dependent MFLPD for the gadolinium rod location perturbation case and the enrichment differencing case that exhibited greatest enhancement in SLCS....72 5-6 Exposure dependent MFLCPR for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS..........................................................73 5-7 Critical eigenvalue at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS.................................................................................76 5-8 MFLPD at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS..............................................................................................79 5-9 MFLCPR at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS..............................................................................................80 vii

PAGE 8

LIST OF FIGURES Figure page 1-1 BWR pressure vessel system......................................................................................5 1-2 A typical BWR fuel assembly and fuel rod................................................................6 1-3 A four fuel assembly group with cruciform control blade.........................................7 2-1 A cross sectional view of the modeled fuel bundle..................................................23 2-2 The geometric setup of the fuel lattice axial zones..................................................24 3-1 Exposure dependent HUCU660 for the DOM at 3.95% enrichment.......................29 3-2 Exposure dependent HUCU660 at varied void fraction and axial zone for the C lattice at an enrichment of 3.95%.............................................................................31 3-3 Exposure dependant HUCU### curve.....................................................................33 3-4 Beginning of cycled HUCU vs. enrichment in the DOM, at 40% void fraction, for a C lattice............................................................................................................34 3-5 The power peaking distributions at 5 GWD/STU, 3.95% enrichment, DOM, C lattice, and 40% void fraction vs. temperature state and boron concentration.........35 3-6 The power peaking distribution at 5 GWD/STU, CU660, DOM, C lattice, and 40% void fraction versus enrichment.......................................................................38 3-7 The power peaking distributions at 5 GWD/STU, HU, DOM, C lattice, and 40% void fraction vs. enrichment.....................................................................................38 3-8 Localized enrichment perturbation map...................................................................40 3-9 Exposure dependent HUCU660 for different localized enrichment perturbation patterns.....................................................................................................................40 3-10 An example of a gross enrichment perturbation map...............................................41 3-11 Exposure dependent HUCU660 at 40% void fraction, in the DOM, with a C lattice........................................................................................................................42 viii

PAGE 9

3-12 Exposure dependent HUCC at 40% void fraction, in the DOM, with a C lattice....42 4-1 Four sample clumped geometries.............................................................................45 4-2 Corresponding 0 GWD/STU gadolinium worth for the patterns displayed in figure 4-1..................................................................................................................46 4-3 Corresponding exposure dependent gadolinium clumping effects on HUCU660 for the patterns displayed in figure 4-1....................................................................46 4-4 The effects of increased number of gadolinium rods on the gadolinium worth at 0 GWD/STU.............................................................................................................48 4-5 The effects of the number of gadolinium rods inserted on HUCU at 0 GWD/STU.............................................................................................................49 4-6 The effect of increasing the gadolinium concentration for 14 gadolinium rods on gadolinium worth at 0 GWD/STU...........................................................................50 4-7 The effect of increasing the gadolinium concentration of 14 gadolinium rods on HUCU at 0 GWD/STU............................................................................................51 4-8 Gadolinium rod placement diagram.........................................................................52 4-9 Gadolinium worth versus location for 0 GWD/STU, 7% gadolinium concentration,...........................................................................................................54 5-1 Reference base core fuel bundle loading map..........................................................58 5-2 The perturbation diagram for the gadolinium rod perturbation cases......................60 5-3 Exposure dependent SLCS for the reference base case, the case in which the perturbation was made to only all of the fresh low enrichment bundles (case 2), and the case in which the perturbation was made to only all of the fresh high enrichment bundles (case 1).....................................................................................61 5-4 Exposure dependent SDM for the gadolinium rod location perturbation cases.......65 5-5 The base case hot axial power shape with superimposed cold axial power shape...66 5-6 Cold axial power shape perturbation diagram..........................................................67 5-7 Exposure dependent SLCS for the gadolinium location perturbation case (case 2) and axial power shape perturbation utilizing enrichment differencing case (case 11) that exhibited the greatest enhancement in SLCS.............................68 5-8 SLCS enhancement utilizing different magnitudes of enrichment differencing between the DOM and VAN....................................................................................69 ix

PAGE 10

5-9 The hot axial power shape for maximum SLCS enhancement utilizing enrichment differencing...........................................................................................70 5-10 Exposure dependent SDM for the gadolinium rod location perturbation case and axial power shaping utilizing enrichment differencing case....................................74 5-11 Exposure dependent SLCS for the gadolinium location perturbation case (case 2), axial power shape perturbation utilizing enrichment differencing case (case 11), inserting a gadolinium rod in the PSZ (case 26) and inserting a gadolinium rod in DOM (case27) that exhibited the greatest enhancement in SLCS........................................................................................................................75 5-12 The hot axial power shape of the most power peaked fuel bundle caused by inserting a gadolinium rod into the PSZ...................................................................77 5-13 The hot axial power shape of the most power peaked fuel bundle caused by inserting a gadolinium rod into the DOM................................................................78 5-14 Exposure dependent SDM for the gadolinium rod location perturbation case, axial power shaping utilizing enrichment differencing case and the axial power shaping utilizing gadolinium placement case...........................................................81 x

PAGE 11

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering OPTIMUM BOILING WATER REACTOR FUEL DESIGN STRATEGIES TO ENHANCE REACTOR SHUTDOWN BY THE STANDBY LIQUID CONTROL SYSTEM By Michael Lorne Fensin August 2004 Chair: Samim Anghaie Major Department: Nuclear and Radiological Engineering Licensing a commercial nuclear reactor core involves a stringent amount of calculations that demonstrate the capability of safe reactor shutdown in the instance of an emergency transient event. In a boiling water nuclear reactor the control blades and standby liquid control system are the two independent redundant safety systems utilized for shutting down the reactor. In past fuel designs lower power cores with smaller cycle lengths resulting from lower core average enrichment caused shutdown by the control blades to be the most limiting strategy of the two modes for reactor shutdown. This led to a lesser focus on fuel design strategies for standby liquid control system margin (SLCS). Advanced modern core designs involve higher powers and increased cycle lengths resulting from higher core average enrichments, therefore causing SLCS to now become a more significant shutdown parameter. This study characterized the most limiting fuel design parameters for maximizing the margin for safe reactor shutdown utilizing SLCS while maintaining the demanded cycle energy requirements. xi

PAGE 12

This study examined perturbation effects of certain parameters in the fuel lattice development stage and the 3-dimesional reactor core simulator modeling. Lattice enrichment perturbation response was examined first to determine the effects of average and local enrichment perturbations on SLCS. Lattice gadolinium perturbation response was next investigated to determine the optimum concentration, number of rods, location and degree of clump of gadolinium rods necessary for enhancing SLCS. Axial power shape perturbation utilizing both gadolinium and/or enrichment differencing in certain axial zones of the fuel bundle was then examined on the full core level to determine the optimum strategy for maximizing SLCS. This study concluded that the necessary strategy for maximizing SLCS depended upon the exposure point at which SLCS was most limiting. Certain perturbations utilizing gadolinium exhibited maximized beginning of cycle SLCS; however, these strategies involved a modified operating strategy to meet the beginning of cycle operational requirements while not maximizing the limiting end of cycle SLCS. Axial power shaping utilizing enrichment differencing maximized end of cycle SLCS and increased beginning of cycle SLCS margin but to a lesser magnitude than the gadolinium perturbation. In all cases the amount of improvement to the margin was limited by a maximum value. Therefore if the desired magnitude of improvement needed is within the achievable limits of the examined techniques, the choice of optimum strategy for enhancing SLCS to a desired value depends upon the magnitude of necessary improvement at the most limiting exposure points. xii

PAGE 13

CHAPTER 1 INTRODUCTION Boiling water nuclear reactor cores are major sources of revenue for power producing utilities. If the utility is able to maximize the amount of energy output from the nuclear reactor while minimizing the cost of the reactor operation then the utility will realize an increase in profit. A utility may choose to maximize the energy output from their nuclear reactors in one of three ways. Either the utility may increase the operating cycle length of the reactor thereby increasing the amount of energy per cycle and increasing the amount of time between refueling outages, or the utility may choose to increase the power level of operation thereby increasing the amount of available distributable energy at any given time, or the utility may choose to utilize a combination of both practices [2]. In either of the operational techniques the utility must increase the installed reactivity in the reactor core in order to meet the desired goal. Increasing the amount of installed reactivity in a given cycle as compared with a previous cycle in order to aggressively improve reactor power output is termed aggressive core loading. Aggressive core loading strategies involve higher fuel batch fractions with fuel bundles of higher average enrichment in order to increase the installed positive reactivity in the reactor core [2]. Greatly increasing the core power output in the internal core locations greatly increases the neutron flux and thus greatly increases the exposures of the interior bundles. A twice-burned fuel bundle in the reactor core is at its least reactive state in bundle life and therefore in ordinary core loadings twice-burned fuel bundles can be an excellent power suppressor utilized to flatten the power 1

PAGE 14

2 distribution in the internal core locations; however, in aggressive core loadings the high neutron flux in the interior of the core may causes twice-burned fuel bundles to exceed thermo-mechanically limited peak exposure. Therefore in aggressive core loadings after bundles have burned two cycles they must be moved to the core periphery in order to inhibit surpassing peak exposure [4]. This leads to only fresh and once burned assemblies loaded in the interior core locations. Most of the gadolinium in a once burned fuel bundle is completely burnt up at the end of the previous cycle therefore these bundles are at the most reactive state. Inability to suppress this energy output decreases the available margin to shutdown the core in an emergency situation. Due to the aggressive core loading geometry, the only available power suppression comes from the fresh fuel bundles that are loaded into the core. However, with increased energy placed in the fresh bundles by increasing average enrichment in order to meet the increased power demand, these fresh bundles will have decreased power suppression capabilities. Therefore with decreased power suppression capabilities, the reactor core becomes more limiting in emergency shutdown capabilities. One of the shutdown systems is the standby liquid control system and the margin in which the reactor core is shutdown utilizing this system is the standby liquid control system margin (SLCS). If the reactor core becomes more limiting in SLCS due to the decreased power suppression capabilities of the core loading strategy, it becomes paramount to then determine the optimum bundle design utilized in order to improve SLCS in an aggressive core loading environment. A reactor core is never licensed without being able to meet all the necessary shutdown criteria, thermal limit characteristics, and cycle length requirements. Therefore

PAGE 15

3 operating reactor cores do not encounter a failure to meet SLCS because the reactor would not be allowed to operate if the reactor could not meet the SLCS requirements. Inability to meet SLCS with a certain reactor core fuel bundle configuration is realized and mitigated in the design phase of the reactor fuel cycle. The designer has many options to improve SLCS but any one option may endure a list of consequences some of which may result in extreme economic concern. The designer may request that the reactor cycle length be decreased; obviously if the utility wishes to increase profit by increasing power output this option is not acceptable. The designer may request the utility to increase the boron concentration or enrichment of B10 in the boron solution utilized by the standby liquid control system. However, this course of action is limited by increased aggravation caused from Nuclear Regulatory Commission (NRC) licensing, availability for the utility to plan for the change economically, ample time to complete the concentration increase in time for the next cycle loading, capabilities of the installed accumulator tank to support the concentration increase and ability to keep the boron soluble in solution. The designer may choose to increase the amount of fuel bundles loaded in the cycle and load more fuel bundles with a smaller enrichment; however, this may cost the utility more than what was budgeted and therefore is not a viable option. The action that is demanded by the utility is for the designer to create a core design that meets the utilities budget and does not increase the amount of bundles that are loaded into the reactor core. Therefore the designer must design a fuel bundle with inherently better SLCS characteristics. In order to accomplish this task efficiently the designer must

PAGE 16

4 know the set of limiting design parameters that may be utilized to enhance SLCS, and have a list of effective techniques that utilize the advantages of those parameters. The purpose of this study was to determine the fuel bundle design parameters that were most limiting in achieving the maximum possible enhancement for SLCS, and to determine the maximum amount of available improvement to SLCS by utilizing certain enhancement techniques that take advantage of those design parameters. This will demonstrate the feasibility of designing a fuel bundle and core operating strategy that has the ability to meet SLCS without incurring the costs of adding extra fuel bundles in the design or decreasing cycle power output requirements. The Boiling Water Reactor System The boiling water reactor (BWR) system is a nuclear system that boils water creating steam that is converted into power. The entire BWR system is composed of a reactor pressure vessel system, a turbine system, a generator system, a condenser system and the auxiliary control and heat removal systems. Steam is created by boiling water in the reactor pressure vessel system. The high quality steam then passes through a turbine, and causes the turbine shaft to rotate. The turbine shaft is connected to a generator and as the turbine shaft rotates the generator converts the mechanical energy of the rotating turbine shaft into electrical energy. Once the steam leaves the turbine system, it is sent to the condenser to be condensed into a sub-cooled fluid and pumped back into the reactor pressure vessel system. The reactor pressure vessel system is of primary concern to the nuclear reactor core designer. Figure 1-1 is an illustration of a typical BWR reactor pressure vessel system. The reactor pressure vessel system consists of the control drive mechanisms, the active

PAGE 17

5 fuel, the jet pumps, the moisture separators and steam driers as well as other inlets for emergency core cooling [16]. Figure 1-1. BWR pressure vessel system [16].

PAGE 18

6 The active reactor fuel length is approximately 12 feet though actual fuel length may vary according to different types of fuel assembly product lines. The fission power of the reactor converts the sub-cooled coolant into steam. The steam then travels through the steam separators and driers to create a high quality steam that is then send to the turbine. Each fuel assembly is composed of the fuel and water rods, intermediate spacer grids, upper and lower tie plates, and flow nozzle. Figure 1-2 displays a typical BWR fuel assembly as well as a typical fuel rod. Each fuel assembly is encased in a zircaloy fuel box in order to limit flow between adjacent fuel assemblies. This allows the flow to any given fuel assembly to be orificed to maintain a constant exit steam quality as well as limit instability in core thermal performance [16]. Figure 1-2. A typical BWR fuel assembly and fuel rod [16].

PAGE 19

7 Fuel assemblies are loaded into the reactor in groups of four with a cruciform B4C control blade loaded in the center of the grouped bundles. Displayed in figure 1-3 is a typical layout of four fuel bundles with a cruciform control rod loaded in the center of the grouped bundles. The fuel assemblies are diagonally symmetric with themselves and are loaded so that there exists 1/8 bundle symmetry with the four grouped bundles. Figure 1-3. A four fuel assembly group with cruciform control blade [16].

