Nuclear Safety OF RBMK Reactors: Università Di Pisa Scuola Di Dottorato in Ingegneria "Leonardo Da Vinci"

UNIVERSITÀ DI PISA
Scuola di Dottorato in Ingegneria “Leonardo da Vinci”
Corso di Dottorato di Ricerca in

SICUREZZA NUCLEARE ED INDUSTRIALE
Tesi di Dottorato di Ricerca
NUCLEAR SAFETY
OF
RBMK REACTORS
Carlo Parisi
Anno 2008
UNIVERSITÀ DI PISA
Scuola di Dottorato in Ingegneria “Leonardo da Vinci”
Corso di Dottorato di Ricerca in

SICUREZZA NUCLEARE ED INDUSTRIALE
Tesi di Dottorato di Ricerca
NUCLEAR SAFETY OF
RBMK REACTORS
Autore:
CARLO PARISI Firma___________________
Relatori:
Prof. Francesco D’Auria (UNIPI) Firma_____________
Dr. Gianni Petrangeli Firma_____________
Prof. Eugenius Uspuras (LEI) Firma_____________
Dr. Sergei L. Soloviev (NIKIET) Firma_____________
Anno 2008
ACKNOWLEDGMENTS
This PhD thesis was developed at the San Piero a Grado Nuclear Research Group
of the University of Pisa. So, my first thank is due to the leader of this Group and
thesis supervisor, prof. F. D’Auria. He gave me the opportunity to perform this job
in a fruitful and stimulating environment, and he always supported and encouraged
the research activities. I appreciated also his example in looking at the RBMK
technology without any prejudice.
I would like to give also special thanks to the other PhD thesis supervisors, Dr. G.
Petrangeli, prof. E. Uspuras and Dr. S. L. Soloviev, for their collaboration,
suggestions and for the trust they showed in my activities.
Dr. V. Malofeev from Kurchatov Institute and Prof. K. N. Ivanov from the
Pennsylvania State Univeristy helped me a lot with their advises on the neutron
transport calculations. I am very grateful to them.
A special mention is due to the NIKIET staff, led by Dr. B. Gabaraev, for their kind
hospitality and assistance provided during my stays in Russia.
I would like to thank also Dr. B. Ivanov, from Westinghouse Electric and Dr. E.
Vanagas, from the Lithuanian Safety Authority for the useful conversations I had
with them on the RBMK HELIOS models.
This work would not be possible if I did not have the opportunity to work in the
serious and at the same time friendly environment of the San Piero a Grado
Nuclear Research Group. For this reason I am grateful to all my present
colleagues and to the whole original team of persons that from 2004 to 2006
contributed with dedicated work to the unquestionable success of the TACIS
Project R2.03/97, from which this PhD work originated.
My final thanks and dedication of this work should go to my parents, and especially
to my father. During all these years he staunchly supported and encouraged me to
perform and finalize the PhD activities, reminding, when difficulties seemed
discouraging, that “..non est ad astra mollis e terra via”.
i
ii
SOMMARIO
La presente Tesi di Dottorato esamina il livello di sicurezza dei reattori di potenza
raffreddati ad acqua bollente e moderati a grafite (reattori RBMK), mediante l’uso di
codici di calcolo best-estimate termoidraulici e neutronici accoppiati. La
disponibilità di tali sofisticati strumenti ha reso possibili, infatti, analisi dettagliate e
realistiche anche per questo tipo di impianti nucleari, notoriamente complessi e
comunemente denominati “reattori di tipo Chernobyl” (in riferimento al gravissimo
incidente occorso ad uno di questi impianti nel 1986). Parte delle attività della
presente Tesi di Dottorato, si sono svolte nell’ambito del Progetto TACIS R2.03/97
“Software development for the RBMK and WWER reactors”, coordinato dal Gruppo
di Ricerca Nucleare San Piero a Grado dell’Università di Pisa in collaborazione con
i progettisti e gli esercenti russi di tali impianti (NIKIET, Kurchatov Institute,
RosEnergoAtom, oggi EnergoAtom Concern OJSC).
Le attività di ricerca hanno contemplato lo sviluppo di una complessa

nodalizzazione termoidraulica dell’unità 3 dell’impianto nucleare di Smolensk, la
sua validazione ed il suo successivo accoppiamento con un modello di cinetica
neutronica tridimensionale. Il codice utilizzato è il RELAP5-3D. I calcoli di
neutronica hanno richiesto, preliminarmente, lo sviluppo di librerie di sezioni d’urto
macroscopiche. Tale attività, è stata svolta in collaborazione con la Pennsylvania
State University, ed ha comportato l’utilizzo del codice di trasporto deterministico
HELIOS.
Dopo la validazione dei modelli sviluppatii, si sono eseguite analisi di transitori a

piena potenza, focalizzandosi su quelli particolarmente critici per la sicurezza (per
esempio, la rottura di un collettore dei canali di potenza o del sistema di
raffreddamento delle barre di controllo). Una speciale enfasi è stata data, anche
mediante l’uso del codice Montecarlo MCNP5, allo studio del transitorio
contemplante il bloccaggio di un canale di potenza.
L’ultima parte delle attività di dottorato si sono concentrate sull’analisi di un

transitorio a bassa potenza, ricostruendo uno scenario incidentale estremo come
quello occorso a Chernobyl. Le sezioni d’urto di cella per lo Xeno sono state
calcolate con il codice al trasporto deterministico DRAGON.
In conclusione, le analisi avanzate effettuate nell’ambito di questa Tesi di

Dottorato, hanno confermato il migliorato grado di sicurezza di tali impianti,
ottenuto grazie alle importanti modifiche effettuate a seguito dell’incidente di
Chernobyl.
iii
ABSTRACT
This PhD thesis is evaluating the safety level of the graphite-moderated boiling
water cooled nuclear power reactors (RBMK reactors) by the use of best estimate
three dimensional neutron kinetics coupled thermal-hydraulics codes. The
availability of such sophisticated tools has allowed detailed and realistic analyses
of these kind of reactors, also known as “Chernobyl-type” reactors. Chernobyl is
the name of a RBMK reactor where, in 1986, a severe accident occurred, leading
to the destruction of the plant and to a major release of radioactivity into the
environment. Parts of the activities of this PhD thesis were developed in the
framework of the European Union funded TACIS Project R2.03/97 “Software
development for the RBMK and WWER reactors”. This project was awarded to the
“Gruppo di Ricerca Nucleare San Piero a Grado” of the University of Pisa and
managed by it in collaboration with the RBMK designers (NIKIET, Kurchatov
Insititute) and the licensee (RosEnergoAtom, now EnergoAtom Concern OJSC).
The research activities dealt with the development and the validation of a
sophisticated thermal-hydraulic nodalization of the Smolensk-3 Nuclear Power
Plant. This thermal-hydraulic model was then coupled with a three dimensional
neutron kinetics model of the core. The code used was RELAP5-3D system code.
Suitable RBMK cross sections libraries were developed in collaboration with the
Pennsylvania State University, using the deterministic lattice physics code
HELIOS.
After the validation of the developed models, the most relevant transients for the
plant safety at full power were calculated, e.g. the group distribution header
rupture, the break of the control and protection system cooling circuit. A special
emphasis was put in the simulation of the single fuel channel transient, using also
the Monte Carlo code MCNP5.
The last part of the PhD activities concerned the analysis of a low power transient.
In particular, the Chernobyl extreme scenario was reconstructed. Xenon fuel cell
cross sections were calculated using the deterministic transport code DRAGON.
Finally, all the analyses performed in the framework of this PhD confirmed the
upgraded level of nuclear safety of the RBMK reactors, obtained also as a
consequence of the relevant hardware modifications implemented in the aftermath
of the Chernobyl accident.
iv
CONTENTS
ACKNOWLEDGMENTS............................................................................... I
SOMMARIO................................................................................................ III
ABSTRACT ................................................................................................ IV
CONTENTS. ................................................................................................ V
NOMENCLATURE ..................................................................................... IX
LIST OF SYMBOLS ................................................................................. XIII
LIST OF FIGURES ....................................................................................XV
LIST OF TABLES....................................................................................XXII
1. INTRODUCTION ..................................................................................... 1
1.1. State of the Art on the RBMK Safety Analyses............................. 1
1.2. Scope 2
1.3. Objectives......................................................................................... 3
1.4. Structure........................................................................................... 3
2. THE DESCRIPTION OF THE RBMK ...................................................... 5

2.1. The Primary System ........................................................................ 8
2.1.1. The pressure boundary ............................................................ 8
2.1.2. The core ................................................................................... 15
2.1.2.1. The fuel .......................................................................................... 15
2.1.2.2. The Absorbers and the Power Control System......................... 18
2.1.2.3. The scram signals ........................................................................ 22
2.1.2.4. The fuel cell neutronic characteristics ....................................... 23
2.1.2.5. Reactivity Control ......................................................................... 26
2.1.2.5.1. Fuel Cell Overmoderation ........................................................ 26
2.1.2.5.2. The Operating Reactivity Margin ............................................. 27
2.1.2.5.3. The Reactivity Coefficients ...................................................... 28
2.1.3. The Balance of Plant ............................................................... 29
2.2. The Confinement System.............................................................. 32
2.2.1. The Reactor Cavity.................................................................. 34
2.2.2. The Accident Localization System ........................................ 35
2.2.3. The Reactor Building and the Turbine Hall........................... 37
v
2.2.4. The fuel loading machine hall ................................................ 38
2.3. The Engineered Safety Features and the Emergency
System……….. ...................................................................................... 38
2.4. The reference plant and the reference conditions: Smolensk-3,
16 October 1996 .................................................................................... 42
2.4.1. Reference fuel load map......................................................... 42
3. THE BACKGROUND FOR RBMK ACCIDENT ANALYSIS ............... ..47