PAGE 20

8 By applying this loading scheme, symmetry may be utilized in modeling the fuel bundle thereby easing the computation time of the fuel assembly parameters [14]. Each BWR fuel assembly will contain fuel rods at certain enrichments that may vary radially and axially as well as gadolinium rods which may vary in placement and concentration radially and axially. The modeling of SLCS will first involve modeling a 2-dimensional slice of a single fuel assembly at certain axial heights and then modeling the entire 3-dimensional reactor utilizing the information from the 2-dimensional model. Boiling Water Reactor History The first two light water cooled nuclear reactor systems commercially available for power production were the boiling water reactor and the pressurized water reactor (PWR). The concept of the commercial PWR was created from the technology developed for submarines by the navy nuclear program [7]. BWR development occurred at Argonne National Laboratory and the Nuclear Energy Division of General Electric (GE) [9]. The PWR concept is generally characterized as a system in which the bulk coolant is sub-cooled and contains boron. The system utilizes an indirect dual-cycle that uses a steam generator. The BWR concept is generally characterized as a system that has boiling in the reactor core, with the bulk fluid containing no boron, utilizing a direct cycle for power conversion (the demonstration BWR/1 plants utilized a dual cycle). The first BWR experiments conducted at Argonne National Laboratory utilized the BORAX in 1953. BORAX-III produced steam-generated electricity for the town of ARCO, ID in 1955. The experimental boiling water reactor (EBWR) was developed in 1957 and ran until 1967. This reactor produced 100 MWt and was utilized to demonstrate the BWR concept for electricity generation utilizing a variety of fuel enrichments. The GE Valecitos boiling water reactor (VBWR) was the first commercial

PAGE 21

9 nuclear power plant to be licensed by the United States Atomic Energy Commission (USAEC). Utilized as an experimental reactor, VBWR examined BWR fuel cycle technology and determined the stable modes of operation. In 1955 Dresden-1 became the first commercial BWR specifically constructed for commercial power [12]. Dresden-1 was a dual cycle plant and fell under the category of BWR/1. BWR/1 designs were basically prototype designs of both dual and direct cycle utilized as demonstration plants that were custom made to meet individual utility specifications. Dual cycle BWR plants eventually fell out of favor because of the enormous capital cost involved in utilizing a steam generator. Ouster creak was the first attempt at standardizing the BWR and marked the beginning of the BWR/2. In 1963 BWR/2 plants were developed incorporating a direct steam cycle as the chosen method of power conversion. The reactor concept utilized internal steam separators and forced flow circulation that pumped core flow through 5 variable speed recirculation pumps. In 1965 GE introduced the BWR/3, the Dresden-2 design, which incorporated the use of internal jet pumps eliminating the need for external flow circulation loops. In 1966 the BWR/4 or Browns Ferry design was introduced. This design was similar to previous designs but incorporated a 20% increase in the core power density improving power producing capability of the reactor and thereby increasing its economic value. The year of 1969 marked the introduction of the Zimmer class of plants better known as the BWR/5. These plants utilized an improved emergency core cooling and recirculation system. Flow control in these reactors was accomplished through use of valve control rather than pump speed control allowing the plants to follow more rapid load change and decrease the capital cost of the control system [12]. The BWR/6 incorporated the used of higher

PAGE 22

10 efficiency steam separators and multi-hole jet pumps as well as improved power flattening through enhanced coolant distribution and burnable Gadolinia zone loading. The original BWR/6 fuel design incorporated an 8X8 fuel assembly rather than the previously utilized 7X7 [12]. The current BWR technologies include the Advanced Boiling Water Reactor (ABWR) and the European Simplified Boiling Water Reactor (ESBWR) both of these concepts utilized passive safety features in order to alleviate the need for complicated control systems. The history of the BWR is one that is based off of evolutionary concepts in order to increase core power and eliminate external moving parts that may fail during reactor operation. History of Fuel Bundle Development Over the years fuel bundle designs have undergone evolutionary changes in order to improve the thermal performance of the fuel and the reactivity features. Each fuel bundle design incorporated a key feature that allowed for increased margin in controlled conditions, transients, and thermal limits. The original fuel bundle concept utilized by the General Electric Company was the 7X7 fuel lattice design. These were fat fuel rods with decreased surface area due to the large size of the fuel rods. The decreased surface as compared to todays designs lead to an increase in the maximum linear heat generation rate (kW/ft) in the fuel bundle leading to an increase in fuel duty [12]. With the creation of the BWR/6 came the inclusion of the 8X8 fuel lattice assembly. Increasing the surface area of the fuel rods by decreasing there width and increasing the amount of fuel rods in the assembly decreased the maximum linear heat generation rate of the fuel rods. In 1988 a fuel design study by Motoo Aoyama, Sadao Uchikawa and Renzon Takeda suggested that extended exposure of fuel assemblies was possible if 9X9 fuel assemblies were utilized with and optimized internal water rod width

PAGE 23

11 thereby increasing the non boiling area inside the fuel lattice to increase moderation capability of the internal portions of the fuel lattice [1]. This design change increased fuel lattice efficiency. Going to a larger amount of fuel rods in the lattice at smaller fuel diameter decreased the linear heat generation of the fuel rod by increasing the surface area of the fuel. The current design utilized today by BWR vendors is the standard 10X10 fuel lattice design. Though the water rod locations vary from vendor to vendor the idea is the same; increase the moderation capabilities of the internal locations of the fuel lattice to boost reactivity in the areas that are most suppressed in power. Because the moderator density varies axially in the fuel assembly, the BWR has a distinct axial power shape. Kazuki Hida and Ritsuo Yoshioka determined that there were optimum axial enrichment distributions that minimized enrichment requirements subject to thermal margins [9]. They proved that increasing enrichment in the top half of the core actually decreases the uranium utilization. Therefore the interpretation of that study determined that fuel utilization was a key design constraint in creating optimum fuel bundles. In order to mainstream fuel designs industry moved away from axial enrichment shaping and fabricated fuel rods of a single enrichment and utilized part length control rods that increased the moderation capability in the top half of the fuel assembly leading to an increase in fuel efficiency. In 1997 Yasushi Hirano, Kazuki Hida, Koichi Sakurada and Munenari Yamamoto created an algorithm for determining optimum enrichment loading schemes for fuel lattices [10]. Holding the position and concentration of the gadolinium rods as the constant, the algorithm optimized the enrichments in the fuel lattice in order to meet

PAGE 24

12 certain thermal limit and local peaking factor criteria. The gadolinium configuration utilized for the study was based off of a configuration that was supposedly optimized only for shutdown margin (SDM) and certain thermal limit criteria based off of previous fuel design experience. However, this method lacked the ability to place the gadolinium and enrichment into the fuel lattice in such an optimum configuration such that all controls were satisfied. If the chosen gadolinium configuration was only optimum for SDM yet not also optimum for SLCS then this method was to only be successful in designing a lattice to meet SDM. What this method was lacking was the rules for understanding how the gadolinium needed to be configured to meet the SLCS, SDM and thermal limit configuration for a given average enrichment. The SLCS Event The standby liquid control system is initiated during anticipated transient without scram (ATWS). The following is a list of the events that occur in the SLCS event: 1. A transient even occurs in which it is necessary to SCRAM the reactor. 2. The reactor control blades fail to insert. 3. A calculated amount of steam is relieved from the reactor to the suppression pool at a rate that will not violate containment. 4. The boron solution is injected into the core at a specified rate and concentration in accordance with 10CFR50.62. 5. The reactor reaches an equilibrium shutdown condition. In 1984 the Nuclear Regulatory Commission (NRC) issued 10CFR50.62, Requirements for reduction of risk from anticipated transients without scam events for light water-cooled nuclear power plants (ATWS rule). The law states: Each boiling water reactor must have a standby liquid control system (SLCS) with the capability of injecting into the reactor pressure vessel a borated water solution at such a flow rate, level of boron concentration and boron-10 isotope enrichment,

PAGE 25

13 and accounting for reactor pressure vessel volume, that the resulting reactivity control is at least equivalent to that resulting from injection of 86 gallons per minute of 13 weight percent sodium pentaborate decahydrate solution at the natural boron-10 isotope abundance into a 251-inch inside diameter reactor pressure vessel for a given core design. [17] This is equivalent 660 ppm boron concentration in current reactor designs. The model utilized for determining if SLCS will satisfy the criteria mentioned in 10CFR50.62 is a steady state point after the transient event has occurred. The purpose of utilizing this method is to demonstrate that the reactor may be safely shutdown after the transient event has occurred. Therefore the event modeled is a cold core (160oC), borated to an acceptable concentration that causes the reactor to be sub-critical by a specified amount. Project Scope This study was conducted at Global Nuclear Fuels in Wilimington, NC. The study included both lattice physics analysis and full core modeling in the 3-d core simulator. The lattice physics work was further subdivided into the enrichment phase and the gadolinium phase. For a reference BWR/3 the following projects were undertaken: 1. Enrichment Phase a. Analyze the effects of homogenous average enrichment perturbation on the ability to maximize k difference between the hot operating condition and cold borated condition. b. Determine the effect of localized heterogeneous enrichment perturbations on the ability to maximize k difference between the hot operating condition and cold borated condition. 2. Gadolinium Phase a. Ascertain the effects of gadolinium clumping on maximizing k difference between the hot operating condition and cold borated condition. b. Resolve the effects of increasing the amount of gadolinium rods on maximizing k difference between the hot operating condition and cold borated condition. c. Analyze the effects of gadolinium concentration on increasing k difference

PAGE 26

14 between the hot operating condition and cold borated condition. d. Determine a methodology for placing gadolinium rods in order to improve k difference between the hot operating condition and cold borated condition. 3. Full Core Modeling Phase a. Establish the maximum SLCS improvement utilizing an altered geometric gadolinium placement within freshly loaded fuel bundles. b. Determine the SLCS attainable from axial power shape perturbations utilizing enrichment differencing in certain axial zones of the fresh fuel bundles. c. Resolve the maximum SLCS gained from inserting extra gadolinium rods into the freshly loaded fuel bundles. d. Conclude the optimum design strategy for maximizing SLCS.

PAGE 27

CHAPTER 2 MODEL AND METHODOLOGIES In order to determine the necessary strategies for enhancing a margin of shutdown it is necessary to have a clear definition of that margin. A model and tools to analyze that model must then be selected that depict the physics of the problem as accurately as possible. After the designation of a model and utilized tools, the design parameters that are to be perturbed within the model must be determined. Finally, all other limiting parameters must be clearly defined so that it may be determined if the improvement to SLCS is feasible and will not cause the nuclear reactor to violate the thermo mechanical limits of the fuel. Standby Liquid Control System and Shutdown Margin The two shutdown parameters utilized for reactor core licensing are SLCS and SDM. SDM is a measure of the amount in which the reactor core is shutdown utilizing all of the control blades excluding the highest reactivity worth control blade. effCHBWEeffkkkSDM (2.1) CHBWE = controlled case highest worth blade excluded If keff = 1 then effkSDM 1 (2.2) Therefore SDM is the reactivity needed to make the system critical or conversely viewed as the amount of reactivity in which the system is shutdown utilizing all of the control blades except for the highest worth control blade. 15

PAGE 28

16 SLCS is a measure of the amount that the reactor core is shutdown utilizing a homogenously dispersed boron poison solution. effBeffkkkSLCS (2.3) B = borated keff If keff = 1 then BkSDM 1 (2.4) Similar to SDM, SLCS is the reactivity needed to make the system critical or conversely viewed as the amount of reactivity in which the system is shutdown utilizing a homogenously dispersed boron poison solution. Both parameters depict the amount in which the reactor is safely shutdown; however, the difference in geometry of the poison causes the physics involved in each shutdown process to be significantly different. Because shutdown margin calculations utilize a control blade, a heterogeneous poison located on the boundary of two sides of the fuel bundle, the power distribution of the SDM case is expected to be skewed radially across the fuel lattice with power peaking occurring in areas furthest from the control blade. SLCS calculations utilize an evenly dispersed boron poison solution; therefore the power distribution in the lattice is expected to follow the power distribution dictated by the actual geometry of the fuel lattice. The two different modes of control have two separate types of lattice reactivity responses. Due to this significant difference in reactivity response, it is possible that the ability to meet specified margin may be satisfied in one mode but not in the other. Understanding the reactivity response to each modes shutdown independently and then

PAGE 29

17 utilizing the commonalities in maximized shutdown in each of the two modes ultimately leads to a fuel design that has maximized margin in both cases. Modeling Tools TGBLA 6 TGBLA 6 is a static, multi-group, 2-dimensional, diffusion theory code with transport corrections factors that assumes infinite lattice behaviors. The steady state multi-group diffusion equation that is solved is [14]: )()()(1)()()()()(''''''rqrrkrrrrrDeggfgggggggggg (1.5) Because of the major differences in fuel bundle design, void fraction history, control blade history, enrichment distribution, gadolinium content and accumulated exposure, the fuel bundles nuclear characteristics in the core are very different both radially and axially. Neighboring fuel bundles also have an influence on the characteristics of the modeled fuel bundle; however, modeling the effects of these neighboring fuel bundles may be a daunting task because each neighboring fuel bundle in the core incurs nuclear characteristics that are unique from every other bundle. Therefore assumptions have to be made in order to be able to achieve an effective and timely approximation of the fuel bundles nuclear behavior [5]. TGBLA 6 makes key assumptions in order to accurately approximate a fuel bundles nuclear characteristics. Because fuel bundle designs may be varied axially in bundle geometry, gadolinium content, enrichment, and void concentration, the influence of axial conditions are considered of primary influence to the fuel bundles nuclear behavior. TGLBA 6 completes 2-dimensional lattice physics calculations at different exposure points for certain axial sections where there exists a known major variation in

PAGE 30

18 fuel bundle geometry. Because in certain defined axial zones the void concentration changes drastically and because the fuel bundle characteristics are also needed for certain temperature states, each fuel axial zone is modeled at 0%, 40%, and 70% void concentration. Though potentially any parameter may be varied by TGBLA 6, the main variables manipulated in a lattice design are the pellet enrichment, number of gadolinium rods and gadolinium concentration in each gadolinium rod. As a result of utilizing 2-dimensional calculations in certain axial zones of the fuel bundle, enrichment distribution and gadolinium content are only varied in the axial zones represented by the 2-dimensional lattice physics calculations [5]. Because the influence of neighboring fuel bundle was assumed a secondary influence on the bundle behavior, TGBLA 6 assumes infinite lattice behavior as a boundary condition. Assuming infinite lattice behavior results in a good approximation of the lattice power peaking distribution as well as an accurate generation of group constants to be later utilized in PANAC11. Utilizing these design constraints and boundary conditions, TGBLA 6 uses the solution of the multi-group diffusion theory equation to generate group averaged cross-sections for 3 energy ranges. Group constants are generated for the fast, epithermal and thermal energy range to be later used by 3-dimensional core simulator PANAC11. PANAC11 After the multi-group cross sections were collapsed and generated by TGBLA6, Panac11 was utilized as the 3-dimensional full core simulator. Panac11 is a static, three-dimensional coupled nuclear-thermal-hydraulic computer program utilized to represent a BWR core by a coarse-mesh nodal, 1-1/2 group (quasi-two group), static diffusion theory approximation. The program was utilized explicitly for detailed three-dimensional

PAGE 31

19 calculations of neutron flux, power distributions, and thermal limits at different exposure steps during reactor core life. The main variable parameters in PANAC11 were the control rod positions, refueling patterns, coolant flows, reactor pressures, reactor power level as well as other operational and design variables [8]. The diffusion equations are solved using the fast energy group. Resonance energy neutronic effects are included in the model by relating the resonance fluxes to the fast energy flux. The thermal flux is represented by an asymptotic expansion using a slowing down source from the epithermal region. A pin power reconstruction model is also implemented to account for the effect of flux gradients across the nodes on the local peaking distribution. Utilized Temperature States, Boron Concentrations and Lattice Types There was a combination of 4 main types of temperature states and boron concentrations investigated in the study. These states included the hot uncontrolled state (HU), cold uncontrolled state (CU0), cold controlled state (CC) and a cold state containing soluble boron (CU###). The HU state was defined to be the operating temperature state with no control blade placed next to the fuel lattice. All cold states were to be defined at a moderator temperature of 160oC, and all the cold uncontrolled states also had no control bladed placed next to the fuel lattice. The CU### condition was designated as a cold lattice containing a homogenously dispersed soluble boron solution at a specified boron concentration. The CC condition represented a cold fuel lattice with a control blade placed in the upper and left side of the lattice. There were two fuel lattice types examined in the enrichment perturbation portion of this study. A C lattice was defined to be a fuel lattice that exhibited the same amount of moderator spacing on all four sides of the lattice while a D lattice exhibited

PAGE 32

20 slightly more moderator spacing in the vicinity of a control blade location. Therefore the radial power distributions of the two different lattices are slightly different in lattice peaking characteristics. Measurement of SLCS and SDM during the Lattice Development Stage SLCS and SDM are both global parameters that describe a margin experienced by the entire nuclear reactor core. Therefore the calculation of these parameters involves utilizing a 3-dimensional core simulation tool. Fuel bundles are generally designed by first utilizing a 2-dimensional fuel lattice physics tool to create average collapsed group cross sections to then be utilized by the 3-dimensional reactor core simulator. An abundant amount of energy groups are utilized in the lattice physics calculation in order to properly model the physics of the lattice. Many bundles within the reactor core will exhibit similar characteristics due to the similar enrichment or gadolinium concentration within the fuel bundle. Therefore by generating these average group cross sections for similar fuel bundles, the 3-dimensional core calculation is significantly faster because the calculation does not involve solving equations at an abundant amount of energies for many different points within the reactor core [14]. The 3-dimensional simulator only utilizes few averaged group cross sections (usually 3 averaged groups). The 3-dimensional core simulation tool utilized for this study, PANAC11 separated the reactor core into a series of 6 in. cubic nodes. A flux was then solved in each individual node. Enhancement to the fuel bundle design therefore involvements modifications to the fuel design in both the lattice physics development stage and the 3-dimensional core simulation stage. Since SDM and SLCS global parameters defined for the entire system,

PAGE 33

21 a separate set of parameters must be utilized for characterizing how fuel improvements in the lattice development stage will affect the full core global parameters. Maximizing SDM and SLCS involves increasing the difference between the hot operating condition and the cold shutdown condition. Therefore in the lattice development stage an enhancement in SLCS and SDM meant and improvement in the difference between the hot operating k and cold shutdown k. The cold shutdown condition related to SDM is defined to be when the lattice is controlled by the placing a control blade next to it. The parameter used to represent the maximized difference between hot operating k and cold shutdown k was designated HUCC, and calculated by the equation: lledColdControngHotOperatikkHUCC (2.5) The cold shutdown condition related to SLCS is when the lattice is controlled by placing a homogeneously dispersed solution throughout the fuel lattice. The parameter used to represent the maximized difference between hot operating k and cold shutdown k was designated HUCU###. The symbol ### represents the amount of parts-per-million of boron in the solution. HUCU### is calculated by the following equation: ppmtdSolutionaColdBoratengHotOperatikkHUCU###### (2.6) Maximizing these parameters in the lattice development stage will maximize the global parameters that these parameters represent in the 3-dimensional core modeling stage. Fuel Bundle Geometry The fuel bundle utilized in the lattice physics calculations was a typical 10X10 boiling water reactor fuel bundle design with 92 fuel rods and two water rods. Figure 2-1

PAGE 34

22 displays a cross sectional view of the fuel bundle geometry. Since each lattice physics calculations was completed utilizing a 2-dimensional model, the fuel bundle had to be sub-sectioned into 2-dimensional axial zones in order to accurately model the sections of the bundle that experience different void concentrations, decreased average moderator density, variable gadolinium and enrichment placement, and different lattice geometries. The axial zones utilized were designated naturally enriched bottom (NAT), power-shaping zone (PSZ), dominant zone (DOM), plenum zone (PLE), vanished rod location zone (VAN), natural vanished rod zone (N-V) and natural top zone(N-T). The NAT was a naturally enriched zone filled with all 92 fuel rods and no gadolinium. This zone represented the first 6 inches of the bottom of the active core length. Both the PSZ and DOM were enriched zones containing 92 fuel rods with gadolinium present in certain locations. The PSZ zone was located on top of the NAT zone and was 48 inches in length. The DOM zone was located on top of the PSZ and was 30 inches long. Though in 2-D geometry these axial zones were identical, the zones are separated due to the different inherent thermo-hydraulic and neutronic characteristics experienced in each zone. The PLE was a 12 inch zone containing 78 fuel rods and gadolinium rods present in needed locations. This zone was used to model the interface of the part length rods and the vanished rod locations began. On top of the PLE was the VAN. The VAN contained 78 fuel rods and 14 vanished rod locations and ranged between 37 and 48 inches in length. On top of the VAN was the N-V. The N-V contained natural enrichment and has pellets existing in 78 fuel rod locations. The N-T was also naturally enrichment but only had pellets in locations where no gadolinium existed in axial portions of that specific rod.

PAGE 35

23 Figure 2-1. A cross sectional view of the modeled fuel bundle. As displayed in figure 2-2, there existed only 4 possible fuel rod geometries. The DOM and PSZ had the same fuel rod geometry, and the VAN and N-V had the same fuel rod geometry. In the N-T the locations marked E are used to represent empty fuel locations in the lattice of where gadolinium rods exist in the N-V. In the N-V, N-T, and VAN the locations marked V are used to represent the vanished rod locations. In the PLE the locations marked E are used to represent the area of the plenum tip of the part length fuel rods. Though 4 different possible fuel rod geometries exist only 3 different types of geometries were modeled in the lattice physics investigations. The DOM, PLE and VAN region were modeled. The N-T was not investigated because this region contained only

PAGE 36

24 natural enrichment and therefore the low power level experienced in this region would never be most limiting in any cold shutdown condition. Figure 2-2. The geometric setup of the fuel lattice axial zones. Thermal Limit Design Considerations Nuclear reactors are designed so that operation will not induce unnecessary risk to the health and safety of the general public. Therefore thermal limits are imposed on

PAGE 37

25 certain core parameters to ensure that radioactive release during reactor operation or any type of transient event does not exceed the acceptable limits imposed by the NRC. Constraining operation to within the thermal limits of the fuel guarantees that during normal operation and emergency transients the fuel integrity will be maintained. For the applicability of this project the two main thermal limits monitored were MFLPD and MFLCPR. These limits are set to limit boiling around the fuel rod locations and limit fuel rod power density in order to preserve fuel integrity [11]. Linear heat generation rate (LHGR) is the amount of power produced per length of fuel and is defined as LengthRodFuelAssmeblyPerRodsFuelAssembliesFuelofNumberOutputPowerThermalMaximumLHGRAverage____________ (1.6) A maximum average LHGR is specified for the utility in order to limit the plastic strain or deformation of the cladding. A limit of 1% deformation of zicaloy cladding is considered a conservative limit below which fuel damage is not expected to occur. New pellets undergo slight densification during irradiation. This causes the gap between the fuel and the cladding to increase and thus decrease thermal conductance. The pellet densification also has a shrinking axial effect. If one of the pellets gets stuck during this process, a gap is created resulting in more fissions occurring in newly exposed faces of the pellets increasing heat flux in that area [6]. Therefore LHGR is adjusted for the possible elevated heat flux and is defined as LTLPPLHGRLHGRdesignit*1*maxlim (1.7) Where: LHGRdesign = Maximum LHGR allowable to prevent clad damage

PAGE 38

26 LT = Total active core length L = Axial position in feet above the bottom of the core maxPP Maximum power spiking penalty In the core simulator LHGR is calculated for certain axial nodes. MFLPD is the maximum fractional limiting power density for the most limiting node and is defined as itnodeLHGRLHGRMFPLDlimmax_ (1.8) As long as MFLPD is less than one LHGR is not exceeded. However, because these calculations are based off of the assumptions of the maximum power spike penalty, and because the designer needs to be certain that MFLPD will never exceed one, the design basis requirement for MFLPD is 0.909 to ensure enough variation between the actual calculation and the measured data [8]. Critical power is the bundle power required to produce transition boiling in a reactor channel. If transition boiling were to manifest in a channel, it may lead to fuel rod dry out in the channel with the inability of the fuel rod surface to rewet. This phenomenon leads to a decrease in the ability of the clad to reject heat to the water through convection and thus heat up the clad to the point of mechanical failure [6]. CPR is the ratio to determine how close the actual power is to transition boiling and is defined as: APCPCPR (1.9) CP = Critical power for transition boiling AP = Actual power.

PAGE 39

27 CPR must always be below 1.0 for safe operation. MFLCPR is the flow adjusted ratio of the operating limit CPR for a specific fuel type to the CPR of that bundle and is defined as: CPRKLimitCPRMFLCPRf*_ (1.10) Kf = Flow adjustment factor Since MFLCPR should never exceed one in any section of the reactor during operation, and since slight uncertainty exists in knowing the actual power of the reactor and the critical power for a specified bundle, the design basis for MFLCPR is set to 0.930 to accommodate these uncertainties [8].

PAGE 40

CHAPTER 3 MAXIMIZING HOT-COLD BORATED k DIFFERENCE UTILIZING ENRICHMENT TGBLA 6 was utilized to understand SLCS improvement by enhancing lattice behavior characteristics utilizing enrichment perturbations. C and D lattices types were examined at 0%, 40% and 70% void fraction. The four system states inspected were HU, CU0, CU###, and CC. Lattices with a homogeneous enrichment distribution were examined to determine the effects of lattice average enrichment on the enhancement of the HUCU### and HUCC. Next, Local and gross enrichment perturbations were also analyzed to determine the effects of these types of enrichment perturbations on HUCU### and HUCC. Since an enormous amount of combinations of temperature states, lattice axial zones, lattice types and void concentrations could be created, the homogenous enrichment work was utilized to determine which of these temperature states, lattice axial zones, lattice types and void concentration were most limiting in order to limit the amount of cases to investigate therefore only examining the most effective strategies for enhancement. Homogeneous Enrichment Distribution A homogeneously enriched distribution was defined to be a fuel lattice with constant enrichment throughout the lattice. Therefore homogenous enrichment perturbations were defined as a change in enrichment to every fuel pin in the lattice by the exact same amount. 28

PAGE 41

29 Determining the Most Limiting Lattice Axial Zone and Void Concentration A variety of tests were completed to determine which lattice parameters were most limiting to HUCU### and HUCC. Figure 3-1 depicts the effects of lattice type and void fraction on HUCU###. Lattice type did not significantly affect HUCU###; however, as void fraction increased HUCU### decreased. For this case, at 0 GWD/STU the difference in HUCU### was solely related to the effective moderator density difference at increased void fraction. At higher void fractions the average moderator density was decreased. Because the average moderator density was decreased fewer neutrons were thermalized and absorbed by the fuel for fission, and an increased amount of neutrons were parasitically captured by the fuel [18]. Therefore HUCU### was smaller for higher void fractions. 00.020.040.060.080.10.120.140.160.18010203040506070Exposure (GWD/STU)HUCU660 0% Void, C Lattice 40% Void, C Lattice 70% Void, C Lattice 0% Void D lattice 40% Void, D Lattice 70% Void, D lattice Figure 3-1. Exposure dependent HUCU660 for the DOM at 3.95% enrichment. Initially U238 was the main parasitic neutron absorber in the fuel due to increased void concentration. When U238 absorbed a neutron it became Pu239. Because Pu239 had a high thermal absorption cross section (a = 1015b) as well as multiple resonance absorption peaks, it became another main parasitic neutron absorber. The increases in

PAGE 42

30 build up of parasitic neutron absorbers lead to a decrease in the effectiveness of boron to capture thermal neutrons [18]. Therefore as the lattice burned, plutonium was built up thus increasing the content of competing neutrons absorbers and therefore decreasing HUCU###. Furthermore, in the higher void history condition it took longer for plutonium production to reach an equilibrium state; therefore at increased void history conditions HUCU### decreases at a faster rate for a longer amount of time. Figure 3-2 illustrates the effects of the combination of axial zone and void fraction on HUCU###. The DOM was the most limiting axial zone because of the maximum amount of fuel rod inventory present in the lattice and minimum amount of volume to place borated water. A lattice geometry that allows for more moderator space allows for more ability to place borated water in the lattice; therefore since the VAN and PLE both have evacuated regions where more space exists to place a boron volume these axial zones were not the most limiting in terms of HUCU###. In order to maximize improvements in HUCU###, enhancements must be made to the most limiting conditions of HUCU###. HUCU### was most limiting in the 70% void fraction and DOM case; however, in the core, on average, the DOM exhibits a 40% void fraction therefore modeling a DOM at 70% void fraction would not have been an accurate realistic model to examine SLCS. The realistic model utilized which was most limiting was determined to be 40% void fraction in the DOM. Therefore since lattice type did not significantly effect HUCU###, and the DOM 40% void fraction was the most realistic limiting condition state, the C lattice type, DOM, 40% void fraction lattice was chosen as the base lattice in which all other perturbations were compared.

PAGE 43

31 00.050.10.150.20.25010203040506070Exposure (GWD/STU)HUCU660 Dom Zone, 0% Void Dom Zone, 40% Void Dom Zone, 70% Void Ple Zone, 0% Void Ple Zone, 40% Void Ple Zone, 70% Void Van Zone, 0% Void Van Zone, 40% Void Van Zone, 70% Void Figure 3-2. Exposure dependent HUCU660 at varied void fraction and axial zone for the C lattice at an enrichment of 3.95%. Understanding the Exposure Dependent HUCU### Curve Understanding the exposure dependence of the HUCU### curve was paramount to determining the appropriate strategy for enhancing HUCU###. Figure 3-3 depicts the two major portions of the exposure dependent HUCU### curve. Portion A encompasses 0 GWD/STU to roughly 11-15 GWD/STU depending upon geometry of the axial zone and void concentration. Portion B encompasses the rest of the HUCU### curve. When a fissile isotope absorbs a neutron, the isotope may either undergo fission or parasitic capture. The capture-to-fission ratio is defined by: f (3.1) In a thermal reactor the majority of the neutrons cause fissions at thermal energies in U235. Therefore for most thermal reactor applications the capture-to-fission ratio is an

PAGE 44

32 explanation of the fission efficiency of the thermal neutrons [13]. As increases k decreases because more thermal neutrons undergo parasitic capture and are removed from the system instead of undergoing a fission event and creating more neutrons [18]. During portion A of the exposure dependent HUCU### curve, HUCU### is decreasing due to increased plutonium production. Pu239 has a capture-fission-ratio of 0.370 (2200 m/s neutron) while U235 has a capture to fission ratio of 0.175 (2200 m/s neutron) [13]. Therefore with an increased plutonium acts as a competing neutron absorber that decreases boron worth thus limiting the effective absorption ability of the boron. The capture-to-fission ratio is a function of the system temperature. Doppler broadening is a phenomenon in which due to the kinetic motion of the target atoms at elevated temperatures the resonance absorption cross sections broaden while the peak magnitude of the cross section decreases, and in most cases slightly preserving the area under the original resonance [3]. Therefore though the effective peak of the cross section has decreased, the width of the resonance is increased and therefore the resonance affects a greater range of energy of neutrons; therefore causing a greater interaction rate in that energy interval and thus leading to more absorption and decreased flux in that energy interval [15]. At higher temperatures there is more kinetic motion of target particles and thus more Doppler broadening of the resonance cross sections. With increased parasitic capture and decrease thermal-to-fast flux ratio, HU k decreases at a much faster rate than CU### during Portion A of the HUCU curve because the worth of the plutonium produced is progressively worth more in the hot operating condition than in the cold condition.

PAGE 45

33 The thermal-to-fast flux ratio is defined by: fast thermal 1 (3.2) As the temperature of the system increases the average moderator density decreases. Because the average moderator de nsity is decreased the neutron spectrum has a higher density in the fast region. In th e hot operating condition the decreased average moderator density causes an increase in 1; therefore fewer neutrons are available for thermal fission events. Since at elevated temperatures there exists greater Doppler Broadening as well as decreased averag e moderator density, the decrease in 1 leads to a decrease in k of the system. Therefore the increasing and decreasing 1 in the hot condition leads to a decrease in th e HUCU### curve because the hot k is decreasing faster in comparison to the cold k. HUCU### Exposure (GWD/STU) A B Figure 3-3. Exposure dependant HUCU### curve. During portion B, as plutonium production reaches an equilibrium concentration the difference between the hot operating condition and cold condition worth becomes

PAGE 46

34 almost constant and therefore during this portion of the curve HUCU### experiences relatively no exposure dependence no exposure dependence. Enrichment and Boron Concentration Effects Figure 3-4 depicts the effects of increasing enrichment on HUCU### at different boron concentrations. For each 1.0% increase in average enrichment 0.0188 HUCU### was lost. Since the derivatives of HUCU### were equivalent at different boron concentrations, the amount of HUCU### gained from a boron concentration increase was linearly dependent on the boron concentration increase and not also affected by the average enrichment. Equation 3.3 calculated 1.84 x 10-4 HUCU660 gained for each 1 ppm or boron introduced to the lattice. Therefore 99.6 ppm of boron was required to compensate for a 1.0% average enrichment increase. EnrichmentSpecificionConcentratBoronHUCUBoronppmgainedHUCU__###__1_### (3.4) 00.050.10.150.20.250.300.40.81.21.622.42.83.23.644.44.85.2EnrichmentHUC U 660 PPM 726 PPM 742 PPM 792 PPM 935 PPM Figure 3-4. Beginning of cycled HUCU vs. enrichment in the DOM, at 40% void fraction, for a C lattice.