3.1. The technological status for RBMK safety.................................. 47
3.1.1. The background ...................................................................... 47
3.1.2. The relevant aspects............................................................... 48
3.2. The safety needs............................................................................ 53
3.2.1. The RBMK acceptance criteria............................................... 55
3.2.1.1. Fuel clad integrity ......................................................................... 56
3.2.1.2. Fuel channel integrity................................................................... 56
3.2.1.3. MCC integrity ................................................................................ 58
3.2.1.4. RC integrity ................................................................................... 58
3.2.1.5. ALS integrity ................................................................................. 59
3.2.1.6. The permissible radiation doses................................................. 61
3.2.2. Recent requirements by RosTechnadzor pertaining to BDBA
analysis including SA..................................................... 62
3.3. Identification and characterization of selected RBMK accident
scenarios and phenomena .................................................................. 65
3.3.1. Thermal-hydraulics of PS ....................................................... 65
3.3.2. Three-dimensional neutron kinetics...................................... 69
4. THE TOOLS, THE METHODOLOGIES AND THE QUALIFICATION OF

3D NK COUPLED TH CODES ANALYSES ............................................. 71
4.1. Neutron Transport codes.............................................................. 71
4.1.1. The Monte Carlo code MCNP5 ............................................... 71
4.1.1.1. General features ........................................................................... 71
4.1.1.2. Monte Carlo Method vs. Deterministic Method ......................... 72
4.1.1.3. The Monte Carlo Method.............................................................. 72
4.1.1.4. Nuclear Data and Reactions ........................................................ 73
4.1.1.5. Tallies and Output ........................................................................ 74
4.1.1.6. Estimation of Monte Carlo Errors ............................................... 74
4.1.2. The deterministic transport codes ........................................ 75
4.1.2.1. The lattice physics code DRAGON ............................................. 76
4.1.2.2. The lattice physics code HELIOS................................................ 77
4.2. The RELAP5-3D system code....................................................... 80
4.2.1. Introduction ............................................................................. 80
4.2.2. Relationship of RELAP5-3D to prior versions ...................... 81
vi
4.2.3. The Thermal-hydraulic part: the RELAP5-3D code .............. 82
4.2.4. The three dimensional neutron kinetics routine: the NESTLE
code ................................................................................. 83
4.2.4.1. The Steady-State and the Transient Problem............................ 83
4.2.4.2. The Cross Section and the feedbacks........................................ 86
4.2.4.3. The GEN neutron cross section model ...................................... 88
4.3. The Cross Section Processing methodology ............................. 91
4.3.1. Introduction ............................................................................. 91
4.3.2. The State of the Art methodologies....................................... 91
4.3.3. The Methodology for the RBMK calculations ....................... 96
4.4. The coupling methodology........................................................... 97
4.5. The procedure for code application including the nodalization
…………………...................................................................................... .98
4.5.1. Introduction ............................................................................. 98
4.5.2. The HELIOS code model......................................................... 99
4.5.2.1. Core geometry .............................................................................. 99
4.5.2.2. The Cross-section library .......................................................... 106
4.5.3. The RELAP5-3D TH nodalization ......................................... 111
4.5.4. The NESTLE 3D NK nodalization and the coupling ........... 123
4.5.4.1. 3D NK-TH Coupling Scheme ..................................................... 124
4.5.4.2. Some considerations on the axial meshing scheme .............. 129
4.5.5. The MCNP5 code model ....................................................... 130
4.5.5.1. Reference Geometry Material.................................................... 130
4.5.5.2. Reference Material data ............................................................. 131
4.5.5.3. MCNP5 Reference Model Geometry ......................................... 132
4.5.5.4. MCNP5 Cross Section Libraries................................................ 133
4.5.5.5. Boundary Conditions ................................................................. 133
4.5.6. The DRAGON code model .................................................... 133
4.5.6.1. Cross Section Libraries ............................................................. 134
4.5.6.2. Boundary Conditions ................................................................. 134
4.5.6.3. Calculation modules................................................................... 134
4.6. The Qualification.......................................................................... 135
4.6.1. Codes qualification ............................................................... 135
4.6.1.1. The RELAP5-3D validation for RBMK analyses ...................... 135
4.6.1.2. The HELIOS code validation for RBMK analyses.................... 136
4.6.1.3. The DRAGON code validation for RBMK analyses ................. 136
4.6.1.4. The MCNP validation for RBMK analyses ................................ 136
4.6.2. Model and Nodalization qualification .................................. 137
4.6.2.1. RELAP5-3D TH nodalization qualification................................ 137
4.6.2.2. RELAP5-3D 3D NK nodalization qualification ......................... 142
4.6.2.3. HELIOS model qualification ...................................................... 146
4.6.2.4. The MCNP5 model qualification................................................ 147
4.6.2.4.1. Sensitivity analyses................................................................ 148
4.6.2.4.2. Geometry Variation ................................................................ 148
4.6.2.4.3. Cross Section Libraries Variation........................................... 148
4.6.2.4.4. Boundary Conditions.............................................................. 148
vii
4.6.2.4.5. Calculation Parameters Variation .......................................... 148
4.6.2.4.6. Results of Sensitivity Analyses .............................................. 150
4.6.2.5. The DRAGON model qualification ............................................ 151
5. THE APPLICATION TO REALISTIC TRANSIENTS

ANALYSES…………. ...........................................................................…153
5.1. Hot Full Power Analyses............................................................. 153
5.1.1. Reactivity Initiated Accidents .............................................. 153
5.1.1.1. CR Withdrawal ............................................................................ 154
5.1.1.2. CR Group Withdrawal................................................................. 161
5.1.2. Decrease of coolant flow events.......................................... 169
5.1.2.1. GDH blockage ............................................................................. 169
5.1.2.2. Fuel Channel Blockage: the System Code Analysis............... 175
5.1.2.3. The Fuel Channel Blockage: the FC criticality calculations .. 179
5.1.2.3.1. Introduction ............................................................................ 179
5.1.2.3.2. Single Fuel Channel - 2.0% Fuel ........................................... 179
5.1.2.3.3. Single FC – 2.4% Fuel ........................................................... 182
5.1.2.3.4. Lattice Cell Analysis ............................................................... 185
5.1.2.3.5. Conclusions............................................................................ 196
5.1.3. LOCA events.......................................................................... 169
5.1.3.1. CPS-LOCA ................................................................................... 197
5.1.3.2. GDH LOCA (rupture after the GDH check valve)..................... 200
5.2. Conclusion ................................................................................... 208
6. THE EXTREME CASE ........................................................................ 209