PAGE 47

35 Power Peaking Distribution The lattice power peaking distribution was a function of the relative distance of fissile material from the moderating regions. Moderation capability of certain lattice regions was a function of the water boundaries of that certain lattice region as well as the temperature state, boron concentration and geometry of poison utilized within the lattice. Power Peaking distributions for a homogenous 3.95% enrichment lattice at 5 GWD/STU are displayed in figure 3-5 for HU, CU0, CC and CU660. Each fuel pin location is identified by the horizontal and vertical location in which the pin resides. The water rod locations were marked with a zero in order to distinguish the water rod locations from the fuel pin locations. 123456789101234567891011.43661.22771.12291.07471.0611.0691.09251.1411.24311.4510.5740.56290.570.58890.61580.64960.69430.78260.99541.3971.30 < x21.22771.00670.90690.86970.86980.89090.91020.93911.0281.243320.56290.59380.62410.66020.70620.76050.79990.84670.98281.30141.30 > x > 1.2031.12290.90690.81880.80060.83190.90320.91620.88410.93921.141530.570.62410.67630.7410.83260.9660.99660.94291.01321.29371.20 > x > 1.1041.07470.86970.80060.81480.9034000.91640.91051.093240.58890.66020.7410.85391.0385001.09391.06281.31561.10 > x > 1.0551.0610.86980.83190.90340.9974000.90350.89141.069950.61580.70620.83261.03851.2382001.12971.08781.34031.05 > x > 1.0061.0690.89090.9032000.99760.90370.83230.87041.062160.64960.76050.966001.29821.16161.04211.08491.36291.00 > x > 0.9571.09250.91020.9162000.90370.81530.80130.87051.07670.69430.79990.9966001.16161.03131.00621.09621.39710.95 > x > 0.9081.1410.93910.88410.91640.90350.83230.80130.81960.9081.124580.78260.84670.94291.09391.12971.04211.00621.02851.14541.46370.90 > x > 0.8591.24311.0280.93920.91050.89140.87040.87050.9081.0081.229690.99540.98281.01321.06281.08781.08491.09621.14541.27661.60350.85 > x > 0.80101.451.24331.14151.09321.06991.06211.0761.12451.22961.439101.3971.30141.29371.31561.34031.36291.39711.46371.60351.91640.80 > x123456789101234567891011.45141.21961.12211.08471.07721.08531.10161.13781.23191.462111.38471.18591.10631.07791.07361.08081.09281.11971.19561.392321.21960.97270.87890.8520.85930.8840.89660.91060.99151.232921.18590.96750.88770.86690.87550.89850.90820.91650.98371.196331.12210.87890.79450.78770.83220.92560.93210.86770.91111.139531.10630.88770.81430.81110.85450.94340.94790.88260.91691.120941.08470.8520.78770.82020.9418000.93270.89771.10441.07790.86690.81110.84440.9612000.94840.90911.094651.07720.85930.83220.94181.0713000.92660.88551.088251.07360.87550.85450.96121.085000.94420.89981.083161.08530.8840.9256001.07170.94260.83350.86121.080661.08080.89850.9434001.08530.96190.85560.87711.076271.10160.89660.9321000.94260.82140.78930.85431.088571.09280.90820.9479000.96190.84550.81260.86881.080981.13780.91060.86770.93270.92660.83350.78930.79660.88161.126481.11970.91650.88260.94840.94420.85560.81260.81620.891.109791.23190.99150.91110.89770.88550.86120.85430.88160.9761.224691.19560.98370.91690.90910.89980.87710.86880.890.97021.1898101.46211.23291.13951.1041.08821.08061.08851.12641.22461.4576101.39231.19631.12091.09461.08311.07621.08091.10971.18981.3894Key 5 GWD,HU,3.95 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,CU0,3.95 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,CC,3.95 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,CU660,3.95 Enrichment, Dom Zone, C Lattice 40% Void Figure 3-5. The power peaking distributions at 5 GWD/STU, 3.95% enrichment, DOM, C lattice, and 40% void fraction vs. temperature state and boron concentration.

PAGE 48

36 Figure 3-5 suggests that there were significant differences between the HU and CU0 power peaking distributions. In the HU state the outer borders of the lattice exhibited the greatest amount of power peaking due to the close proximity of large areas of water to that location and therefore displaying the greatest moderation capabilities. Due to the moderation of the internal water rod locations, fuel pins located near the internal water rods exhibited a higher relative power than locations that were located away from the borders of the lattice and away from the internal water rod locations. The lack of moderation for the fuel pins located away from the borders of the lattice and away from the internal water rod locations caused these locations to exhibit the lowest relative power. In the CU0 state, power was raised in the highly moderated areas. With no voids, the outer borders of the lattice and the internal water rod locations exhibit a greater amount of power peaking than the areas of the lattice that were between these locations. Because of this effect, the importance of fuel pins located away from the borders and water rod locations were significantly decreased, and if a perturbation were to be made to a fuel lattice in order to improve inherent HUCU0 characteristics these locations would not play an important role. There were also significant differences in power peaking distribution for different poison types. In the CC state a high anisotropy of power peaking was exhibited due to poison residing at the corners of the lattice. Fuel pins located closest to the control blade exhibit the greatest power suppression; therefore if a poison introduction was necessary for power suppression in the HU state, placing that poison away from the greatest power suppressed pins in the CC condition will achieve the greatest improvement in HUCC.

PAGE 49

37 In the CU660 state the boron caused the power to be suppressed in highly moderated regions thereby flattening out the power distribution of the lattice. Because boron was present in the moderator, regions that had greater power peaking due to increased moderation capability also consequently had greater power suppression from the boron dispersed within the moderator. Because of the flatter power distribution in the CU660 as compared with the HU state, placing power suppressors in peaked locations corresponding to the HU state did not necessarily have as severe an impact on the cold borated state. However, this leads to a distinct design advantage because it may be possible switch locations of a distributed poison and have a miniscule effect on HU but a major effect on CU###. Therefore in order to maximize HUCU###, power suppressors must be placed in areas where the CU### state exhibits a higher power peak than the power peak in the HU state. As enrichment increased the power peaking distribution in the lattice became more skewed. Figure 3-6 displays that as average enrichment was increased in the CU### state the power peaked more in the border regions and internal water rod locations of the lattice. Figure 3-7 displays that as average enrichment was increased in the HU state the power also peaked more in the border regions and internal water rod locations of the lattice. Therefore there exist fewer locations in which moving a distributed poison will not greatly affect the HU state while greatly affecting the CU### state. Unfortunately this demonstrates that in a power up rate or increased exposure design, it will be harder for the designer to create a design that improves SLCS and decreases the power peak in the HU state. Figure 3-6. The power peaking distributions at 5 GWD/STU, CU660, DOM, C lattice, and 40% void fraction vs. enrichment.

PAGE 50

38 123456789101234567891011.19231.0951.05951.04711.04621.05031.05571.0671.09961.192911.45141.21961.12211.08471.07721.08531.10161.13781.23191.462121.0950.97290.93360.92370.92970.94190.9470.9510.98221.098621.21960.97270.87890.8520.85930.8840.89660.91060.99151.232931.05950.93360.89630.89580.92450.97750.98020.94010.95111.066231.12210.87890.79450.78770.83220.92560.93210.86770.91111.139541.04710.92370.89580.91680.9912000.98040.94741.055341.08470.8520.78770.82020.9418000.93270.89771.10451.04620.92970.92450.99121.0681000.97790.94241.050151.07720.85930.83220.94181.0713000.92660.88551.088261.05030.94190.9775001.06830.99160.92510.93041.046161.08530.8840.9256001.07170.94260.83350.86121.080671.05570.9470.9802000.99160.91740.89650.92461.047271.10160.89660.9321000.94260.82140.78930.85431.088581.0670.9510.94010.98040.97790.92510.89650.89720.93461.059781.13780.91060.86770.93270.92660.83350.78930.79660.88161.126491.09960.98220.95110.94740.94240.93040.92460.93460.9741.095391.23190.99150.91110.89770.88550.86120.85430.88160.9761.2246101.19291.09861.06621.05531.05011.04611.04721.05971.09531.1923101.46211.23291.13951.1041.08821.08061.08851.12641.22461.4576123456789101234567891011.34151.1681.09651.07041.06651.07331.08461.10931.17711.34811.43131.2041.11661.08621.08161.08911.10161.13031.21411.439921.1680.9730.89950.87970.88760.90890.91840.9270.98861.177521.2040.960.87470.85340.86270.88730.89690.90430.97661.21531.09650.89950.83120.82770.86790.94860.95330.89550.92731.110131.11660.87470.79690.79430.84090.93880.94290.86870.90481.131941.07040.87970.82770.85790.9645000.95370.91911.085941.08620.85340.79430.83090.9588000.94340.89791.103851.06650.88760.86790.96451.0768000.94930.90991.075151.08160.86270.84090.95881.095000.93970.88871.091961.07330.90890.9486001.07720.96510.86890.8891.068661.08910.88730.9388001.09540.95960.84210.86451.084871.08460.91840.9533000.96510.85890.8290.88131.072871.10160.89690.9429000.95960.83210.79590.85561.089881.10930.9270.89550.95370.94930.86890.8290.83280.90141.099381.13030.90430.86870.94340.93970.84210.79590.7990.87731.120691.17710.98860.92730.91910.90990.8890.88130.90140.97531.171291.21410.97660.90480.89790.88870.86450.85560.87730.96311.2086101.3481.17751.11011.08591.07511.06861.07281.09931.17121.3451101.43991.2151.13191.10381.09191.08481.08981.12061.20861.43725 GWD,CU660,0.71 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,CU660,3.2 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,CU0,3.95 Enrichment, Dom Zone, C Lattice 40% Void5 GWD, CU660, 4.9 Enrichment, Dom Zone, C Lattice, 40% Void Figure 3-6. The power peaking distribution at 5 GWD/STU, CU660, DOM, C lattice, and 40% void fraction versus enrichment. 123456789101234567891011.22971.12191.07021.04431.03651.04181.05651.08361.1341.238711.43661.22771.12291.07471.0611.0691.09251.1411.24311.451.30 < x21.12190.99720.94610.92450.92370.93510.94840.96741.01261.132821.22771.00670.90690.86970.86980.89090.91020.93911.0281.24331.30 > x > 1.2031.07020.94610.89970.88750.90650.95070.96140.94280.96751.082731.12290.90690.81880.80060.83190.90320.91620.88410.93921.14151.20 > x > 1.1041.04430.92450.88750.89430.9499000.96150.94861.055741.07470.86970.80060.81480.9034000.91640.91051.09321.10 > x > 1.0551.03650.92370.90650.94991.0071000.95090.93541.041251.0610.86980.83190.90340.9974000.90350.89141.06991.05 > x > 1.0061.04180.93510.9507001.00720.95010.90680.92411.03661.0690.89090.9032000.99760.90370.83230.87041.06211.00 > x > 0.9571.05650.94840.9614000.95010.89460.88790.9251.043871.09250.91020.9162000.90370.81530.80130.87051.0760.95 > x > 0.9081.08360.96740.94280.96150.95090.90680.88790.90010.94671.069881.1410.93910.88410.91640.90350.83230.80130.81960.9081.12450.90 > x > 0.8591.1341.01260.96750.94860.93540.92410.9250.94670.99781.121591.24311.0280.93920.91050.89140.87040.87050.9081.0081.22960.85 > x > 0.80101.23871.13281.08271.05571.04121.0361.04381.06981.12151.2291101.451.24331.14151.09321.06991.06211.0761.12451.22961.4390.80 > x123456789101234567891011.38911.20631.11161.06681.05371.06141.0841.12951.22171.402211.4881.24931.13461.08371.06951.07781.10171.15271.26451.501721.20631.01020.91850.88280.88210.90170.92080.94951.03131.221821.24931.00040.89340.85550.85660.87930.89820.92621.02161.264931.11160.91850.83670.81830.84620.91060.92410.89810.94971.129831.13460.89340.79980.78260.81750.89650.90870.86880.92641.153341.06680.88280.81830.82960.9089000.92420.92111.084541.08370.85550.78260.80030.8993000.90890.89861.102651.05370.88210.84620.90890.9928000.91090.90211.062151.06950.85660.81750.89931.0045000.89690.87981.078961.06140.90170.9106000.9930.90910.84660.88271.054661.07780.87930.8965001.00470.89970.8180.85731.070971.0840.92080.9241000.90910.830.81890.88351.067871.10170.89820.9087000.89970.80090.78330.85651.085381.12950.94950.89810.92420.91090.84660.81890.83740.91941.112881.15270.92620.86880.90890.89690.8180.78330.80080.89461.136591.22171.03130.94970.92110.90210.88270.88350.91941.01131.207891.26451.02160.92640.89860.87980.85730.85650.89461.0021.2517101.40221.22181.12981.08451.06211.05461.06781.11281.20781.3909101.50171.26491.15331.10261.07891.07091.08531.13651.25171.4913Key 5GWD,HU,0.71 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,HU,3.2 Enrichment, Dom Zone, C Lattice 40% Void5 GWD,HU,3.95 Enrichment, Dom Zone, C Lattice 40% Void5 GWD, HU, 4.9 Enrichment, Dom Zone, C Lattice, 40% Void Figure 3-7. The power peaking distributions at 5 GWD/STU, HU, DOM, C lattice, and 40% void fraction vs. enrichment.

PAGE 51

39 Heterogeneous Enrichment Distribution After determining the effects of average enrichment on the behavior of HUCU###, heterogeneous enrichment perturbations were then examined in order to determine if enrichment changes to individual fuel pins could affect HUCU###. The two major types of enrichment perturbations investigated were localized enrichment perturbations and gross enrichment perturbations. Localized enrichment perturbations were considered to be small sets of fuel pins that were either increased or decreased in enrichment by a certain amount holding the rest of the fuel lattice at constant enrichment. Gross Enrichment perturbations were considered to be a large lump of fuel pins that were either increased or decreased in enrichment by a certain amount holding the average enrichment of the entire lattice constant. Localized Enrichment Perturbation Localized enrichment perturbation patterns were generated based on the power peaking distribution map. The fuel pin locations were set into groups based on locations exhibiting similar power peaking in the homogeneously enriched lattice calculation. Since lattice power peaking was determined to be dependent on enrichment, 3.95% enrichment was chosen as the distribution for which the determination of the pattern type was made. Figure 3-8 displays the distribution of perturbations made to the lattice. Each group of numbers represents a group of fuel pins that were either increased or decreased by 1.0% enrichment as the rest of the lattice was kept at a constant enrichment. Localized enrichment perturbations had no effect on HUCU### as depicted in figure 3-9. The minute difference of 0.00373 HUCU### in figure 3-9 was considered only a function

PAGE 52

40 of the 0.2% average enrichment difference exhibited between each pattern utilized. Therefore localized enrichment perturbation did not greatly affect HUCU### behavior. 12345678910112344443212257888775233799977873448997WW774548976WW7846487WW679847477WW799848378779997392577888752101234444321 Figure 3-8. Localized enrichment perturbation map. 00.020.040.060.080.10.120.14010203040506070Exposure (GWD/STU)HU-CU660 Map 1, Pattern 1 Map 1, Pattern 2 Map 1, Pattern 3 Map 1, Pattern 4 Map 1, Pattern 5 Map 1, Pattern 6 Map 1, Pattern 7 Map 1, Pattern 8 Map 1, Pattern 9 0.074957 0.078689Maximum Change in Hot Uncontolled k-infinity Due To Cold Uncontrolled k-infinity = 0.003732 Which is Soley a Function of Average Enrichment Change (0.2%). Figure 3-9. Exposure dependent HUCU660 for different localized enrichment perturbation patterns.

PAGE 53

41 Gross Enrichment Perturbation Gross enrichment perturbations were next examined in order to determine how these type of lattice perturbations would affect shutdown behavior. Figure 3-10 displays an example pattern of gross enrichment perturbations and table 3-1 lists the enrichment perturbations made to that example pattern. Table 3-1. Gross enrichment perturbation scheme for figure 3-10. Pattern Enrichment Perturbation (1-2) Pattern Enrichment Perturbation (1-2) 1 1.6%-4.9% 7 4.9%-1.6% 2 2.4%-4.9% 8 4.9%-2.4% 3 3.2%-4.9% 9 4.9%-3.2% 4 4.4%-4.9% 10 4.9%-4.4% 5 3.2%-4.4% 11 4.4%-3.2% 6 3.6%-4.4% 12 4.4%-3.6% 12345678910111111111112111111112231111111122411111WW222511111WW2226111WW222227111WW222228111222222291222222222101222222222 Figure 3-10. An example of a gross enrichment perturbation map. Gross enrichment perturbation demonstrated no effect on HUCU###. Though HUCU### was not a function of localized and gross enrichment perturbation, HUCC was highly dependent upon these perturbations. Figure 3-11 and 3-12 display the difference in effect of gross lattice perturbation skewing. Notice in figure 3-11 no effect was noticed on HUCU###; however, in figure 3-12 HUCC was highly dependent upon enrichment distribution.