6.1. Introduction.................................................................................. 209
6.2. The Chernobyl event ................................................................... 209
6.3. Literature review .......................................................................... 214
6.4. Low Power analyses – the model upgrade ............................... 214
6.4.1. Xenon modelling ................................................................... 215
6.4.2. Power Control modelling...................................................... 215
6.5. The Power-reduction transient................................................... 216
6.6. The “Chernobyl-like” event ........................................................ 220
6.6.1. Reference Transient.............................................................. 221
6.6.1.1. Positive Reactivity Perturbation ............................................... 226
6.6.1.2. No Scram actuation .................................................................... 227
6.6.2. Conclusions........................................................................... 227
7. CONCLUSIONS .................................................................................. 228
viii
NOMENCLATURE
0D Zero-dimensional
1D One-dimensional
2D Two-dimensional
3D Three-dimensional
AA Additional Absorber
Accumulator (part of the ECCS, also identified as hydro-
ACC
accumulator)
ACRR Annular Core Research Reactor (Sandia National Lab)
ADF Assembly Discontinuity Factor
AECL Atomic Energy of Canada Limited
AHTLM Adaptive High-order Table Look-up Method
ALS Accident Localization System
ALT Accident Localization Tower
ANM Analytical Nodal Method
ANS American Nuclear Society
ANS Advanced Neutron Source (Oak Ridge National Lab)
AOO Anticipated Operational Occurrences
ATR Advanced Test Reactor
ATWS Anticipated transient without scram
AZ-1, AZ-3 to RBMK Emergency Protection Signals (Scram or Power
AZ-6 reduction)
AZM (Signal of) Emergency Protection by Power
BAZ RBMK scram key (Fast functioned emergency protection)
BDBA Beyond Design Basis Accident
BE Best Estimate
BIC Boundary and Initial Conditions
BPLU Border Profile Lower Upper (Matrix Solver)
BRU-K, -D, -TK Safety Valves on Steam Lines
BWR Boiling Water Reactor
CHF Critical Heat Flux
ChNPP Chernobyl Nuclear Power Plant
CMFD Coarse Mesh Finite Difference
CAMP Code Assessment and Maintenance Program
CANDU Canadian deuterium uranium (reactor)
CCPM; Current-Coupling Collision Probability Method
CPM Collision Probability Method
CPS Control and Protection System
CPU Central Processing Unit
CR Control Rod
CRCC Control Rod Cooling Circuit
CTV Control Throttling Valve
DBA Design basis accident
DC Down-Comer
DF Discontinuity Factor
DOE Department of Energy (U.S.A.)
ix
DREG Diagnostic Parameter Recording Program
DS Drum Separator, see also SD or Steam Drum
EC The European Commission
ECCS Emergency Core Cooling System
ENDL Evaluated Nuclear Data Library
ENDF Evaluated Nuclear Data File -
ESF Engineered Safety Features
FASS Fast Acting Scram System
FDM Finite Difference Method
FLOP Floating Operation
GDH Group Distribution Header
The San Piero a Grado Nuclear Research Group of the
GRNSPG
University of Pisa, Italy
DBA Design Basis Accident
Department of Mechanical, Nuclear and Production
DIMNP
Engineering of the University of Pisa, Italy
DS Drum separator
ECCS Emergency Core Cooling System
EPS Emergency Protection Signal
EREC Electrogorsk Research and Engineering Center
ET Equivalent Channel
FASS Fast Acting Scram System
FC Fuel Channel
FDM Finite Difference Method
FLOP Floating Point Operation
FP Fission Product
FSP Fixed Source Problem
Gosatomnadzor, the former Russian Nuclear Safety Authority,
GAN
now Rostechnadzor
The Russian Research Center “Kurchatov Institute” or see
KI
RRC KI
HFBR High Flux Beam Reactor
HFIR High Flux Isotope Reactor
IAEA International Atomic Energy Agency
ICAP International Code Assessment and Applications Programme
ICV Isolated control valve
IGR Impulse Graphite Reactor
INL Idaho National Laboratory
LAC Local Automatic Control
LANL Los Alamos National Laboratory
LAP Local Automatic Protection
LAR Local Power Control
LEI Lithuanian Energy Institute
LOCA Loss of coolant accident
LOFT Loss of Fluid Test Reactor
LNPP Leningrad Nuclear Power Plant
LTC Leak-Tight Compartment
LWL Lower Water Line
LWR Light Water Reactor
x
MCC Main circulation circuit
MCP Main circulation pump
MCR Manual Control Rod
MFCC Multipass Forced Circulation Circuit
MPTR Multiple Pressure Tube Rupture
MSV Main Safety Valve
NEM Nodal Expansion Method
“N. A. Dollenzhal” Institute for Power Engineering (RBMK
NIKIET
designer)
NK Neutron Kinetics
NPR New Production Reactor
NRC Nuclear Regulatory Commission (U.S.A.)
NRU National Research Universal Reactor (Chalk River Lab)
NPP Nuclear Power Plant
ORM Operating Reactivity Margin
ORNL Oak Ridge National Laboratory
PDDMS Power Density Distribution Monitoring System
PDMS-A Power Density Sensors of the Axial Monitoring
PH Pressure Header
PHWR Pressurized heavy water reactor
PSA Probabilistic Safety Analysis
PSP Pressure Suppression Pool
PSU The Pennsylvania State University
PT Pressure Tube
PWR Pressurized Water Reactor
Reactor Bolsoi Mochnosti Kipyashiy (Large Power Boiling
RBMK
Reactor)
RC Reactor Cavity
RCPS see MCR
RCSRV Reactor Cavity Steam Relief Valve
RCVS Reactor Cavity Venting System
RDFMG The Reactor Dynamics and Fuel Management Group
RIA Reactivity Initiated Accident
RRCC Radial Reflector Cooling Channels
RRC KI The Russian Research Center “Kurchatov Institute”
SA Severe Accident
SCS Sprinkler Cooling System
SCV Stop-control valve
SD Steam Drum
SDC Steam Corridor Distribution
SHR Shortened Control Rod
SL Steam Line
SOR Successive Over Relaxation
SPTR Single Pressure Tube Rupture
SR Safety Control Rod (components of FASS)
SRV Steam Relief Valve
SS Steady State
SV Stop Valve
SWL Steam Water Line
xi
TACIS Technical Aid to the Commonwealth of the Independent States
TH Thermal-Hydraulic
TPEN Triangle-based Polynomial Expansion Nodal Method
URC Under Reactor Compartments
UNIPI The University of Pisa
VATESI Lithuanian Nuclear Power Safety Inspectorate
VVER see WWER
XSEC Cross-section (for 3D Neutron Kinetics calculations)
WWER Water Cooled Water Moderated Energy Reactor
xii
LIST OF SYMBOLS
Greek Symbols
aφ Void Reactivity Coefficient (pcm/% void)
aw Power Reactivity Coefficient (pcm/MWth)
at Doppler Reactivity Coefficient (pcm/°C)
ac Graphite Temperature Reactivity Coefficient (pcm/°C)
acool Coolant Temperature Reactivity Coefficient (pcm/°C)
β Delayed Neutron Fraction
f Neutron flux (n/cm2*s)
ε Fast Fission Factor
εR Error in the neutron flux calculation
γ Gamma Particle
h Fission Neutron per Neutron Absorption into the Fuel
n Number of Fission Neutrons
µ Cosine of the direction angle for neutrons
ρ Reactivity
ρw Material Density (Kg/m3)
ξ Average Logarithmic Energy Loss
si Stress Intensity (MPa)
sXe Microscopic Xenon Absorption Cross Section (*106 barn)
Σa Macroscopic Absorption Cross Section (cm-1)
Σf Macroscopic Fission Cross Section (cm-1)
Σsg‘g Macroscopic Group-to-Group Scattering Cross Section (cm-1)
n Σf Macroscopic Neutron Generation Cross Section (neutron* cm-1)
τth Fermi age (cm2)
zi Strain Rate Intensity
Roman Symbols
$ Reactivity (1$ = 1 β )
B Neutron Flux Distribution Buckling (cm-1)
(12)
C(6) Graphite (Nuclear Grade)
D Diffusion Coefficient (cm)
f Thermal Utilization Factor
eV Electron Volt
k Multiplication factor (eigenvalue)
j Neutron current
ji Rupture Strain Power (W/Kg)
L Diffusion Length (cm)
Lgil Average i-directed leakage for group g in node l
M Migration Length (cm)
n Neutron Particle
xiii
p Resonance Escape Probability Factor
Tw Pressure Tube Temperature (K)
u Neutron Lethargy
xiv
LIST OF FIGURES
Fig. 1 – Overall view of the RBMK NPP system....................................................... 5
Fig. 2 – Smolensk-3 NPP view ................................................................................. 6
Fig. 3 – RBMK Main Circulation Circuit (MCC) flow diagram ................................... 8
Fig. 4 – Key elevations in the MCC of the RBMK system (Smolensk 3)................. 9
Fig. 5 – The MCP for RBMK NPP. ......................................................................... 10
Fig. 6 – RBMK MCP (CVN-8) performance data (left: high temperature conditions,
265 ºC; right: low temperature conditions, 90 °C)......................................... 11
Fig. 7 – Geometrical configuration of RBMK GDH ................................................. 12
Fig. 8 – Geometrical configuration of RBMK SD: a) overall view; b) cross section13
Fig. 9 – Steam Drum – component specifications.................................................. 14
Fig. 10 – Layout of the two DS and related connection.......................................... 15
Fig. 11 – Geometrical configuration of RBMK core: a) detail of active region; b)
overall view .................................................................................................... 16
Fig. 12 – Geometrical configuration of a) RBMK fuel bundle, b) fuel rod............... 17
Fig. 13 – Geometrical and material configuration of RBMK FC ............................. 17
Fig. 14 – Geometrical and material configuration of regular and cluster type AAC of
RBMK core..................................................................................................... 18
Fig. 15 – Longitudinal view of RBMK CRs (Short, Safety and Manual types)........ 20
Fig. 16 – Cross sectional view of RBMK regular CR.............................................. 21
Fig. 17 – Flow diagram of the CPS cooling system................................................ 21
Fig. 18 – Fuel Cell basic model .............................................................................. 24
Fig. 19 – Thermal neutron flux distribution ............................................................. 25
Fig. 20 – Fast neutron flux...................................................................................... 25
Fig. 21 – Neutron flux distribution........................................................................... 26
Fig. 22 – kinf versus moderation ratio of Moderator area on Fuel area................... 27
Fig. 23 – Flow diagram of the BOP system of the RBMK core .............................. 30
Fig. 24 – Flow diagram of SL from the SD region till the MSVs ............................. 31
Fig. 25 – Flow diagram of RBMK SL from the MSIV till the turbines...................... 31
Fig. 26 – Sketch of the confinement system of the RBMK NPP with main sub-
systems (numbers refer to components identified in Fig. 3.1, where
applicable)...................................................................................................... 32
Fig. 27 – Cross sectional view of the RBMK confinement system ......................... 33
Fig. 28 – Longitudinal view of the RC of a RBMK NPP.......................................... 34
Fig. 29 – Gas removal system of the RC of a RBMK NPP..................................... 35
Fig. 30 – Sketch of the Ignalina ALS ...................................................................... 37
Fig. 31 – Main (or Short-term) ECCS subsystem, RBMK damaged core side
delivery system .............................................................................................. 39
Fig. 32 – Long-term ECCS subsystem, RBMK damaged core side delivery system
....................................................................................................................... 40
Fig. 33 – Long-term ECCS subsystem, RBMK un-damaged core side delivery
system............................................................................................................ 41
Fig. 34 – The reference map of fuel load of the Unit-3 of Smolensk NPP (16.10.96)
....................................................................................................................... 42