PAGE 54

42 00.020.040.060.080.10.120.140.16010203040506070Exposure (GWD/STU)HUCU660 Map 6, Pattern 2 Map 6, Pattern 8 Figure 3-11. Exposure dependent HUCU660 at 40% void fraction, in the DOM, with a C lattice. 00.050.10.150.20.25010203040506070Exposure (GWD/STU)HUCC Map 6, Pattern 2 Map 6, Pattern 8 Figure 3-12. Exposure dependent HUCC at 40% void fraction, in the DOM, with a C lattice. Placing a higher enrichment closer to the control blade location allowed for greater power suppression and thus enhanced HUCC due to the increased control the blade exhibited over the maximum power producing section of the lattice. Therefore though

PAGE 55

43 enrichment perturbation was not limiting in HUCU###, distorting the enrichment distribution had an effect on HUCC. Though SLCS does not depend on local or gross enrichment perturbation, SDM is sensitive to this type of perturbation. However, the designer may only enhance HUCU### by perturbing average enrichment of the entire lattice therefore as long as the average of the enrichment of the lattice satisfies SLCS requirements the enrichment may be perturbed to meet SDM without violating SLCS.

PAGE 56

CHAPTER 4 MAXIMIZING HOT-COLD BORATED k DIFFERENCE UTILIZING GADOLINIUM There were four different isolated studies examined for the purpose of determining the optimum strategies for utilizing gadolinium to enhance HUCU###. Gadolinium rods were examined in a variety of clumped geometries in order to determine the lumped spatial self-shielding effects. After the effects of self-shielding were determined, the effects of increasing the amount of gadolinium rods on HUCU### were investigated. Gadolinium concentration was next analyzed. Finally, gadolinium rod placement was examined to determine the optimum gadolinium locations for enhancing HUCU### without diminishing HUCC. Spatial Self-Shielding Effects of Gadolinium Rods on HUCU### Gadolinium rods were clumped in a variety of geometries to determine the effects of lumped spatial self-shielding on HUCU###. Four samples of examined clumped gadolinium rod geometries are displayed in figure 4-1. Clumping the gadolinium rods decreased the BOL gadolinium worth because the gadolinium rods were effectively spatially self-shielding each other from the impinging neutron flux. The self-shielding of the gadolinium decreased the effective surface area utilized for neutron absorption [15]. Because the thermal-to-fast flux ratio was much higher in the cold state than in the hot state more neutrons were likely to be thermally absorbed in the gadolinium in the cold condition; therefore decreasing the effective surface area for neutron absorption decreases 44

PAGE 57

45 the effectiveness of the power suppression from the gadolinium rods and thus decreasing HUCU###. Figure 4-1. Four sample clumped geometries. Figure 4-2 depicts the effects of gadolinium spatial self-shielding on gadolinium worth. Highlighted in red are the patterns corresponding to those displayed in figure 4-1. As gadolinium clumping increased, gadolinium worth decreased. Face adjacent clumping of all four sides of a gadolinium rod resulted in a 38% decrease in gadolinium worth, and face adjacent clumping of two sides resulted in a 19% decrease in gadolinium worth. Diagonal face adjacent clumping lead to a 7% decrease in gadolinium worth.

PAGE 58

46 -0.3-0.25-0.2-0.15-0.1-0.05010111213Pattern Gadolinium Worth (dk/k) HU Gad Worth CU0 Gad Worth CU660 Gad Worth CU 935 Gad Worth CC Gad Worth Figure 4-2. Corresponding 0 GWD/STU gadolinium worth for the patterns displayed in figure 4-1. 00.020.040.060.080.10.12010203040506070Exposure (GWD/STU)HUCU660 Pattern 10 Pattern 11 Pattern 12 Pattern 13 Figure 4-3. Corresponding exposure dependent gadolinium clumping effects on HUCU660 for the patterns displayed in figure 4-1.

PAGE 59

47 Because clumping a group of gadolinium rods decreased the effective surface area for absorption, the exposure time required to burn out the gadolinium increased. In figure 4-3 as clumping increased the exposure point in which gadolinium burns out also increased. Also depicted in figure 4-3 is the decrease in HUCU### as a function of increase clumping. Therefore in order to design an optimum lattice to enhance HUCU### gadolinium rods must be spaced as far apart as reasonably achievable and face adjacent and diagonal adjacent clumping must be eliminated. The Effects of Increasing the Amount of Gadolinium Rods on HUCU### In a power up-rate and an increased exposure cycle, extra positive reactivity must be installed in the fuel bundle. Placing extra positive reactivity will cause a greater skewing of the lattice power peaking as well as violating beginning of cycle critical eigenvalue requirements. In order to decrease the beginning of cycle eigenvalue to the critical requirements and decrease power peaking to improve lattice efficiency and meet thermal margins, gadolinium must be placed in the fuel bundle. As more positive reactivity is installed, more gadolinium rods at higher concentrations are needed. The effects of increasing the amount of gadolinium rods in the lattice on gadolinium worth are demonstrated in figure 4-4. The gadolinium rod placement geometry was held constant while 8 to 18 rods were placed in the lattice. As the amount of gadolinium rods placed in the lattice increased, the degree of gadolinium clumping decreased due to the size limitations of the lattice. In the increase of 15 gadolinium rods to 16 gadolinium rods, a clumped geometry was utilized that resulted in a decrease in gadolinium worth. Each gadolinium rod insertion for the hot condition was worth -0.0103 k/k while each gadolinium rod insertion in the CU660 case was worth -0.0095

PAGE 60

48 k/k leading to 0.5 mk/k difference in gadolinium worth between the hot and cold lattice states. -0.3-0.25-0.2-0.15-0.1-0.05089101112131415161718Number of RodsGadolinium Worth (dk/k) HU Gad Worth CU0 Gad Worth CU660 Gad Worth CU935 Gad Worth CC Gad Worth Figure 4-4. The effects of increased number of gadolinium rods on the gadolinium worth at 0 GWD/STU. Increasing the amount of gadolinium rods decreased hot and cold k by increasing the amount of neutrons removed from the system by absorption. HUCU### also decreased as the number of gadolinium rods increased. Figure 4-5 displays BOL decrease in HUCU660 as a function of increasing the amount of gadolinium rods placed in the lattice. The thermal-to-fast flux ratio is higher in the cold state than in the hot state due to the cold states increased moderator density, and the thermal-to-fast flux ratio is lower in higher boron concentration due to decreased thermal neutron availability after boron capture. Gadolinium is dominantly a thermal neutron absorber therefore in the increased

PAGE 61

49 thermal-to-fast flux ratio gadolinium was a more effective absorber. The decreased thermal-to-fast flux in the hot state as compared to the cold borated states results in a decrease in HUCU### as each gadolinium rod was inserted because the gadolinium was worth more per rod insertion in the cold state than in the hot state. Gadolinium rod worth was also a function of the boron concentration utilized in the cold condition. At 0 GWD/STU HUCU660 changes -0.0017 per rod insertion (10-13 ppm boron equivalence) while HUCU935 changes -0.0025 per rod insertion (16-20 ppm boron equivalence). This demonstrates that when a utility decides to go to a power up-rate or increased exposure cycle, the increased amount of gadolinium needed to offset the increased installed reactivity will result in a decrease in the HUCU### parameter on the lattice level resulting in a decrease in SLCS margin on the core wide level. y = -0.0025x + 0.1197y = -0.0017x + 0.070300.020.040.060.080.10.1202468101214161820Number of Gadolinium RodsHUCU HUCU660 HUCU935 Figure 4-5. The effects of the number of gadolinium rods inserted on HUCU at 0 GWD/STU.

PAGE 62

50 The Effects of Increasing the Gadolinium Concentration on HUCU### The effects of increasing gadolinium concentration of a given gadolinium configuration on HUCU### was next examined to determine if gadolinium concentration was a design constraint for HUCU###. Increasing the gadolinium concentration of the lattice had similar results to increasing the amount of rods in the lattice. Figure 4-6 displays the increase in gadolinium worth as a function of increasing concentration. For CU660 at 0 GWD/STU a 1% increase in gadolinium concentration for 14 gadolinium rods is worth -0.002343 k/k. For HU at 0 GWD/STU a 1% increase in gadolinium concentration for 14 gadolinium rods is worth -0.004587 k/k. -0.3-0.25-0.2-0.15-0.1-0.0500%1%2%3%4%5%6%7%8%9%Gadolinium ConcentrationGadolinium Worth (dk/k) HU Gad Worth CU0 Gad Worth CU660 Gad Worth CU935 Gad Worth CC Gad Worth Figure 4-6. The effect of increasing the gadolinium concentration for 14 gadolinium rods on gadolinium worth at 0 GWD/STU. The HU state exhibited 2 times greater worth in increasing 1% in gadolinium worth than the CU state; therefore increasing the gadolinium concentration will decrease HUCU###. Figure 4-7 exhibits the decrease in HUCU### as the gadolinium

PAGE 63

51 concentration is increased. For a 1% Change In Concentration for 14 Gadolinium Rods at 0 GWD/STU, HU660 changes -0.004504 (33 ppm boron equivalent) and HU935 changes -0.004906 (36 ppm boron equivalent). y = -0.4504x + 0.0795y = -0.4906x + 0.119800.020.040.060.080.10.120%1%2%3%4%5%6%7%8%9%Gadolinium ConcentrationHUCU HUCU660 HUCU935 Figure 4-7. The effect of increasing the gadolinium concentration of 14 gadolinium rods on HUCU at 0 GWD/STU The Importance of Gadolinium Rod Location The lattice power distribution is never uniform. The effectiveness of gadolinium to suppress power while increasing HUCU### was highly dependent upon the location in the lattice in which the gadolinium was placed. Since many parameters contributed to the power distribution within the lattice, determining an optimum location for gadolinium placement resulted from satisfying all the parameters that were most limiting. The power peaking distributions for the HU, CU### and CC states differed therefore determining an optimum location for placing gadolinium involved placement in areas that maximized

PAGE 64

52 improvement to the most limiting state without violating the parameter requirements of the other states. Two gadolinium rods were placed in series of different locations throughout the lattice to determine the areas in which HUCU### and HUCC could be maximized (two gadoliniums rods were used in order to preserve mirror symmetry of the lattice design). Figure 4-8 presents the locations examined and corresponding case numbers of the gadolinium location tests. 1234567891012145113123681012134379WW145469WW1517658WW161820710WW1619218111214151821222391317202310 Figure 4-8. Gadolinium rod placement diagram. The effectiveness of the gadolinium placement was a function of the power distribution. The power distribution was a function of the moderation ability and poison geometry. In the CC state as gadolinium rods were placed further from the edges of the control blade the worth of the gadolinium increased. This was due to the decreased competition for neutrons between the gadolinium and the control blade. If the

PAGE 65

53 gadolinium was placed to close to the control blade then effectively the gadolinium and blade were spatially self-shielding each other and therefore decreased both of the poisons effective worth. In the CU### case, the difference in worth of a certain gadolinium rod location was not a function of boron poison geometry (assuming no clumping) because the boron poison geometry was uniform; however, the difference in worth of certain gadolinium rod locations was related to the moderation capability of the fuel lattice. Areas of the lattice exhibiting more moderation created more thermal neutrons, leading to a higher power peaking. Gadolinium was worth more in these areas of increased moderation capability due to the increased amount of thermal neutrons available for absorption. Figure 4-9 displays the difference in gadolinium worth as a function of gadolinium location corresponding to the patterns in figure 4-8. Certain pattern changes caused opposing worth differences in different temperature and boron geometry states. In the change from pattern 4 to pattern 5 and in the change from pattern 15 to pattern 16, CC gadolinium worth increased while CU660 and HU gadolinium worth decreased. In the change from pattern 10 to pattern 11, CC gadolinium worth greatly decreased while CU660 worth slightly decreased and HU worth slightly increased. Placing a gadolinium rod in a higher power peaked area resulted in up to a 5% increase in gadolinium worth. In the CC case, increasing the distance of the gadolinium rod from the center of the control blade increased gadolinium worth by 0.00525 until the water rods in the center of the lattice were reached. Once the water rods were reached in the CC lattice (on the lower right diagonal half of the lattice), the power distribution and gadolinium rod worth become independent of the effects of the control blade.

PAGE 66

54 Altering a current lattice design to improve SDM and SLCS while maintaining HU must include shifting gadolinium locations where both the CU### and CC states have the most significant worth improvement while HU only has slight gadolinium worth increase. Improving CC gadolinium worth involves placing the gadolinium away from the control blade so that the gadolinium does not compete for neutrons with the boron in the control blade for. Enhancing the CU### worth involves spacing out the gadolinium so that spatial self-shielding does not occur and placing the gadolinium in the areas of highest power peaking exhibited by the cold power shape (areas of greatest moderation capabilities). Enhancing the HU worth also involves spacing out gadolinium and placing them in areas of highest power peaking corresponding to the HU power shape. The optimum gadolinium pattern for any given amount of gadolinium rods is the design that meets all three of these criteria. -0.05-0.045-0.04-0.035-0.03-0.025-0.02-0.015-0.01-0.005001234567891011121314151617181920212223PatternGad Worth (dk/k) HU Gad Worth CU0 Gad Worth CU660 Gad Worth CC Gad Worth Figure 4-9. Gadolinium worth versus location for 0 GWD/STU, 7% gadolinium concentration,

PAGE 67

55 Fuel Lattice Design Conclusions Certain parameters in the 2-dimensional fuel lattice design have a significant contribution to HUCU###. Lattice enrichment may be utilized to enhance HUCU###. While local lattice enrichment perturbations do not contribute to HUCU###, decreasing the lattice average enrichment decreases the power distribution skewing of the lattice and therefore increases HUCU###. However, with the demand for increased cycle lengths at higher powers, unfortunately higher average enrichments are needed to meet these requirements. Therefore with the needed increased average enrichment of the bundles will result in an increase in power skewing of the lattice thus decreasing HUCU###. The addition of competing thermal neutron poisons decreases HUCU###. Fuel designs with a tighter pitch between fuel rods lead to an increased production of plutonium. Plutonium is a competing thermal neutron poison. Therefore creating smaller fuel rods with a tighter pitch may increase the heat transfer of the fuel bundle, but also wa greater amount of plutonium is generated and therefore HUCU### is compromised. Also utilization of mixed oxide fuels (MOX) introduces an increased plutonium inventory in the core therefore further decreasing HUCU###. Gadolinium is also a thermal neutron poison. Enough positive reactivity must be installed into the reactor core at the beginning of cycle in order to meet the cycle length requirements. Gadolinium must installed in each of the fuel bundles in order to make sure that with the installed reactivity the reactor core is critical throughout operation. Therefore though gadolinium competes for thermal neutrons thereby decreasing HUCU###, it is a necessary component of reactor operation. In order to enhance HUCU### while maintaining a cycle operation goal, only certain fuel lattice parameters may be varied. Cycle length and power level is dependent

PAGE 68

56 upon installed reactivity; therefore average enrichment of the fuel is a fixed parameter if the number of fresh bundles utilized in the design is fixed. Plutonium production is related to power level, fuel lattice pitch, and isotope content of the fuel. In most cases all of those are fixed. The only design parameter with room for enhancement is gadolinium. If gadolinium is utilized effectively in certain locations of the fuel bundle, maximized differences between the HU and CU### may be created; therefore HUCU### is improved leading to an improvement in SLCS on the full core level. However, in increased average enrichment cores more gadolinium rods are needed to meet critical eigenvalue requirements; therefore resulting in fewer locations to manipulate gadolinium rod placement for improvement in HUCU###. Therefore if gadolinium has already been placed in the areas of maximized HUCU### enhancement further techniques must be utilized on the full core level to enhance HUCU###.

PAGE 69

CHAPTER 5 FULL CORE SLCS MODELING The reference base reactor core analyzed was a generic BWR/3. The reactor core was quarter core symmetric meaning that only a quarter of the full core had to be modeled to accurately represent characteristics of the full core. Figure 5-1 was the base reference core in which all perturbations were compared. Two different average bundle enrichments of fresh fuel were loaded into the core for the investigated cycle. The high enrichment bundles (fuel type 19) were 4.18% enriched, and the low enriched bundles (fuel type 20) were 3.89% enriched. Three major types of perturbations utilized to enhance SLCS were investigated on the full core level. The perturbations were selected based on the knowledge generated from the lattice physics analysis, and fell into two distinct characteristic types. The first type involved making a perturbation to the entire bundle. Gadolinium rod placement perturbations to the entire bundle were examined in order to determine the maximum achievable enhancement to SLCS. The second type involved perturbing the axial power shape. Based on the fact that average enrichment was also a dominant parameter in enhancing HUCU### in the lattice physics calculations, axial power shaping techniques utilizing enrichment differencing in certain axial zones was next examined to determine the maximum achievable enhancement to SLCS by perturbing the cold axial power shape. Gadolinium insertion in certain axial zones was also examined to determine if this method was also effective in enhancing SLCS by perturbing the axial power shape utilizing a poison. 57

PAGE 70

58 33899 733557 632011 631230 629071 633219 630031 626946 625807 726128 725187 733611 732365 630011 730073 717908 1715550 1615517 1616187 1615726 1633451 628541 629470 727398 727574 717496 160 190 190 200 1933992 633626 625332 725539 618506 1718019 170 190 1918275 160 2016968 1633476 633533 618102 1717534 1615903 1617983 170 190 1918023 170 2018156 170 2034030 728585 625349 717563 1617073 170 200 190 1918657 170 2018564 160 2018368 1632422 629444 725517 615879 160 2018404 160 1918791 170 2017545 170 2018513 160 2029955 727372 718511 1717980 170 190 1918610 170 2018586 170 2017211 160 2018702 1633233 630085 727593 718030 170 190 1918809 170 2018404 160 2018625 170 2018470 170 2033890 730081 617919 1717512 160 190 1918661 170 2018590 170 2017333 170 2018608 170 2017902 1733527 626959 615552 160 190 1918024 170 2017542 170 2018629 170 2017094 170 2018054 170 2031928 625804 715484 160 1918292 160 2018558 160 2017214 160 2018646 170 2017372 170 2018539 1731253 626087 716242 160 200 2018166 170 2018537 160 2018472 170 2018059 170 2018176 170 2029057 625179 715725 160 1916977 160 2018402 160 2018682 160 2017898 170 2018529 170 2018485 17 Fresh BundleOnce Burnded BundleTwice Burned Bundle XXXX ZZZZ Bundle Ex p osureBundle T yp e Figure 5-1. Reference base core fuel bundle loading map. In each perturbation case the critical eigenvalue, thermal limits, and SDM were monitored in order to determine if the enhancement to SLCS would violate the requirements of these margins. Calculated eigenvalue was monitored for each case to determine if calculated eigenvalue varied more than 0.001 k from the base case critical

PAGE 71

59 eigenvalue. At BOC rods patterns may always be adjusted in order to make the reactor critical and have critical eigenvalue deviate less than 0.001 k; however, at EOC when all the rods are pulled out of the core and no other form of positive reactivity may exist in the core to supply reactivity for criticality any decrease in critical eigenvalue as compared from the base case resulted in loss of cycle exposure and decreases of cycle energy. MFLPD was observed to make sure no perturbation resulted in a MFLPD greater than 0.909 as the BWR design basis requires. MFLCPR was also monitored to be certain that no perturbation caused a MFLCPR to become greater than the design basis requirements of 0.930. A SLCS enhancement that leads to decreased cycle energy and results in loss of cycle exposure was unacceptable due to the $ 1,000,000 at day cost involved in shutting the reactor down early. Furthermore, a core that does not meet thermal limits may not be licensed; therefore though SLCS may be improved through a certain modification, if that modification leads to unacceptable thermal margin, the core will not be licensed to operate. The optimum enhancement for SLCS involves an enhancement that meets thermal limits and does not deplete EOC calculated eigenvalue. Enhancing SLCS by Perturbing the Location of Gadolinium Rods A gadolinium placement modification to certain lattice axial zones was made in each fresh fuel type separately. The lattice physics calculations ensured that HUCU### improved with enhanced gadolinium location loading; therefore applying the technique to each fuel type individually determined the limiting effects of core radial and axial power weighting incurred on the enhancement of SLCS. The gadolinium rods that were interchanged were chosen based on the fact that the cold power peaking map displayed an

PAGE 72

60 increased worth for the new rod locations while the hot power peaking displayed a lesser change in worth. Figure 5-2 displays the base DOM gadolinium geometry and the areas circled in red correspond to where the gadolinium pins were interchanged with normal fuel pins in the perturbed cases. These perturbations were made to each axial zone individually and then to the entire bundle for each fresh bundle type. ABCDEFGHIJ11.602.803.203.953.953.953.953.953.952.8022.802.803.203.953.603.958.003.954.406.003.953.9533.203.204.403.004.404.906.004.404.406.004.404.406.004.9043.953.954.404.908.003.95WR-4.904.904.9053.953.604.906.003.954.90--4.904.904.9063.953.958.004.40WR-4.904.904.904.908.004.9073.953.954.406.00--4.904.904.908.004.904.9083.954.406.004.404.904.904.904.908.004.904.908.004.9093.953.954.406.004.904.904.908.004.904.908.004.904.90102.803.954.904.904.904.904.904.904.903.20 #.###.## #.## % U235 Enrichment % Gadolinium Enrichment Figure 5-2. The perturbation diagram for the gadolinium rod perturbation cases. SLCS margin was improved by the greatest amount when every axial zone containing gadolinium was perturbed. Figure 5-3 displays the maximum SLCS enhancement exhibited by each bundle type perturbation. Case 1 and Case 2 represent perturbations made to every axial zone in the bundle containing gadolinium. Case 1 represents when the perturbation was only made to the high enrichment bundles, and Case 2 represents when the perturbation was only made to the low enrichment bundles. Case 1 exhibited a 0.0095 improvement in BOC SLCS margin while Case 2 exhibited a 0.0341 increase in BOC SLCS margin. The low enrichment bundles represented 69% of the total loaded batch fraction and the majority of these bundles

PAGE 73

61 resided in the high power peak locations in the interior core region; therefore any perturbations made to these bundles demonstrated a more pronounced enhancement than perturbations made to the high enrichment bundles. 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSLCS Base Case 1 Case 2 Most Limiting Base SLCS Figure 5-3. Exposure dependent SLCS for the reference base case, the case in which the perturbation was made to only all of the fresh low enrichment bundles (case 2), and the case in which the perturbation was made to only all of the fresh high enrichment bundles (case 1). Though SLCS margin was enhanced, other limiting factors were greatly affected. Table 5-1 displays the effects of the enrichment perturbation on eigenvalue, and highlighted in red are the points in which eigenvalue deviated more than 0.001 k from the critical eigenvalue. All exposure points ending in an A represented the exposure step in which the control blade was shifted into the next pattern configuration. Due to the slight increase in gadolinium utilization caused by the perturbation, BOC eigenvalue decreased; and due to the saved positive reactivity from BOC, mid-cycle eigenvalue increased. In order to increase BOC eigenvalue and decrease mid-cycle eigenvalue the

PAGE 74

62 control blade patterns were manipulated to offset this reactivity imbalance. Table 5-2 displays the exposure dependent MFLPD. Case 2 improved most limiting MFLPD below the most limiting base case MFLPD after the rod pattern adjustment. Table 5-3 displays the exposure dependent MFLCPR. The most limiting MFLCPR in case 2 was also improved below the base case after the rod pattern adjustment. Therefore after the control blade adjustments were made both cases could meet critical eigenvalue requirements, but only case 2 could also meet MFLPD requirements as well. Table 5-1. Critical eigenvalue at specified exposure points for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS. Exposure Critical Eigenvalue (MWD/STU) Base Case 1 Case 1 Fix Case 2 Case 2 Fix 0 1.0136 1.0123 1.013 1.0104 1.0129 181 1.0142 1.0129 1.0139 1.011 1.0146 907 1.0141 1.0129 1.0139 1.0111 1.0138 1814 1.0128 1.0117 1.0125 1.0101 1.0124 2722 1.0133 1.0123 1.0131 1.0108 1.0132 2722A 1.0122 1.0112 1.0112 1.0095 1.0126 3629 1.0128 1.0119 1.0119 1.0104 1.0136 4536 1.0118 1.0111 1.0111 1.0098 1.012 5443 1.0126 1.012 1.0121 1.0108 1.0134 5443A 1.0119 1.0113 1.0114 1.0103 1.012 6350 1.013 1.0126 1.0126 1.0116 1.0135 7258 1.012 1.0117 1.0118 1.0108 1.0125 8165 1.0134 1.0133 1.0134 1.0126 1.0135 8165A 1.0123 1.0122 1.0123 1.0115 1.0123 9072 1.0134 1.0133 1.0134 1.0129 1.0137 9979 1.0135 1.0135 1.0136 1.0134 1.0138 10886 1.0148 1.0149 1.0149 1.0153 1.0152 10886A 1.0147 1.0148 1.0148 1.0152 1.0152 11794 1.0149 1.0153 1.0152 1.0163 1.0155 12570 1.0163 1.0168 1.0167 1.0182 1.0169 12701 1.0157 1.0164 1.0162 1.0179 1.0164 13608 1.0159 1.0168 1.0165 1.0186 1.0167 13608A 1.0169 1.0178 1.0175 1.0195 1.0176 14061 1.0159 1.0168 1.0165 1.0188 1.0167 14334 1.0164 1.0174 1.0171 1.0195 1.0171 14570 1.0163 1.0173 1.017 1.0193 1.0169

PAGE 75

63 Table 5-2. Exposure dependent MFLPD for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS. Exposure MFLPD (MWD/STU) Base Case 1 Case 1 Fix Case 2 Case 2 Fix 0 0.846 0.829 0.815 0.866 0.821 181 0.845 0.85 0.832 0.858 0.799 907 0.806 0.811 0.796 0.82 0.779 1814 0.807 0.814 0.803 0.821 0.786 2722 0.781 0.789 0.778 0.795 0.763 2722A 0.742 0.753 0.754 0.753 0.712 3629 0.714 0.725 0.724 0.729 0.712 4536 0.717 0.726 0.726 0.734 0.702 5443 0.703 0.711 0.71 0.72 0.682 5443A 0.838 0.842 0.84 0.858 0.822 6350 0.849 0.857 0.853 0.878 0.824 7258 0.847 0.865 0.859 0.884 0.825 8165 0.889 0.917 0.912 0.923 0.886 8165A 0.847 0.865 0.859 0.892 0.841 9072 0.852 0.863 0.863 0.887 0.878 9979 0.784 0.786 0.791 0.806 0.836 10886 0.683 0.686 0.685 0.699 0.723 10886A 0.72 0.724 0.729 0.717 0.755 11794 0.704 0.713 0.711 0.718 0.706 12570 0.704 0.709 0.706 0.73 0.693 12701 0.704 0.708 0.705 0.736 0.691 13608 0.854 0.855 0.851 0.878 0.852 13608A 0.734 0.742 0.738 0.762 0.737 14061 0.841 0.84 0.834 0.871 0.837 14334 0.746 0.748 0.741 0.789 0.736 14570 0.763 0.766 0.759 0.805 0.753 Utilizing any type of gadolinium insertion will only enhance BOC SLCS. Though BOC SLCS was enhanced in case 2 by gadolinium location improvement, once the gadolinium burned out (~11,000 MWD/STU for this specific case) the perturbed case became limiting again in SLCS. Therefore if SLCS is enhanced utilizing this method, the designer must consider if the gadolinium burn out point is acceptable as well. If at the gadolinium burn out point SLCS is not acceptable in magnitude, another method must be utilized to enhance SLCS.

PAGE 76

64 Table 5-3. Exposure dependent MFLCPR for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS. Exposure MFLCPR (MWD/STU) Base Case 1 Case 1 Fix Case 2 Case 2 Fix 0 0.711 0.723 0.727 0.707 0.714 181 0.721 0.729 0.727 0.713 0.715 907 0.726 0.733 0.733 0.717 0.72 1814 0.737 0.744 0.741 0.731 0.721 2722 0.741 0.748 0.744 0.734 0.727 2722A 0.73 0.737 0.737 0.717 0.744 3629 0.736 0.743 0.743 0.725 0.754 4536 0.729 0.734 0.734 0.72 0.736 5443 0.734 0.738 0.738 0.726 0.744 5443A 0.773 0.77 0.77 0.778 0.767 6350 0.775 0.774 0.774 0.78 0.769 7258 0.731 0.733 0.734 0.734 0.735 8165 0.732 0.732 0.732 0.731 0.737 8165A 0.745 0.744 0.744 0.744 0.75 9072 0.746 0.743 0.743 0.748 0.749 9979 0.754 0.752 0.751 0.758 0.752 10886 0.767 0.766 0.764 0.773 0.763 10886A 0.757 0.756 0.754 0.766 0.755 11794 0.79 0.793 0.792 0.789 0.785 12570 0.824 0.828 0.828 0.838 0.817 12701 0.825 0.83 0.829 0.836 0.818 13608 0.829 0.836 0.834 0.841 0.821 13608A 0.825 0.833 0.831 0.827 0.817 14061 0.822 0.83 0.828 0.836 0.815 14334 0.822 0.818 0.814 0.84 0.816 14570 0.808 0.803 0.8 0.825 0.802 Enhancing SLCS at the cost of SDM was not an acceptable option if SDM was already a limiting constraint from the original design. Figure 5-4 demonstrates that the modifications made to the low enriched fresh bundles did not diminish SDM below the most limiting value of the base case. For this perturbation, BOC SDM was improved 0.0190 at the cost of decreasing EOC SDM; however, the decrease in EOC SDM did not fall below the most limiting SDM value, and therefore the perturbation yielded acceptable SDM consequence.

PAGE 77

65 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSDM Base Case 1 Case 2 Most Limiting Base SDM Figure 5-4. Exposure dependent SDM for the gadolinium rod location perturbation cases. Utilizing a gadolinium location perturbation may be utilized to enhance BOC SLCS. However, the magnitude of the improvement is limited by the ability to separate the gadolinium and move it into areas of greater effective worth. As the amount of gadolinium rods in the lattice increases, the ability to move gadolinium rods to more effective locations decreases. As fuel bundle designs move to higher average enrichments more gadolinium rods are needed in the fuel bundles to counteract the increased installed reactivity. More gadolinium rods in the fuel bundle causes this technique to be less effective due to the inability to move the gadolinium to locations of greater effective worth due to the space constraints of the fuel lattice. Axial Power Shape Characteristics The base case most limiting power peak bundle location was bundle (15, 11). Because lattice physics work determined that the DOM was the most limiting geometry for HUCU###, decreasing the cold power peak in the DOM decreases SLCS.