Fig. 35 – Control Rods Cartogram.......................................................................... 43
xv
Fig. 36 – Core Channels Flow Map (Kg/s) ............................................................. 46
Fig. 37 – Best estimate and conservative safety analysis: definitions of acceptance
criteria and safety margins, [12]..................................................................... 55
Fig. 38 – Pressure tube rupture temperature versus internal pressure.................. 57
Fig. 39 – Specific rupture strain power as a function of tube temperature............. 57
Fig. 40 – A typical neutron history simulation by MNCP5 (from [24])..................... 73
Fig. 41 – HELIOS package..................................................................................... 77
Fig. 42 – Flow scheme of library generation. ......................................................... 79
Fig. 43 – Overview of the nested iterative solution strategy................................... 86
Fig. 44 – Cross Section dependence ..................................................................... 92
Fig. 45 – Cross-section interdependence between fuel temperature and moderator
temperature.................................................................................................... 93
Fig. 46 – Cross-section calculation points for the polynomial fitting procedure ..... 94
Fig. 47 – Areas of inaccurate cross-section calculation using the polynomial fitting
procedure ....................................................................................................... 94
Fig. 48 – The PSU transient cross-section representation..................................... 95
Fig. 49 – RBMK Libraries Cross Section – Thermal absorption trend for
composition 29 (see section 4.5.2) ................................................................ 96
Fig. 50 – RBMK Libraries Cross Section – Thermal fission trend for composition 29
(see section 4.5.2) ......................................................................................... 97
Fig. 51 – Codes and Models developed for 3D NK TH RBMK analyses ............... 98
Fig. 52 – HELIOS modeling of RBMK FC cell ...................................................... 100
Fig. 53 – HELIOS modeling of RBMK MCR cell................................................... 100
Fig. 54 – Core neutron kinetic mesh..................................................................... 101
Fig. 55 – Two-dimensional channel type map (core left part) .............................. 104
Fig. 56 – Two-dimensional channel type map (core right part) ............................ 105
Fig. 57 – Interpolation Scheme for Cross Sections parameters.......................... 107
Fig. 58 – Macroscopic cross-section table’s structure.......................................... 110
Fig. 59 – Sketch of the reference RELAP5-3D TH nodalization of Smolensk-3 (only
left MCC showed) ........................................................................................ 112
Fig. 60 – Sketch of the simplified RELAP5-3D TH nodalization (Left and Right
MCC)............................................................................................................ 113
Fig. 61 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of FW and
SL systems................................................................................................... 114
Fig. 62 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of down-
comers up to the water lines. ....................................................................... 115
Fig. 63 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of core, FC
exit lines and CPS cooling system............................................................... 115
Fig. 64 – Relap5 UNIPI model of Smolensk 3 NPP: nodalization sketch of ECCS.
..................................................................................................................... 117
Fig. 65 – RELAP5-3D Core neutron kinetic mesh................................................ 124
Fig. 66 – TH / 3D NK & Lattice Codes meshes coupling...................................... 125
Fig. 67 – Radial Arrangement of TH Channels for General Purpose Nodalization
..................................................................................................................... 126
Fig. 68 – Correlation map between TH channels, TH Zones and NK Nodes for
General Purpose Nodalization ..................................................................... 127
Fig. 69 – CR Numbers & Types – RELAP5-3D Numeration ................................ 128
Fig. 70 – Relative Axial Power: SS sensitivity, FD method calculations .............. 129
Fig. 71 – Relative Axial Power: SS sensitivity, NEM calculations ........................ 130
xvi
Fig. 72 – Fuel Channel Cross-View...................................................................... 130
Fig. 73 – Fuel Cell in XY plane (at Z=10) ............................................................. 132
Fig. 74 – Fuel Cell in YZ and XZ plane (at X=0 and Y=0) .................................... 132
Fig. 75 – DRAGON RBMK FC model................................................................... 133
Fig. 76 – DRAGON code: calculation scheme used for RBMK calculations........ 134
Fig. 77 – RELAP5-3D Smolensk 3 NPP: nodalization qualification at steady state
level, comparison between adopted and reference (design) volume vs height
curve of MCC. .............................................................................................. 138
level, comparison between steady state and reference (nominal conditions)
pressure distribution along the MCC. .......................................................... 140
void fraction at SD inlet, right part, channels 1-24. ...................................... 140
graphite axial temperature at outer surface. ................................................ 141
FC void fraction (2.01 MW) .......................................................................... 141
level, radial temperature distribution of graphite at various elevations along
the FC stack. ................................................................................................ 142
Fig. 83 –RELAP5-3D Smolensk 3 Axial Power ................................................... 142
Fig. 84 – Smolensk 3 nodalization qualification for 3D NK-TH Relap5/Nestle code:
axial power distribution in selected FC (coordinate 40-47).......................... 143
Fig. 85 – RELAP5-3D Radially Averaged Power ................................................. 143
Fig. 86 – RELAP5-3D Relative Radial Power ...................................................... 144
Fig. 87 – 3D Thermal Neutron Flux Distribution (n/(cm2*sec)) ............................. 144
Fig. 88 – kinf trend versus cycle number ............................................................... 147
Fig. 89 – Shannon Entropy trend versus cycle number ....................................... 147
Fig. 90 – 3 x 3 Lattice .......................................................................................... 149
Fig. 91 – 5 x 5 Lattice ........................................................................................... 149
Fig. 92 – CR Cartogram for RELAP5-3D RIA calculations................................... 154
Fig. 93 – Radial Arrangement of TH Channels for CR Withdrawal Nodalization . 156
Fig. 94 – Correlation map between TH channels, TH Zones and NK Nodes for CR
Withdrawal Nodalization .............................................................................. 157
Fig. 95 – Total Reactor Power.............................................................................. 158
Fig. 96 – 1 FC equivalent Power .......................................................................... 158
Fig. 97 – 1FC Equivalent Coolant Mass Flow – 214, 259, 265 channel .............. 159
Fig. 98 – Void Fraction at the exit of FC 214, 259, 265........................................ 159
Fig. 99 – Clad Temperature at Hot Spot............................................................... 160
Fig. 100 – Fuel CL temperature............................................................................ 160
Fig. 101 – Graphite Temperatures ....................................................................... 161
Fig. 102 – Radial Arrangement of TH Channels for CR Bank Withdrawal
Nodalization ................................................................................................. 163
Fig. 103 – Correlation map between TH channels, TH Zones and NK Nodes for CR
Bank Withdrawal Nodalization ..................................................................... 164
Fig. 104 – Reactor Power..................................................................................... 165
xvii
Fig. 105 – 1 FC Equivalent Power........................................................................ 165
Fig. 106 – Equivalent 1FC mass flow rate............................................................ 166
Fig. 107 – Top FC void fraction ............................................................................ 166
Fig. 108 – Hot Spot Clad Temperatures............................................................... 167
Fig. 109 – Hot Spot Fuel CL temperature ............................................................ 167
Fig. 110 – Hot Spot Graphite Outer Temperature ................................................ 168
Fig. 111 – 1 FC - Equivalent axial Power ............................................................. 168
Fig. 113 – 1FC Equivalent Mass Flow Rate (channels 265, 259, 214) ................ 170
Fig. 114 – 1FC Equivalent Mass Flow Rate (channels 244, 175, 669) ................ 171
Fig. 115 – Total Mass Flow in not-affected GDHs (Right Half) ............................ 171
Fig. 116 – Mass Flowrate in the affected GDH .................................................... 172
Fig. 117 – Void Fraction at top of affected GDH FCs........................................... 172
Fig. 118 – Hot Spot Clad Temperatures - FC 214, 259, 265, 175, 244 ............... 173
Fig. 119 – Hot Spot Fuel CL Temperatures - FC 214, 259, 265, 175, 244 .......... 173
Fig. 120 – Pressure Tubes temperatures at Hot Spot – FC 214, 259, 265, 175.. 174
Fig. 121 – Graphite Temperatures ....................................................................... 174
Fig. 122 – Relap5-3D/Nestle coupled 3D NK TH FC blckage analysis of Smolensk
3 RBMK NPP: flow-rate in the affected FC.................................................. 175
3 RBMK NPP: core power ........................................................................... 176
Fig. 124 – Mass Flowrate in the other channels of the affected GDH.................. 176
Fig. 125 – Void Fraction at the Blocked Channel Exit .......................................... 177
Fig. 126 – Affected FC power and power (per unit FC) in neighbouring FC. ....... 177
Fig. 127 – Rod surface temperature at different elevations in the affected FC.... 178
Fig. 128 – PT temperature at different elevations in the affected FC................... 178
Fig. 129 – 2.0% fuel cell criticality during a FCB event considering occurrence of all
phenomena, of channel voiding, of Doppler effect ...................................... 180
Fig. 131 – 2.0% fuel lattice criticality during a FCB event on central channel
considering occurrence of all phenomena, of channel voiding, of Doppler
effect ............................................................................................................ 186
effect ............................................................................................................ 190
Fig. 133 – 2.4% fuel lattice w/ central 2.0% FC: criticality during a FCB event on
central channel considering occurrence of all phenomena, of channel voiding,
of Doppler effect........................................................................................... 193
Fig. 135 – Power in FC of 11th GDH of right MCC part ........................................ 198
Fig. 136 – Equivalent Mass Flowrate in FC of 11th GDH of right MCC part ......... 198
Fig. 137 – Hot Spot Clad Temperature in FC 214, 259, 265, 669........................ 199
Fig. 138 – Void fraction in CPS Cooling Channel................................................. 199
Fig. 140 – Mass Flow rate in affected Channels 214, 259, 265 ........................... 201
Fig. 141 – Mass Flow rate in affected Channels 669, 175, 244 ........................... 202
Fig. 142 – Mass Flowrate at GDH rupture............................................................ 202
Fig. 143 – Imposed Mass flow rate in damaged MCC part .................................. 203
xviii
Fig. 144 – Void Fraction at Core Inlet................................................................... 203
Fig. 145 – Void Fraction at core outlet.................................................................. 204
Fig. 146 – Clad Temperature in FC 214 ............................................................... 204
Fig. 151 – PT Hot Spots temperatures ................................................................. 207
Fig. 152 – MCR configuration of Chernobyl-4 reactor.......................................... 210
Fig. 153 – Chernobyl axial distribution of the thermal neutron flux density at h.22:00
(curve 3) and h. 00:30 (curve 4) before the accident [83]............................ 213
Fig. 154 – CR bank 2 actuation logic.................................................................... 215
Fig. 155 – Simulation of Chernobyl event in the Smolensk NPP: Reconstructed-
imposed reactor power before the event start (t = 1240 s in the scale above)
..................................................................................................................... 216
Fig. 156 – Simulation of Chernobyl event in the Smolensk NPP: MCR operated by
power control logic - insertion ...................................................................... 217
Fig. 157 – Simulation of Chernobyl event in the Smolensk NPP: Xenon buildup 217
Fig. 158 – Simulation of Chernobyl event in the Smolensk NPP:MCPs mass flow,
RHS and LHS............................................................................................... 218
Fig. 159 – Simulation of Chernobyl event in the Smolensk NPP: SD liquid level 218
Fig. 160 – Simulation of Chernobyl event in the Smolensk NPP: FW mass flow per
SD, LHS and RHS ....................................................................................... 219
Fig. 161 – Simulation of Chernobyl event in the Smolensk NPP: Void fraction at the
core outlet, right and left side....................................................................... 219
Fig. 162 – Simulation of Chernobyl event in the Smolensk NPP: Reactor Power
before the Test ............................................................................................. 221
Fig. 163 – Simulation of Chernobyl event in the Smolensk NPP: CR before the test.
..................................................................................................................... 221
Fig. 164 - Simulation of Chernobyl event in the Smolensk NPP: MCP activation 222
Fig. 165 – Simulation of Chernobyl event in the Smolensk NPP: Left and Right side
Mass Flow (Kg/s) ......................................................................................... 222
Fig. 166 – Simulation of Chernobyl event in the Smolensk NPP: FW perturbation
..................................................................................................................... 223
Fig. 167 – Simulation of Chernobyl event in the Smolensk NPP: MCPs speed... 223
Fig. 168 – Simulation of Chernobyl event in the Smolensk NPP: MCP flow rates224
Fig. 169 – Simulation of Chernobyl event in the Smolensk NPP: Circuit Flow Rate
..................................................................................................................... 224
Fig. 170 – Simulation of Chernobyl event in the Smolensk NPP: Void Fraction at
core exit during the transient........................................................................ 225
Fig. 171 – Simulation of Chernobyl event in the Smolensk NPP: Power during the
transient ....................................................................................................... 225
Fig. 172 – Sensitivity with introduction of additional positive reactivity: Reactor
Power ........................................................................................................... 226
Fig. 173 – Sensitivity with introduction of additional positive reactivity: Core
Reactivity...................................................................................................... 226
Fig. 174 – Reactor Power without scram actuation.............................................. 227
xix
xx
LIST OF TABLES
Tab. 1 – Key parameters of the Smolensk-3 RBMK-1000 ....................................... 7
Tab. 2 – CPS Modes of Operation ......................................................................... 22
Tab. 3 – Comparison between diffusion parameters of Graphite, Light Water and
Heavy Water .................................................................................................. 23
Tab. 4 – Comparison of diffusion parameters for RBMK and LWRs...................... 23
Tab. 5 – Neutronic characteristics of a RBMK fuel cell .......................................... 24
Tab. 6 – Comparison between BWR and RBMK core parameters ........................ 26
Tab. 7 – Typical Reactivity coefficient for the Smolensk-3 NPP [8]. ...................... 29
Tab. 8 – Characteristics of FC and control rods ..................................................... 43
Tab. 9 – CR Standard Position............................................................................... 44
Tab. 10 – RBMK generations. ................................................................................ 60
Tab. 11 – ALS evolution in the various RBMK generations (see Tab. 10) and
acceptance criteria......................................................................................... 61
Tab. 12 – List of key-subjects and phenomena utilized for the assessment of
Korsar code and comparison between VVER and RBMK............................. 66
Tab. 13 – List of RBMK specific thermal-hydraulic phenomena for MCC. ............. 67
Tab. 14 – List of RBMK specific design features for the core neutron kinetics ...... 70
Tab. 15 – Range of the physical parameters to be considered for Fuel and Non-
fuel cells ......................................................................................................... 76
Tab. 16 – Definition of fuel channel types ............................................................ 101
Tab. 17 – Definition of non-fuel channel types ..................................................... 103
Tab. 18 – Composition numbers in axial layers for each channel type................ 108
Tab. 19 – Normalized axial exposure ................................................................... 109
Tab. 20 – Range of independent variables .......................................................... 109
Tab. 21 – Key to macroscopic cross-section tables ............................................ 110
Tab. 22 – Relap5 UNIPI model of Smolensk 3 NPP: correspondence between
nodes and NPP components (left side). ...................................................... 118
nodes and NPP components (right side). .................................................... 120
Tab. 24 – Nuclear densities in fresh fuel, 1024 atoms/cm3 ................................... 131
Tab. 25 – Geometry and material composition of fuel cell ................................... 131
Tab. 26 – RELAP5-3D Smolensk 3 NPP: nodalization: qualification at steady state
level, key MCC geometric parameters......................................................... 137
Tab. 27 – RELAP5-3D Smolensk 3 NPP: nodalization qualification at steady state
level, key MCC thermal-hydraulic parameters............................................. 138
Tab. 28 – RELAP5-3D Axial Peaking Factor........................................................ 143
Tab. 29 – RELAP5-3D Channels Radial Averaged Power................................... 145
Tab. 30 – RELAP5-3D Reactor core Neutron Kinetics Parameters..................... 145
Tab. 31 – RELAP5-3D adopted delayed neutron fraction data............................ 145
Tab. 32 – Comparison between HELIOS and MCNP codes of the kinf for a RBMK
lattice (Temp. 300 K, fresh fuel)................................................................... 146
Tab. 33 – Void Effects by HELIOS and MCNP Codes, absolute units................. 146
Tab. 34 – Results for the reference calculations .................................................. 148
Tab. 35 – Results from sensitivity calculation ...................................................... 150
xxi
Tab. 36 – Comparison between NIKIET and GRNSPG criticality calculations .... 150
Tab. 37 – Comparison between RRC-KI and GRNSPG results for voiding effect151
Tab. 38 - Comparison of the reference criticality calculations Monte Carlo versus
DRAGON code............................................................................................. 151
Tab. 39 – Comparison between RRC-KI and DRAGON results for voiding effect152
Tab. 40 – Cell parameter variation during a FCB event by RELAP5-3D code .... 179
Tab. 41 – 2.0% fuel cell criticality during a FCB event considering occurrence of all
Tab. 42 – 2.0% fuel cell reaction rates and fluxes for runs Xi to X3..................... 181
Tab. 43 – 2.0% fuel cell – Ratio of Tallies, % ....................................................... 181
Tab. 45 – 2.4% fuel cell reaction rates and fluxes for runs Xi to X3..................... 184
phenomena, of channel voiding, of Doppler effect, % ................................. 184
Tab. 47 – 2.0% fuel lattice criticality during a FCB event on central channel
effect ............................................................................................................ 186
Tab. 48 – 2.0% lattice fuel cells reaction rates and fluxes for runs Xi and X1 ..... 187
Tab. 49 – 2.0% lattice fuel cells reaction rates and fluxes for runs X2 and X3 .... 187
Tab. 50 – 2.0% lattice fuel cell – Ratio of Tallies, % ............................................ 188
effect ............................................................................................................ 189
Tab. 52 – 2.4% lattice fuel cells reaction rates and fluxes for runs Xi and X1 ..... 191
Tab. 53 – 2.4% lattice fuel cells reaction rates and fluxes for runs X2 and X3 .... 191
Tab. 54 – 2.4% lattice fuel cells – Ratio of Tallies, %........................................... 192
Tab. 55 – 2.4% fuel lattice w/ central 2.0% FC: criticality during a FCB event on
central channel considering occurrence of all phenomena, of channel voiding,
of Doppler effect........................................................................................... 193
Tab. 56 – 2.4% lattice fuel w/ 2.0% central FC: reaction rates and fluxes for runs Xi
and X1.......................................................................................................... 194
Tab. 57 – 2.4% lattice fuel w/ 2.0% central FC: reaction rates and fluxes for runs
X2 and X3 .................................................................................................... 195
Tab. 58 – 2.4% lattice fuel cell w/ 2.0% central FC: Ratio of Tallies, %............... 195
Tab. 59 – Chernobyl-4 accident, sequence of the events .................................... 211
Tab. 60 – Sequence of the events (actual versus reconstructed) ........................ 220
xxii
1. INTRODUCTION
1.1. State of the Art on the RBMK Safety Analyses