PAGE 78

66 The cold power shape and the hot power shape were not the same. Figure 5-5 is the most limiting radial power peaking base case axial power distribution relative to its radial power peaking. Superimposed in blue over figure 5-5 is an example of the cold power shape. Though not to scale, the superimposition displays the difference in where the power peaking resides in the two conditions. The BOC cold power shape was a cosine shape peaked in the DOM, and the hot power shape was a modified Bessel function peaked in the PSZ. Therefore in the axial power shape perturbations this characteristic was utilized to maximize SLCS. 016324864809611212814416000.20.40.60.811.21.41.61.8Relative Power PeakAxial Location (in.) Base (15,11) (15,10) (14,11) (15,12) Example Cold Power Shape Calculated Hot Power Shape Figure 5-5. The base case hot axial power shape with superimposed cold axial power shape. Enhancing SLCS through Axial Power Shaping Utilizing Enrichment Differencing The position of the cold axial power peak is the axial portion of the fuel bundle exhibiting the least amount of power suppression in the cold condition; therefore the cold axial power peak region is also the most limiting HUCU### region. As previously displayed in figure 3-2, the DOM was the most limiting region for HUCU### due to the decreased availability of borated water locations in the lattice. The VAN exhibited the

PAGE 79

67 greatest HUCU### due to the increased availability to place borated water in the lattice as a result of the vanished rod locations. Since the region of the cold axial power peak was the most limiting in HUCU###, shifting the cold axial power peak out of the DOM and into the VAN should increase SLCS. The cold axial power peak may be decreased utilizing enrichment differencing in the DOM and VAN. Figure 5-6 illustrates the goal of enrichment differencing. By increasing the enrichment in the VAN and decreasing the enrichment in the DOM, the DOM axial power peak is decreased thereby increasing SLCS. N-T N-V VAN PLE DOM PSZ NAT Base Case Power Peak Decreased Power Peak in DOM Zone Base CasePerturbed Case Utilizin g Axial Enrichment Differencin g Figure 5-6. Cold axial power shape perturbation diagram. Axial power shaping, utilizing enrichment differencing, enhanced SLCS at BOC and at the gadolinium burn out exposure point. Figure 5-7 displays the SLCS

PAGE 80

68 enhancement utilizing axial power shape perturbation by enrichment differencing as well as SLCS enhancement utilizing gadolinium placement perturbations. Case 11 was the case that utilized enrichment differencing. Gadolinium perturbations enhanced SLCS only at BOC; however, axial power shape perturbations utilizing enrichment differencing in the DOM and VAN enhanced SLCS at both BOC and the gadolinium burn out exposure point. The BOC SLCS margin enhancement utilizing enrichment differencing was 0.035 and the gadolinium burn out exposure point enhancement was 0.0142. Therefore if improvement in SLCS is necessary in both BOC and gadolinium burn out exposure point the enrichment differencing method is the preferred method. 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSLCS Base Case 2 Case 11 Most Limiting Base SLCS Figure 5-7. Exposure dependent SLCS for the gadolinium location perturbation case (case 2) and axial power shape perturbation utilizing enrichment differencing case (case 11) that exhibited the greatest enhancement in SLCS. In the gadolinium perturbation case, the fresh bundles reached a maximum peak power point once the gadolinium burned out. Once the gadolinium had burned out, the main parameters affecting the cold power shape were the enrichment distribution and

PAGE 81

69 axial leakage of the bundle. Since the axial leakage of the fuel bundle was a function of axial height (a fixed parameter) and controlled utilizing top and bottom natural zones, axial enrichment distribution was the main mode for altering the power shape at the gadolinium burn out exposure point. Increasing the enrichment distribution in the VAN and decreasing the enrichment in the DOM resulted in a decreased DOM cold power peak at the gadolinium burn out exposure point due to the decreased availability of enrichment in the DOM. Figure 5-8 presents the SLCS enhancement as a function of enrichment difference between the DOM and VAN. The maximum amount of SLCS margin enhancement for BOC and the gadolinium burn out exposure point occurs at 0.30% enrichment difference between the DOM and VAN. 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSLCS Base (0.01) Case 17 (0.35) Case 11 (0.31) Case 18 (0.20) Case 19 (0.10) Case 20 (0.04) Most Limiting Base SLCS Figure 5-8. SLCS enhancement utilizing different magnitudes of enrichment differencing between the DOM and VAN.

PAGE 82

70 An optimum enrichment difference arises from the fact that reactivity worth is a flux weighted. As the enrichment was increased in the VAN, the flux increased in the VAN; and as the enrichment decreases in the DOM, the flux decreases in the DOM. Therefore the 0.30 enrichment difference represented the optimum decrease in flux weighting of the DOM and increase in flux weighting of the VAN that resulted in the greatest average cold borated negative reactivity insertion. Axial power shaping utilizing enrichment differencing alters the hot power shape and mode in which the core burns. The BOC hot axial power profile utilizing enrichment differencing is displayed in Figure 5-9. By increasing the VAN zone enrichment while decreasing the DOM zone enrichment, the DOM and PSZ zones exhibited a decrease in power peak while the VAN zone experiences and increase in power peaking. 016324864809611212814416000.20.40.60.811.21.41.61.8Relative Power PeakAxial Location (in.) Base (15,11) (15,10) (14,11) (15,12) Decreased Power Peak In DOM and PSZ Zones Increased Power Peak In VAN Zone Figure 5-9. The hot axial power shape for maximum SLCS enhancement utilizing enrichment differencing.

PAGE 83

71 The change in axial power shape slightly altered the calculated eigenvalue and thermal margins. Table 5-4 demonstrates that critical eigenvalue of the enrichment perturbations case did not vary more than 0.001 k from the base case critical eigenvalue; therefore utilizing this method did not warrant a rod pattern adjustment. The final calculated eigenvalue for the enrichment differencing case was 0.0007 k less than the critical base case eigenvalue; however, the increased mid cycle energy created could have been suppressed by utilizing a rod pattern adjustment if determined necessary. Table 5-4. Critical eigenvalue at specified exposure points for the gadolinium rod location perturbation case and enrichment differencing case that exhibited greatest enhancement in SLCS. Exposure Critical Eigenvalue (MWD/STU) Base Case 2 Fix Case 11 0 1.0136 1.0129 1.0142 181 1.0142 1.0146 1.0143 907 1.0141 1.0138 1.0143 1814 1.0128 1.0124 1.0128 2722 1.0133 1.0132 1.0133 2722A 1.0122 1.0126 1.013 3629 1.0128 1.0136 1.0136 4536 1.0118 1.012 1.0125 5443 1.0126 1.0134 1.0134 5443A 1.0119 1.012 1.0121 6350 1.013 1.0135 1.0133 7258 1.012 1.0125 1.0127 8165 1.0134 1.0135 1.0142 8165A 1.0123 1.0123 1.013 9072 1.0134 1.0137 1.0141 9979 1.0135 1.0138 1.014 10886 1.0148 1.0152 1.0153 10886A 1.0147 1.0152 1.0152 11794 1.0149 1.0155 1.015 12570 1.0163 1.0169 1.0162 12701 1.0157 1.0164 1.0156 13608 1.0159 1.0167 1.0153 13608A 1.0169 1.0176 1.0163 14061 1.0159 1.0167 1.0151 14334 1.0164 1.0171 1.0157 14570 1.0163 1.0169 1.0156

PAGE 84

72 MFLPD for the axial power shaping utilizing enrichment differencing case also was under the acceptable limit for all exposure points through out the cycle. Table 5-5 displays that most limiting MFLPD decreased by 0.026 from the base case. Therefore utilizing this technique improves the MFLPD of the cycle. Table 5-5. Exposure dependent MFLPD for the gadolinium rod location perturbation case and the enrichment differencing case that exhibited greatest enhancement in SLCS. Exposure MFLPD (MWD/STU) Base Case 2 Fix Case 11 0 0.846 0.821 0.825 181 0.845 0.799 0.834 907 0.806 0.779 0.792 1814 0.807 0.786 0.799 2722 0.781 0.763 0.771 2722A 0.742 0.712 0.718 3629 0.714 0.712 0.7 4536 0.717 0.702 0.699 5443 0.703 0.682 0.681 5443A 0.838 0.822 0.827 6350 0.849 0.824 0.83 7258 0.847 0.825 0.819 8165 0.889 0.886 0.863 8165A 0.847 0.841 0.818 9072 0.852 0.878 0.839 9979 0.784 0.836 0.793 10886 0.683 0.723 0.69 10886A 0.72 0.755 0.735 11794 0.704 0.706 0.707 12570 0.704 0.693 0.692 12701 0.704 0.691 0.691 13608 0.854 0.852 0.829 13608A 0.734 0.737 0.72 14061 0.841 0.837 0.805 14334 0.746 0.736 0.718 14570 0.763 0.753 0.728 Table 5-6 shows no significant difference realized in MFLCPR as compared with the base case. Therefore utilizing this technique does not deplete the cores to meet any thermal margin requirements.

PAGE 85

73 Table 5-6. Exposure dependent MFLCPR for the gadolinium rod location perturbation cases that exhibited greatest enhancement in SLCS. Exposure MFLCPR (MWD/STU) Base Case 2 Fix Case 11 0 0.711 0.714 0.724 181 0.721 0.715 0.719 907 0.726 0.72 0.725 1814 0.737 0.721 0.734 2722 0.741 0.727 0.741 2722A 0.73 0.744 0.742 3629 0.736 0.754 0.75 4536 0.729 0.736 0.737 5443 0.734 0.744 0.744 5443A 0.773 0.767 0.763 6350 0.775 0.769 0.767 7258 0.731 0.735 0.741 8165 0.732 0.737 0.741 8165A 0.745 0.75 0.753 9072 0.746 0.749 0.751 9979 0.754 0.752 0.756 10886 0.767 0.763 0.763 10886A 0.757 0.755 0.753 11794 0.79 0.785 0.787 12570 0.824 0.817 0.821 12701 0.825 0.818 0.821 13608 0.829 0.821 0.825 13608A 0.825 0.817 0.82 14061 0.822 0.815 0.818 14334 0.822 0.816 0.821 14570 0.808 0.802 0.807 The effect of axial enrichment differencing had a similar effect on SDM as the lattice geometric placement perturbation. Figure 5-10 displays the exposure dependence effects of these perturbations on SDM at different exposure points. SDM for both perturbations did not fall below the most limiting base case value; therefore both enhancements may be utilized to enhance SLCS if BOC SDM were to be in a limiting condition. However, only axial power shaping utilizing enrichment differencing also improves the gadolinium burn out exposure point limiting condition.

PAGE 86

74 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSDM Base Case 2 Case 11 Most Limiting Base SDM Figure 5-10. Exposure dependent SDM for the gadolinium rod location perturbation case and axial power shaping utilizing enrichment differencing case. Enhancing SLCS by Means of Axial Power Shaping Utilizing Gadolinium Insertion Enrichment differencing was not the only method for perturbing the axial power shape in order to decrease the power peak in the DOM. Axial power shaping from the utilization of an additional gadolinium pellets in certain axial zones was also analyzed in order to determine if the negative reactivity insertion from adding additional gadolinium was more favorable than shifting the axial enrichment distribution to create similar types of perturbations. The concept of utilizing the negative reactivity of a gadolinium rod insertion to decrease the power peak in the most limiting zone was similar to the concept of decreasing enrichment in the most limiting axial zone. In both cases a negative reactivity insertion in the limiting axial zone caused the power to peak to decrease in that zone in which the gadolinium was inserted thus decreasing the worth of that axial zone to SLCS.

PAGE 87

75 As displayed in figure 4-5 increasing the amount of rods in a lattice decreased HUCU###; however, k of the lattice also decreased therefore leading to an improvement of SLCS due to the decreased cold k. Figure 5-11 displays the gain in BOC SLCS utilizing a gadolinium rod insertion as compared with the other types of perturbations examined. Case 26 represented a gadolinium insertion made in the PSZ, and Case 27 represented a gadolinium rod insertion made in the DOM. Because the cold power shape peaks in the DOM, inserting a gadolinium rod into the PSZ does not have as drastic of an effect on SLCS as placing a gadolinium rod in the DOM. 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSLCS Base Case 2 Case 11 Case 26 Case 27 Most Limiting Base SLCS Figure 5-11. Exposure dependent SLCS for the gadolinium location perturbation case (case 2), axial power shape perturbation utilizing enrichment differencing case (case 11), inserting a gadolinium rod in the PSZ (case 26) and inserting a gadolinium rod in DOM (case27) that exhibited the greatest enhancement in SLCS. The DOM gadolinium rod insertion yielded the greatest increase in BOC SLCS as compared with the other perturbations. Inserting a gadolinium rod in the DOM enhanced SLCS by 0.422 while inserting a gadolinium in the PSZ only enhanced SLCS by 0.189.

PAGE 88

76 Table 5-7. Critical eigenvalue at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS. Exposure Critical Eigenvalue (MWD/STU) Base Case 26 Case 26f Case 27 Case 27f 0 1.0136 1.0127 1.0133 1.0116 1.013 181 1.0142 1.0129 1.0136 1.012 1.0139 907 1.0141 1.0131 1.0138 1.0121 1.0141 1814 1.0128 1.012 1.012 1.0111 1.0121 2722 1.0133 1.013 1.013 1.0121 1.0131 2722A 1.0122 1.0121 1.0121 1.0112 1.0119 3629 1.0128 1.0129 1.0129 1.0122 1.013 4536 1.0118 1.0121 1.0121 1.0116 1.0117 5443 1.0126 1.013 1.013 1.0126 1.0128 5443A 1.0119 1.0122 1.0122 1.0118 1.0119 6350 1.013 1.0133 1.0133 1.0129 1.0131 7258 1.012 1.0124 1.0125 1.0119 1.0121 8165 1.0134 1.0139 1.014 1.0133 1.0136 8165A 1.0123 1.0128 1.0128 1.0122 1.0124 9072 1.0134 1.0137 1.0137 1.0132 1.0134 9979 1.0135 1.0135 1.0135 1.0134 1.0135 10886 1.0148 1.0146 1.0145 1.0149 1.0147 10886A 1.0147 1.0145 1.0145 1.0148 1.0147 11794 1.0149 1.0143 1.0142 1.0151 1.0147 12570 1.0163 1.0155 1.0161 1.0165 1.016 12701 1.0157 1.0148 1.0154 1.0159 1.0154 13608 1.0159 1.0147 1.0151 1.016 1.0154 13608A 1.0169 1.0156 1.0162 1.017 1.0164 14061 1.0159 1.0146 1.0149 1.016 1.0153 14334 1.0164 1.0151 1.0148 1.0167 1.016 14570 1.0163 1.015 1.0146 1.0165 1.0158 Placing negative reactivity into one region of the bundle without introducing positive reactivity into some other region will cause a decrease in BOC eigenvalue. Therefore a rod pattern change was utilized in this method in order to achieve acceptable BOC eigenvalue requirements. Table 5-7 displays the calculated eigenvalue as compared with the base case for inserting a gadolinium rod in the DOM zone and in the PSZ before. Inserting a gadolinium rod in the PSZ greatly reduced the BOC reactivity of the bundle; therefore the rod patterns had to be adjusted to meet the BOC condition. The

PAGE 89

77 decrease in integrated power realized in the bundle due to the gadolinium insertion in the PSZ caused the core to fall 0.0017 k short of EOC critical eigenvalue requirements. However, inserting a gadolinium rod in the DOM zone caused only a 0.0005 k decrease in EOC critical eigenvalue. Distorting the hot axial power shape altered the thermal margins of the core. Figure 5-12 and figure 5-13 displays the distorted hot axial power shape caused from inserting a gadolinium rod in the PSZ and in the DOM. 016324864809611212814416000.20.40.60.811.21.41.61.8Relative Power PeakAxial Location (in.) Base (15,11) (15,10) (14,11) (15,12) Figure 5-12. The hot axial power shape of the most power peaked fuel bundle caused by inserting a gadolinium rod into the PSZ. When inserting a gadolinium rod into the PSZ, the decreased power peak in the PSZ causes the hot axial power shape to flatten. When inserting a gadolinium rod into

PAGE 90

78 the DOM, the extreme decreased power peak in the DOM leads to an increased relative power peak in the PSZ. 016324864809611212814416000.20.40.60.811.21.41.61.8Relative Power PeakAxial Location (in.) Base (15,11) (15,10) (14,11) (15,12) Figure 5-13. The hot axial power shape of the most power peaked fuel bundle caused by inserting a gadolinium rod into the DOM. Table 5-8 displays MFLPD for the PSZ and DOM gadolinium insertions as compared to the base case. Inserting a gadolinium rod in the PSZ decreased BOC MFLPD by 0.083, and also decreased most limiting MFLPD by 0.033. However, inserting a gadolinium rod in the DOM yielded no significant enhancement in MFLPD. In both cases there was no significant alteration in MFLCPR as displayed in table 5-9. Therefore inserting an extra gadolinium rod into a certain axial zone of the fuel bundle does not hinder the ability to meet thermal margins.