The Russian designed Water-Cooled Graphite-Moderated reactors RBMK were the
subjected to an extensive campaign of studies, researches and upgrades in the
aftermath of the Chernobyl disaster (26 April 1986).
In particular, immediate actions focused on plant modifications, and concerned

about a complete review of the neutronic design (e.g., increase of Uranium
enrichment, insertion of Additional Absorber (AA)), hardware modification (e.g.,
modification of Control Rods (CR) design, improvement of the core electronic
calculator) and plant operating procedures [1], [2].
After the Soviet Union collapse (26 December 1991) and during the whole ‘90s, the
European Commission (EC) and the United States Department of Energy (US-
DOE) financed a series of international projects (e.g., the TACIS Projects from the
European side), in collaboration with Russian Institutions, for a further improvement
of the safety of RBMK Nuclear Power Plant (NPP) and for the uniformities of safety
technology methods and understanding [3], [4], [5].
Objectives of these Projects were NPP hardware upgrading, personnel training,

nuclear safety codes improvements, accident analyses and accident management
procedures development.
The TACIS project R2.03/97 “Software Development for the WWER and RBMK
reactors”, managed by the Gruppo di Ricerca Nucleare San Piero a Grado
(GRNSPG) of the University of Pisa (UNIPI) during 2005 and 2006, concerned the
nuclear safety codes improvements and the accident management procedures
developement. An important part of the work for this PhD thesis was developed in
the framework of this Project, in collaboration with Russian RBMK designers
(NIKIET), researchers (Russian Research Center “Kurchatov Institute”) and
licensee (RosEnergoAtom, Russian nuclear electric utility, now Energoatom
Concern OJSC).
At the beginning of the present PhD thesis (January 2005), RBMK systems were
thoroughly analyzed by a series of mainly thermal-hydraulic or stand-alone
neutronic codes, e.g. [87], [88], [89]. Coupled codes analyses were also executed
by few leading Western Institutions (e.g., Idaho National Laboratory, USA, or GRS,
Germany, see [98], [99], [100]) and by Lithuanian Institutions (e.g., Lithuanian
Energy Institute (LEI), see [26], [77], [78]).
Chernobyl-type events were anymore simulated and most of the analyses found in
the literature, were executed in the aftermath of the Chernobyl disaster, with
simplified (for the today’s standards) numerical tools (see section 6.3).
1
1.2. Scope
The safety of nuclear power plants, understood as its capability to keep the
radiation exposure of personal and population within specified limits, is ensured by
maintaining the integrity of safety barriers, which are part of the plant defence in
depth concept. A series of barriers prevents the release of radioactive fission
products from their source beyond the reactor containment and into the
environment. In analyzing the NPP safety, it is essential to assess the integrity of
these barriers and to decide to what degree the response of the whole NPP and its
systems to a certain initiating event is acceptable from the viewpoint of the plant
safety. For the sake of simplicity and clarity, the integrity of the safety barriers is
related to certain threshold values, which are referred to as acceptance criteria.
Essentially these are the design limits for design basis accidents (DBA), adopted
with a conservative margin so that the safety barrier integrity is guaranteed as long
as the parameters do not exceed the relevant criteria.
Safety analysis for an RBMK NPP should assess the integrity of the following
barriers in the path of radioactivity transport and release:
• Fuel matrix.
• Fuel cladding.
• Circulation circuit pressure boundary and, in particular, the components
most susceptible to damage, namely fuel channel (pressure) tubes.
• Metal structures forming the reactor cavity.
• Structural components of the leak-tight Accident Localization System (ALS)
compartments and other buildings of the NPP housing circulation circuit
pipelines.
Should any safety barrier fail, thus opening the pathway for the release of
radioactivity beyond the plant boundaries, the amount of radioactivity and the
population exposure should be assessed. Beyond design basis accidents (BDBA)
are analyzed for the following purposes:
a) To assess the degree of reactor protection and the time available for taking
countermeasures.
b) To determine the emergency and other signals available to the operator for
identifying the plant status and to devise appropriate accident management
steps.
c) To develop a package of organizational and technical measures
(management strategy) for prevention and mitigation of the accident
consequences.
d) To assess the possible consequences of the BDBA as input information for
planning protection of the population and personnel.
According to the requirements of the Russian nuclear regulatory authority

RosTechnadzor (Russian Federal Ecological, Technological and Nuclear
Inspectorate, the former Gosatomnadzor (GAN)), both deterministic and
2
probabilistic approaches for BDBA analysis should be used. According to
requirements OPB-88/97:
• The estimated probability of an event with large release should be

less than 10-7 within one reactor year.
• The estimated probability of severe reactor core degradation or
melting should be less than 10-5 within one reactor year.
However, if some initiating events can lead to severe consequences and the inside
features of reactor are not able to prevent this, accident mitigation means should
be foreseen without drawing the attention to the probability of these events.
The deterministic approach for analysis BDBA is therefore also very important.
According to the Russian nuclear regulatory authority RosTechnadzor the
deterministic approach should be based on the method of postulating accidental
conditions. This method is based on determination of connections between the
plant conditions, level of severity of accident consequences (how many physical
barriers are violated: fuel assembly including fuel pellet and fuel cladding, pressure
tube, reactor cavity, main circulation circuit, accident localization system) and
availability of critical safety functions.
Thus, according to requirements OPB-88/97 of the Russian nuclear regulatory

authority RosTechnadzor, a list of BDBA scenarios for further detailed
investigations should be developed. The analysis results will be the basis for the
development of the accident management program for RBMK reactors.
1.3. Objectives
The key objective and the key products of the present PhD Thesis are:
1) the development of a set of advanced Best-Estimate (BE) computational
models (i.e, codes input-decks) for the safety analyses of the RBMK,
focusing on the core behavior;
2) the execution of safety analyses, investigating a broad spectrum of
accidents, comprehensive of both DBA and BDBA
1.4. Structure
The road-map of the PhD activities is outlined in the Chapter 4 and hereafter the
structure of this document is presented. Eight main chapters (including this one)
constitute the PhD thesis report. The contents and the reasons for each main
chapter are outlined in the following.
1) Chapter 2. This chapter deals with the description of the RBMK system
focusing on the safety features. The description essentially duplicates
information available in the literature. However, the self-standing nature of
the PhD thesis report and the need to avoid repeating NPP data in the
other chapters suggested the consideration of this chapter. Some basic
3
considerations about the neutronics characteristics were derived during the
PhD activities and are reported here.
2) Chapter 3. The status of RBMK safety technology and the related needs
are discussed in this chapter. The outcomes contributed to finalizing the
research and development activities. Namely, accident scenarios were
selected whose expected phenomena are envisaged to encompass the
relevant RBMK safety technology areas for the core. The list of transients
by itself constitutes a result from the PhD Thesis.
3) Chapter 4. This Chapter deals with the presentation of the chain of codes
used and the methodologies for codes application. A short description of
codes used is also given. Then, the description of the developed codes
input decks and their qualification according to GRNSPG/UNIPI criteria are
reported.
4) Chapter 5. This Chapter reports the results for some Hot Full Power (HFP)
realistic transient analyses. Calculations results for CR and CR group
withdrawal, Group Distribution Header (GDH) Loss of Coolant Accident
(LOCA) and GDH Blockage, Control and Protection System (CPS) LOCA
and Fuel Channel (FC) blockage accident are given. Results were
obtained mainly by the use of 3D Neutron Kinetics (NK) coupled thermal-
hydraulics (TH) system code RELAP5-3D©. In the case of the FC
Blockage scenario, calculations were supported by the use of Monte Carlo
MCNP5 code.
5) Chapter 6. The Chapter deals with the results of the accident analyses
performed at low power conditions, in order to simulate a “Chernobyl-type”
event (Loss of Main Circulation Pumps (MCP) at low power with Xenon
poisoned reactor). Calculations were performed using RELAP5-3D 3D NK
TH code. DRAGON lattice physics code was used in order to derive Xenon
cell nuclear Cross Sections (XSecs).
6) Chapter 7. The Chapter reports the list of the main results achieved during
the execution of the PhD activities.
7) References. The list of documents mentioned in the main body of this
report is given.
4
2. THE DESCRIPTION OF THE RBMK
Twelve RBMK units are installed in Russia and Lithuania, [6]. In the case of Russia
the RBMK contribute for about 50% of the electricity generated by the nuclear
source. Information of general validity for the RBMK systems is provided below.
However, specific information is related to the reference reactor for the PhD
activities, i.e. Smolensk 3 NPP. This is a “3rd generation RBMK reactor”
constructed in the Russian Federation about 400 Km Southwest of Moscow. The
unit 3 was put into operation on January 17, 1990. Relevant information for the
chapter is derived from refs. [7] and [9], the last one dealing with the Ignalina NPP.
Additional details about RBMK generations can be found in section 3.2.1.5.
An overall sketch of the RBMK NPP can be seen in Fig. 1 and Fig. 2 and key
system data are given in Tab. 1. Information is provided hereafter about system
hardware relevant for safety technology.
1) Core
2) Water pipe
3) Reactor bottom plate
4) Group Distribution Header
5) Shield
6) Steam Drum
7) Steam Water Pipe
8) Reactor top plate
9) Refuelling machine
10) Reactor lid structure
11) Top reactor channel
12) Downcomers
13) Pressure Header
14) Suction header
15) Main Coolant Pump
Fig. 1 – Overall view of the RBMK NPP system
5
Fig. 2 – Smolensk-3 NPP view
6
Tab. 1 – Key parameters of the Smolensk-3 RBMK-1000
Key-parameters of core
Power, MWth Total Left Part Right Part
3213 1629 1583
Flow rate Kg/s 10118 5158 5104
Number of FC 1568 794 774
Number FC with AA 90 45 45
Fuel Channel parameters
Full length, mm 18325
Number of channels 1660
Upper part diameter, mm and thickness, mm 95x5
Middle part diameter, mm and thickness, mm 88x4
Lower part diameter, mm and thickness, mm 60x5.5
Fuel rod overall length, m 3.46
Active core length, m 6.92
Spacer/support grids 20 per FC
Key dimensions of the down-comers
Diameter, mm and thickness, mm 325 x 15
Average length, m 29.15
3
Total volume, m 47.7
Key dimensions of the Steam Header
Diameter, mm and thickness, mm 1020x60
Length, mm 21 080
3
Volume m 17.3
3
Volume cylindrical part, m 13.4
o
Key MCP thermal-hydraulic parameters (at 265 C)
3
Q, m /h (Kg/s) 8000±200 (2.22)
Head, m 200±20
P inlet, bars 70.5
Power, MW 4.3±0.3
Speed, rpm 1000
Torque, N-m 39400
Key-parameters of Group Distribution Header
Length, mm 5400
Design pressure, bar 100
Design flow, t/h 1700
3
Total volume of GDH, m 16.5
Key-parameters of steam lines
Average length, m 32
Design flow, t/h 40
Design pressure, bar 75
Key-parameters of Drum Separators
Number per reactor 4
DS length, m 30.98
DS internal diameter, m 2.6
3
DS volume, m 150.2
Operation pressure, bar 68.9
Design pressure, bar 73.5
Nominal water level (above the central axis) mm 380±50
7
2.1. The Primary System
2.1.1. The pressure boundary
The MCC consists of two loops, whose components are arranged symmetrically
with respect to the vertical axis of the reactor. Each loop has two separator drums
(1) (number in parenthesis refers to legend of Fig. 3, which separate the steam
from the steam-water mixture exiting from the core block. The separator drums are
horizontal cylindrical vessels interconnected both at the lower liquid level and at the
upper steam level.
15
1
14
13
2
3
8 10
From 12
4 ECCS
7 11
5 6
1-separator drum, 2-downcomers, 3-suction header, 4-suction piping of the MCP,