PAGE 91

79 Table 5-8. MFLPD at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS. Exposure MFLPD (MWD/STU) Base Case 26 Case 26f Case 27 Case 27f 0 0.846 0.771 0.763 0.873 0.847 181 0.845 0.772 0.762 0.875 0.839 907 0.806 0.748 0.738 0.832 0.795 1814 0.807 0.768 0.768 0.829 0.811 2722 0.781 0.765 0.765 0.790 0.776 2722A 0.742 0.707 0.706 0.746 0.738 3629 0.714 0.710 0.709 0.716 0.707 4536 0.717 0.722 0.722 0.718 0.718 5443 0.703 0.698 0.697 0.703 0.700 5443A 0.838 0.827 0.825 0.840 0.836 6350 0.849 0.827 0.824 0.855 0.845 7258 0.847 0.816 0.813 0.853 0.840 8165 0.889 0.861 0.856 0.894 0.882 8165A 0.847 0.807 0.804 0.854 0.840 9072 0.852 0.841 0.839 0.849 0.850 9979 0.784 0.812 0.815 0.776 0.789 10886 0.683 0.710 0.713 0.687 0.688 10886A 0.72 0.757 0.761 0.715 0.727 11794 0.704 0.716 0.722 0.712 0.711 12570 0.704 0.677 0.684 0.714 0.705 12701 0.704 0.677 0.684 0.709 0.702 13608 0.854 0.838 0.866 0.841 0.831 13608A 0.734 0.703 0.720 0.749 0.741 14061 0.841 0.821 0.859 0.822 0.807 14334 0.746 0.719 0.710 0.744 0.737 14570 0.763 0.736 0.727 0.755 0.749 Therefore when a designer chooses to utilize this method for enhancing SLCS, the designer must decide which parameters are most necessary for achieving the required result. If the designer is experiencing limiting MFLPD and willing to compromise cycle energy to meet this requirement, then placing a gadolinium rod in the PSZ is the better choice. If the designer does not have limiting MFLPD, then inserting a gadolinium rod in the DOM zone is the better choice. Therefore the choice of one method or the other depends on the thermal margins and the critical eigenvalue requirements.

PAGE 92

80 Table 5-9. MFLCPR at specified exposure points for the gadolinium insertion into the DOM (case 27) and gadolinium insertion into the PSZ (case 26) that exhibited greatest enhancement in SLCS. Exposure MFLCPR (MWD/STU) Base Case 26 Case 26f Case 27 Case 27f 0 0.711 0.712 0.715 0.694 0.692 181 0.721 0.708 0.712 0.683 0.683 907 0.726 0.717 0.721 0.692 0.695 1814 0.737 0.728 0.728 0.711 0.709 2722 0.741 0.737 0.737 0.724 0.723 2722A 0.73 0.724 0.724 0.708 0.717 3629 0.736 0.733 0.733 0.723 0.732 4536 0.729 0.729 0.729 0.723 0.724 5443 0.734 0.738 0.738 0.732 0.734 5443A 0.773 0.772 0.772 0.771 0.77 6350 0.775 0.775 0.775 0.774 0.773 7258 0.731 0.741 0.742 0.73 0.733 8165 0.732 0.741 0.742 0.73 0.733 8165A 0.745 0.752 0.752 0.742 0.744 9072 0.746 0.748 0.748 0.745 0.744 9979 0.754 0.751 0.75 0.754 0.751 10886 0.767 0.758 0.757 0.768 0.763 10886A 0.757 0.748 0.747 0.758 0.753 11794 0.79 0.786 0.786 0.79 0.789 12570 0.824 0.819 0.818 0.824 0.823 12701 0.825 0.819 0.818 0.825 0.823 13608 0.829 0.824 0.822 0.829 0.826 13608A 0.825 0.818 0.815 0.825 0.822 14061 0.822 0.817 0.814 0.822 0.819 14334 0.822 0.81 0.805 0.824 0.815 14570 0.808 0.796 0.791 0.809 0.8 SDM was neither greatly enhanced nor greatly decreased utilizing gadolinium insertion. Figure 5-14 displays SDM as a function of exposure for the base case and all three types of perturbations. Therefore if a designer was limited in EOC SDM then inserting a gadolinium rod should be utilized in order to enhance BOC SDM. The decision to enhance SLCS margin by inserting a gadolinium rod into a certain axial zone of a fuel bundle is dependent upon the preexisting limiting conditions of the fuel design. If the fuel designer decides that maximizing BOC SLCS without concern for

PAGE 93

81 SLCS at the gadolinium burn out exposure point is the most limiting design characteristic, then inserting a gadolinium rod into the DOM will suffice as a solution to enhancing BOC SLCS margin. If the designer cannot afford loss in EOC SDM, and only needs a minimal improvement in thermal margins as well as minimally enhanced BOC SLCS, then adding a gadolinium in PSZ at the cost of cycle energy may be an adequate solution to enhancing BOC SLCS. 00.0050.010.0150.020.0250.030.0350.040200040006000800010000120001400016000MWD/STUSDM Base Case 2 Case 11 Case 26 Case 27 Most Limiting Base SDM Figure 5-14. Exposure dependent SDM for the gadolinium rod location perturbation case, axial power shaping utilizing enrichment differencing case and the axial power shaping utilizing gadolinium placement case. Enhancing BOC SLCS margin utilizing gadolinium perturbations may cause the gadolinium burn out exposure point to become the most limiting in SLCS. If SLCS at the gadolinium burn out exposure point is of acceptable magnitude, then the designer has utilized an acceptable technique for enhancing SLCS. If, however, SLCS at the gadolinium burn out exposure point is of unacceptable magnitude, then the gadolinium insertion techniques are not feasible methods for improving SLCS. Therefore the

PAGE 94

82 decision to utilize this method is solely dependent upon the limitations of the gadolinium burn out exposure point.

PAGE 95

CHAPTER 6 CONCLUSIONS SLCS is a core wide phenomenon that is dependent upon the HUCU### characteristics of each fuel bundle. Introducing fresh bundles into the core with inherently enhanced HUCU### characteristics will improve SLCS. HUCU### is improved by manipulating design parameters on the lattice design level as well as in the full core design. Therefore understanding the most limiting design parameters in both design aspects and the capability of those parameters to increase HUCU### is paramount to improving SLCS. The ability of a certain type of fuel lattice design perturbation to enhance SLCS was determined by the limiting characteristics of that lattice perturbation. HUCU### was highly dependent upon average enrichment. As average enrichment of the fuel bundle was increased HUCU### decreased thereby decreasing SLCS on the full core level. Increasing enrichment has a greater impact per percent increase of reactivity in the cold, collapsed void lattice state then in the hot Doppler broadened voided operating state. Localized enrichment perturbations did not affect HUCU### therefore when a designer creates a lattice with SLCS in mind they need only be concerned with the average enrichment of the lattice and not how the local enrichment is schemed. Gadolinium rods also had a significant impact on HUCU### lattice behavior and therefore significantly impacted SLCS. Gadolinium geometries that were clumped and incurred significant spatial self-shielding decreased HUCU### while gadolinium geometries that were spread out limiting the self-shielding exhibited an increased 83

PAGE 96

84 HUCU###. Increasing the amount of gadolinium rods in the fuel lattice decreased relative HUCU###; however, inserting the gadolinium also reduced k thereby actually improving SLCS by decreasing the worth of the bundle to the entire core. Therefore increasing the amount of gadolinium rods in the bundle had a diminishing return. Increasing the gadolinium concentration also decreased the relative HUCU###; however, increasing the gadolinium concentration also reduced k thereby also improving SLCS for the whole core. Optimum locations for gadolinium rod placement exist for certain amounts of gadolinium rods. These optimum placement locations are realized by understanding the difference in power peaking between the hot and cold homogenously enriched power shapes (the power shape realized explicitly from geometry of the fuel bundle and flux level) and placing gadolinium rods in areas where the difference in power peak between the two states is the greatest. After understanding the 2-dimensional lattice physics calculations, perturbations were made to fuel bundles in the full core simulator in order to determine effects on full core criticality and thermal limits. Perturbing the placement of the gadolinium rods in order to maximize gadolinium worth utilized in the cold borated condition improved BOC SLCS at the expense of decreased BOC critical eigenvalue. Therefore after perturbing gadolinium locations to maximize negative reactivity, the control blade patterns must be adjusted in order to introduce enough positive reactivity in the hot condition to meet the critical eigenvalue requirements. Perturbing the axial enrichment distribution in order to decrease the power peaking in the axial zone most limiting to HUCU### decreased that axial zones flux importance to the SLCS calculation and thereby improved both the BOC and the gadolinium burn out exposure point SLCS.

PAGE 97

85 However, utilizing this method causes an increased complexity in manufacturing of the bundle and therefore leading to an increased production cost. Inserting an extra gadolinium rod into a certain axial zone in order to also perturb the axial power shape improved BOC SLCS without decreasing EOC SDM. However, utilizing gadolinium perturbations only helped improve the BOC SLCS and did not enhance the gadolinium burn out exposure point SLCS. SLCS may always be improved by increasing the boron concentration or boron enrichment in the SLCS tank. However, if the utility is limited by time, cost or aggravation then utilizing an acceptable design technique in order to enhance SLCS margin is solely dependent upon the limiting characteristics of the core behavior and the acceptable sacrifice in margin of those parameters. Unfortunately, not all core situations will have a possible remedy for SLCS. The greatest increase in BOC SLCS utilizing any of the mentioned techniques was roughly 0.5% and the gadolinium burnout point maximum improvement was 0.14%. Therefore utilities exhibiting marginal SLCS fuel design difficulties that wish to have power output increases in their following cycles, increasing the average enrichment and gadolinium content in their core, will need to understand the limitations of the inherent fuel design. Utilities must then realize that an increase in boron concentration of their SLCS tank or utilizing enriched boron is needed if they wish to accommodate SLCS while not incurring the extra cost per cycle of loading extra bundles to flatten the power distribution and reduce SLCS.

PAGE 98

CHAPTER 7 FUTURE WORK The purpose of this study was to conduct a sensitivity analysis in order to determine limiting fuel design characteristics for SLCS. The methodology developed by Yasushi Hirano, Kazuki Hida, Koichi Sakurada and Munenari Yamamoto utilized a fixed gadolinium pattern and then generated an optimal enrichment distributions for a 2-dimensional BWR fuel lattice [10]. Since this study proved that radial enrichment distribution was not a factor in SLCS and that SLCS was only limited by average enrichment, the possibility exists to expand on the enrichment distribution tool and develop a tool that determines an optimum SLCS gadolinium placement for a given lattice average enrichment. The tool would basically compare homogenously enriched hot and cold lattice power distributions and determine an optimum gadolinium scheme based on the maximum difference in the two power distributions. Because the placement of the gadolinium for SLCS is basically decoupled from the enrichment distribution, this problem does not become over-constrained, and therefore it is possible to obtain an optimum gadolinium configuration for SLCS while creating an optimum enrichment distribution for thermal limit and fuel efficiency requirements. The optimum enrichment distribution methodology was also a 2-dimensional methodology. This study concluded that axial enrichment and gadolinium perturbations may be utilized to improve SLCS. In modern core design strategy 2-dimensional lattice calculations are completed and then the group constants from the 2-dimensional codes are utilized by the full core simulators because of computational time constraints and 86

PAGE 99

87 memory requirements of the processor. With computers getting faster and distributed parallel computing schemes becoming more optimized, core design may reach a point where full 3-dimensional bundles are modeled assuming an infinite bundle approximation (or a more brilliant scheme) to get group constants for the full core simulator. When this technology is available, utilizing the design criteria from this study for the SLCS portion, a full bundle axial and radial enrichment and gadolinium configuration optimization methodology may be devised that creates the optimum fuel bundle for SLCS, SDM, thermal margin and fuel utilization. This will create an automated core design environment thus freeing the designers time to allow for examination of other pressing issues in the design strategy.

PAGE 100

LIST OF REFERENCES 1. Aoyama, Mooto, Sadao Uchikawa and Renzo Takeda, Reactivity Control Method for Extended Burnup of Boiling Water Reactor Fuel Bundles, Journal of Nuclear Science and Technology, 26, pp.403-410, April 1989. 2. Cochran, Robert and Nicholas Tsoulfanidis, The Nuclear Fuel Cycle: Analysis and Management, American Nuclear Society, La Grange Park, Illinois, 1999. 3. Dresner, Lawerance, Resonance Absorption in Nuclear Reactors, Pergamon Press, New York, New York, 1976. 4. Duderstadt, James and Louis Hamilton, Nuclear Reactor Analysis, John Wiley & Sons, Inc., New York, New York, 1976 5. General Electric Company, TGBLA06A; General Electric Lattice Physics Method, DRF A00-05526, October, 1994. (Proprietary Information) 6. General Physics Corporation, BWR Generic Fundamentals: Chapter 9 Core Thermal Limits, Columbia, Maryland 1993. (Proprietary Information) 7. Glasstone, Samuel and Walter H. Jordan, Nuclear Power and its Environmental Effects, American Nuclear Society, La Grange Park, Illinois, 1980. 8. Global Nuclear Fuels, PANAC11 Users Manual, UM-0021 Rev. 1, February 2001. (Proprietary Information) 9. Hida, Kazuki and Ritsuo Yoshioka, Optimal Axial Enrichment Distribution of the Boiling Water Reactor Fuel Under the Haling Strategy, Nuclear Technology, 80, pp. 423-430, March 1988. 10. Hirano, Yasushi, Kazuki Hida, Koichi Sakurada, and Munenari Yamamoto, Optimization of Fuel Rod Enrichment Distribution to Minimize Rod Power Peaking throughout Life with BWR Fuel Assembly, Journal of Nuclear Science and Technology, 34, pp. 5-12, January 1997. 11. Kazimi, Mujid and Neil Todreas, Nuclear Systems 1: Thermal Hydraulic Fundamentals, Taylor and Francis, Bristol, PA, 1993 12. Lahey, R.T. and F.J. Moody, The Thermal Hydraulics of a Boiling Water Nuclear Reactor, American Nuclear Society, La Grange Park, Illinois, 1979. 88

PAGE 101

89 13. Lamarsh, John R., Introduction to Nuclear Engineering, Addison-Wesley Publishing Company, Inc., Melano Park, California, 1983. 14. Lewis, E.E. and W.F. Miller, Computational Methods of Neutron Transport, Nuclear Society, La Grange Park, Illinois, 1993. 15. Raharjo, R. and Mark Williams, Space-Dependent Resonance Self-Shielding, Nuclear Science and Engineering, 126, pp. 19-34, May 1997. 16. Rust, James H., Nuclear Power Plant Engineering, S.W. Holland Company, Atlanta, GA, 1979. 17. USNRC 10CFR Part 50, Section 50.62, Requirements for Reduction of Risk from Anticipated Transients Without Scram (ATWS) Events for Light-Water-Cooled Nuclear Power Plants, Washington D.C., January 1, 1996. 18. Weinberg, Alvin M. and Eugene P. Wigner, The Physical Theory of Neutron Chain Reactors, The University of Chicago Press, Chicago 1958.

PAGE 102

BIOGRAPHICAL SKETCH Michael Lorne Fensin was born in the city of Miami, Florida, on February 2, 1980. He served as president for the Bnai Brith Youth Organization city of Miami counsel and state of Florida region from May 1997-98. Michael is a member of the American Nuclear Society (ANS) and served as treasurer for the University of Floridas ANS chapter. He further served as president for the American Nuclear Societys honors society, for University of Florida. Michael was a recipient of the bright futures scholarship, dean of the college of engineering scholarship, and the national academy of nuclear training fellowship. During his undergraduate and masters degree work, Michael worked for the University of Florida as well as a variety of businesses in the nuclear industry. He served as a laboratory technician for the University of Floridas neutron activation analysis laboratory. Michael also interned for a summer at Southern Nuclear Companys Alvin W. Vogtle Electric Generating Plant in the area of reactor engineering and reactor operations. In the following summer Michael interned at Global Nuclear Fuels where he collaborated his thesis efforts with work completed at that facility. After completing his masters thesis requirements Michael will continue on for a PhD at the University of Florida in the area of nuclear space power and propulsion. His PhD thesis will be in collaboration with work at Los Alamos National Labs. 90