5-MCP, 6-pressure piping of the MCP, 7-bypass between headers, 8-pressure
header, 9-group distribution header with flow limiter, check valve and mixer, 10-
water bottom piping, 11-fuel channel before the core, 12-fuel channel within the
core, 13-fuel channel above the core, 14-steam water pipes, 15-steam pipelines.
Fig. 3 – RBMK Main Circulation Circuit (MCC) flow diagram
In the bottom section of each separator drum (SD) a feed-water header is installed
which provides through special mixers feed-water to the down-comer pipes. The
separated water mixed with the feed-water reaches the suction header (3) through
24 down-comer pipes per each loop (2). The flow path continues through four
suction pipes (4) connected to the four Main Circulation pumps (MCP). During
normal reactor operation, only three pumps are operating in each loop, the fourth is
a reserve. They are single stage vertical and centrifugal type.
8
From the MCP, water flows through pressure header pipes (6) to the pressure
header, PH, (8). The suction and pressure headers are connected by bypass lines
(7), provided with a gate and a check valve. The bypass ensures that natural
circulation of the coolant takes place in case of main circulation pumps shut-off.
From the pressure header (8) water continues through twenty two pipes to twenty
two group distribution headers (GDH) (9). Mechanical filters are provided inside the
pressure header while at GDH exit there are a flow limiter, a check valve and a
mixer to which ECCS are connected.
Each GDH is connected to liquid pipes (10) leading to an average of forty-five fuel
channels (12). The flow in each pipe, and therefore in each fuel channel (12), is set
by isolation and control valves and is measured by a ball flow-meter. The steam-
water mixture generated in the fuel channel flows through the steam-water pipes
(14) to the separator drums (1).
The elevations for the most important components of the MCC are presented
schematically in Fig. 4. The total, top-to-bottom elevation of the primary system is
over 30 m. The elevation driving the natural circulation loop, measured from the
bottom of the core to the bottom of the separator drums, is ~ 21m. These large
elevation heads determine the flow parameters of the system under natural
circulation conditions.
Fig. 4 – Key elevations in the MCC of the RBMK system (Smolensk 3)
9
From the separator drums the generated steam is directed to the turbines.
Discharge steam from the turbines is accumulated in condensers, from there the
condensate flows down through filters, heaters and deaerators to the main feed
water pump and is finally returned to the separator drums.
The purification and the cooling of the primary cooling circuit water is performed by
the Purification and Cooling System (PCS) which is an equivalent of the Chemical
and Volume Control System (CVCS) in Western LWR. Part of the water is taken
from the MCC, cooled down and filtered by a mechanical filter and by an ion-
exchanger in the purification bypass. The treated water then joins the feed-water
flow. The reactor also contains a number of channels for control rods and metering
devices. These are cooled by a separate circulation system, which is called the
Control Rod Cooling Circuit (CRCC).
The twelve down-comer pipes direct the water from the separator drums to the
suction header, thus each of the two circulation loops contains twenty-four down-
comers. The geometrical dimensions of each down-comer are given in Tab. 1. The
suction headers collect water from the down-comers and supply coolant to four
suction pipes in such a way a mixing of the water from the drum separator occurs.
The geometrical dimensions of the suction headers are also given in Tab. 1.
For the forced circulation of cooling water through the RBMK reactor type CVN-8
(‘wet stator’ type) MCP are employed, Fig. 5. The CVN-8 type is centrifugal,
vertical, single-stage pump with a sealed shaft. Four pumps are installed per each
loop, three of them are normally in operation, and the fourth is kept as a reserve.
The basic parameter of the pump at nominal conditions are reported in Tab. 1 and
the pump curves for hot and cold water are given in Fig. 6, left and right sides,
respectively.
1-service platform, 2-
flywheel, 3- electric motor,
4- junction coupling, 5-
support of electric motor,
6- foundation frame, 7-
tank of the pump, 8- water
outlet, 9- water inlet
Fig. 5 – The MCP for RBMK NPP.
10
N, kW
H, m 40 92
5000 250 90
35
4500
88
4000 2 200 30
3 86
3500
N 25 η
3000 150 84
H
2500 20 82
2000 100
∆h 80
1500 15 4
1000 50 78
10
500 1 76
0 0 5
74
5 6 7 8 9 10 11 12 Q, 103 m3/h
0 72
3 3
5 6 7 8 9 10 11 12 Q 10 m /h
∆ h, al. η,%
40 90 N, kW H, m
7000 250
35 80
30 70
6000
60
25 η 200
4 50 5000
20
∆h 40
15
30 4000 H 3 150
N
10 20
5 3000
1 10 100
0 0
3 3 2000
5 6 7 8 9 10 11 12 Q, 10 m /h 2
50
1000
0 0
1. Cavitations curvecurve
1 - cavitation Q-Dh (Q-∆h) 5 6 7 8 9 10 11 12
2.power
2- Power curve
curve Q-N
(Q-N) Q 103 m3/h
1- cavitation curve (Q-∆h)
3-
3. Hydraulic performance (Q-H)
hydraulic performance curve curve Q-H 2- power curve (Q-H)
4-
4.MCP
MCP discharge–efficiency relationship
discharge-efficiency (Q-η) Q-h
relationship
3- hydraulic performance curve (Q-N)
4- MCP discharge–efficiency relationship (Q-η)
Fig. 6 – RBMK MCP (CVN-8) performance data (left: high temperature conditions,
265 ºC; right: low temperature conditions, 90 °C)
The main function of each PH is to collect the water that comes from the pressure
pipes connected to all main pumps exit and supply the coolant to the twenty-two
GDH through twenty-two pipes. The pressure header has 1040 mm outside
diameter and 70 mm thick walls.
The coolant is distributed to individual GDH by means of twenty-two pipes 325x15

mm. Each pipe has a manual control gate valve, a check valve and a mixer to mix
(in case of accident) the cold water from the Emergency Core Cooling System
(ECCS) and the hot water from the MCC. The check valve prevents back-flow from
the fuel channels in case of failure of the pressure header. Mixers protect the MCC
from thermal or hydraulic shocks. Flanges designed to prevent pipe whip in the
event of a pressure surge are fixed to the structural beams of the plant and to a
special framework
The coolant is supplied to the individual fuel channels via group distribution header,
Fig. 2.6 which are horizontal cylinders, Tab. 3.1 The GDH are securely fastened to
support structures to prevent any sliding in case of failure. Each header distributes
coolant to 40-43 bottom water pipes. These pipes are provided with isolation and
11
control valves between the GDH outlet and the fuel channel inlet. Isolation and
control valves are used to adjust channel flow on the basis of channel power. Flow
rates can be controlled by varying the flow-area of the valves. This is achieved by
manual operation from a separate room in the vicinity of the reactor block. The
bottom water pipes leading into the reactor block drive the water into the fuel
channels that remove the heat generated in the fuel assemblies. The liquid water
starts to boil and the steam-water mixture flows from the top of the fuel channels by
way of individual steam - water pipe to the separator drums. The exiting steam-
water pipes include several bends. This aids in reducing gamma streaming.
1 – isolation and control valve, 2 – ball type flow-rate

meter, 3 – coolant water pipe leading to the fuel
channel, 4 – group distribution header
Fig. 7 – Geometrical configuration of RBMK GDH
12
a)
b)
Fig. 8 – Geometrical configuration of RBMK SD: a) overall view; b) cross section
The separator drum has the following functions:

• separation of steam from the steam-water mixture flowing from the fuel
channels,
• mixing of the separated water with feed-water,
• storing of coolant for the MCC.
The steam water mixture arrives at the separator drum, through inlet pipes (9), (see
Fig. 9 for numbering). A part of the steam is separated in the distribution
compartments because the flow looses its kinetic energy in impact to the special
plates (3).
13
Fig. 9 – Steam Drum – Component Specifications
This steam then penetrates the submerged perforated sheet (4) and the
“barbotage” layer above it. Final separation occurs because of gravity force.
The separated steam goes through the perforations of the upper shields (5) into
the steam-flow piping, the separated liquid flows downward from pipes at the
bottom (13). The feed-water line has a nominal diameter of 500 mm (10). It enters
into the separator drum at a 45 degree angle, and extends to a distribution header
in the lower part of the drum. The feed-water is injected from the header (6) (7) to
the down-comer (13) via jet spray nozzle in order to cool the water to be supplied
to the MCP. Both the steam and liquid containing regions of the two separator
drums are connected by number of pipes. The non uniform generation of power in
fuel channels can lead to a non homogeneous steam-water distribution in the
steam drum. This requires design features which serve to reduce both transverse
and longitudinal variations of the steam content. This is accomplished by a
submerged perforated sheet (4) with a 150 mm thick downward frame. A down-flow
passages is provided between the frame and the drum wall for the part of the
water, which penetrates the perforations together with steam. The down flow
passages functions as a hydraulic lock against any penetrations of steam at the
side of the perforated sheet. The sink is covered by safety plates spaced at 75 mm
from the frame. Traverse and circumferential variations of pressure at the entrance
of the steam pipes are reduced by a similar perforated shield in the upper part of
the drum (5) and by inside diameter bushing installed in the steam outlet pipes.
The two separator drums within each loop are interconnected both in the liquid and
steam region, Fig. 10. There are five connecting (300x15) mm pipes in the steam
14
zone and four pipes (300x15) mm in the water zone. The total length of the pipes is
19.8m in the water zone and 16.2m in the steam zone. These connections ensure
that equal water levels and steam pressures are maintained in both drums.
1 – separator drums, 2 –
connecting pipes at the
water level, 3 – connecting
pipes at the steam level, 4 –
submerged perforated sheet
Fig. 10 – Layout of the two DS and related connection
2.1.2. The core

2.1.2.1. The fuel
The RBMK core is mounted in a graphite stack with a lattice of vertical channels
loaded by fuel assemblies, additional absorbers and regulators. A ring of periphery
channels forms a radial reflector. Top and bottom reflectors are core-wide and are
0.5 m width each one. The graphite stack is enclosed in a space, named reactor
cavity, formed by lower and upper metal constructions and ringed by a steel
cylindrical structure, see section 2.2.1. The graphite stack of the RBMK-1000
reactors serves several functions. The primary one is the neutron moderation and
reflection, but it also provides structural integrity and in the event of a temporary
cooling malfunction, a relatively large heat capacity. A general view of the core
region and of a portion of the graphite stack are given in Fig. 11. More details in
Tab. 1.
The FC together with the fuel rod is reported in Fig. 12. A key feature for the RBMK
is the FC housed by graphite blocks. The core contains 1570 fuelled channels
separated from its nearest neighbors by the wall of the pressure tube and graphite
blocks. Each pressure tube has considerable autonomy. For example the coolant
flow rate of each tube is controlled online by an individual isolation and control
valve. In such a way any pressure tube can be isolated from the rest of the primary
cooling system while the reactor remains under operation. This peculiarity makes
possible to change fuel clusters online and also has a significant impact on the
potential consequences of loss-of-coolant accidents.
A fuel assembly consists of an upper and a lower fuel bundle, symmetrically placed
around the core middle plane. Each section is formed by 18 active rods and a
central steel rod that ensures structural rigidity together with 11 spacer grids
equally distributed along the section. Each fuel rod has a zirconium alloy clad that
recovers the fuel pins, see Fig. 12. The top, centre and bottom segments of a
typical reactor fuel channel are shown schematically in Fig. 13. The central
15
segment usually called Pressure Tube (PT) is made of a zirconium-niobium alloy
(Zr+2.5% Nb). The top (3) and bottom (11) segments are stainless steel. The
choice of zirconium-niobium for the centre part was made because of the relatively
low thermal neutron absorption cross section of the material and its adequate
mechanical and anticorrosive properties at high temperature (up to 350 oC).
1 – top ring,
2 – diaphragm,
3 – bottom ring,
4 – bushing,
5 – steel support
plates,
6 – graphite rods,
7 – shield plates,
8 – standpipe,
9 – guide pipe,
10 – reinforcing tube
(reflector cooling),
11 – outer steel shell,
12 – ring plate
Fig. 11 – Geometrical configuration of RBMK core: a) detail of active region; b)

overall view
The interaction of fast neutrons can lead to dimensional changes in various

materials. For example, in graphite moderated reactors, initial accumulation of the
fast neutron dose produces a gradual shrinkage of the graphite blocks. For the
RBMK reactors this results in a decrease of the bore diameter through which the
fuel channel passes. For the pressure tube made of a Zr+2.5% alloy, the effect is
opposite, due to thermal and irradiation effects the tube diameter increases. As a
result, the gap between the pressure tube and the graphite, which has a nominal
thickness of 1.5 mm is gradually diminished leading to an eventual closure of the
gap itself.
Referring to the Smolensk 3 plant data, [7], the reactor core is composed by 2488
graphite columns, of which 1570 are fuel channel columns and 314 are non-fuel
channels columns. The non-fuel channel columns are subdivided as follows: 211
Control and Protection System (CPS) channels, 12 ADC, 90 AAC and 1 water
column. In addition, 604 are Radial Reflector Channels (RRC). The reference
height for FC is 8.0 m, however the core active region is 7.0 m and the graphite top
and bottom reflector have 0.3 m width.
16
1 – suspension bracket, 2 – top plug, 3 –
adapter, 4 – connecting rod,
5 – fuel element, 6 – carrier rod, 7 – end
sleeve, 8 – end cap, 9 – retaining nut.
a) b)
Fig. 12 – Geometrical configuration of a) RBMK fuel bundle, b) fuel rod
1 – steel biological shield

plug, 2,10 – top and bottom
metal structures,
respectively, 3 – top part of
the fuel channel, 4 –
welding-support ledge, 5 –
fuel assembly support
bracket, 6 – encasement
cylinder, 7 – seal plug, 8 –
graphite cylinder, 9 –
central part of the channel,
11 – bottom part of the
channel, 12 – thermal
expansion bellows
compensator, 13 – stuffing
box
Fig. 13 – Geometrical and material configuration of RBMK FC
17
2.1.2.2. The Absorbers and the Power Control System
AAC are located symmetrically around core midline. 90 AAC are placed inside the
core in order to increase the Operative Reactivity Margin (ORM). Regular AAC and
cluster AAC (two FC) shall be distinguished that have the configuration (cross
section) depicted in Fig. 14. Namely the cluster-type additional absorber consists of
two bundles, where inner displacer is surrounded by 18 small-size, stainless-steel
tubes, 2 tubes are empty while 16 tubes are filled with boron carbide (density 1678
Kg/m3). The height of the column is 3.50 m and the distance between boron
columns of two bundles in the middle of the core is 0.044m. The displacer is a tube
with a graphite block inside.
1 – support tube, 2 – absorbing

sleeves, 3 – channel tube
a) regular
1 – graphite displacer, 2 –
support tube, 3 – absorbing pins,
4 – empty tube, 5 – channel
tube
b) cluster
Fig. 14 – Geometrical and material configuration of regular and cluster type AAC of
RBMK core
Control rods are inserted in special channels of the CPS that is hydraulically
independent from the fuel channel system. The CPS can be kept in operation in
case of emergency when the FC fail. There are three types of control rods: a) 32
short CR (SHR), b) 24 safety or fast scram CR (SR), c) 155 manual CR (MCR).
SHR move upward from the bottom when inserted, all others move downward from
the core top. A sketch of the three CR types is given in Fig. 15 and a cross section
(for regular CR) in Fig. 16.
18
a) Short CR b) Safety CR
19
c) Manual CR
Fig. 15 – Longitudinal view of RBMK CRs (Short, Safety and Manual types)
20
Fig. 16 – Cross sectional view of RBMK regular CR
Control rods are placed in an independent cooling loop, as already mentioned, with
its own pumps and heat carrier (see Fig. 17).
1 – top storage tank

circulation tank; 2 –
Circulation pump (CPS-P); 3
– Emergency tanks; 4 –
Distribution header; 5 – CPS
channel, 6 Reflector cooling
channel , 7 – Scram channel,
8 – CPS discharge collector,
9 – CPS heat exchanger, 10
– Supporting device, 11 –
CPS-P bypass with valves
for flow-rate regulation, 12 –
channel (train) for insertion
of nitrogen in the scram
channel, 13 – line for de-
aeration, 14 – excess water
sink overflow pipe, 15 –
mechanical filter, 16 –
Distribution header, 17 –
purification bypass, 18 –
Reflector cooling channel
distribution header.
Fig. 17 – Flow diagram of the CPS cooling system
The piping of Control Rod Cooling Circuit (CRCC) cools the channels of the CPS
rods, fission chambers, Power Density Distribution Monitoring System (PDDMS)
21
sensor and Radial Reflector Cooling Channels (RRCC). The coolant must be
distributed in the circuit to ensure that proper temperature fields prevail in the
channels of the control meters and of the graphite moderator. The cooling in this
circuit is gravity driven, see Fig. 17. The cooling circuit includes: a) 211 CPS
channels containing rods and safety instrumentation, b) 12 channels containing the
in-core Power Density Sensors of the Axial Monitoring (PDMS-A), c) 604 channels
of RRCC.
2.1.2.3. The scram signals

Several modes of scram operation are possible for the Smolensk-3 NPP and are
summarized in Tab. 2. The BAZ mode is initiated from both neutronic signals
(excess power and reduced period), two process trips (increase of pressure in
selected rooms and increase of reactor cavity pressure) and by manual action. The
more normal shutdown mode of AZ-1 is initiated by 16 process parameters and by
failure of more than 3 local protection zones to reduce power satisfactorily. The
other modes are initiated by a variety of upset conditions (see [8] for more info).
Tab. 2 – CPS Modes of Operation
Rods Used /
Name Conditions Power reduction
Insertion Times
24 SR + All CRs Operating power to
BAZ 5 initiating signals
<2.5s 12-14 s zero
24 SR + All CRs Operating power to
AZ-1 17 initiating signals
7s 12-14 s zero
3 initiating signals
LAC CR power deviation
From 100% to 50 %
AZ-3 0.4 m/s insertion <1% (LAR)
at 2%/s
0.2 M/s withdrawal power deviation
<5% (LAP)
From 100% to 60%
AZ-4 LAC CRs 6 initiating signals
at 1%/s
AZ-5 NOT USED
Operates until LAP
Local Protection
AZ-6 LAC CRs signal disappears
1 initiating signal
(up to 50% power)
The following should be noted concerning the CPS CR grouping:

• 24 SR belong to the fast active scram system (FASS)
• 9 MCR are automatically moved for controlling the local power (LAR)
• 32 ShR (inserted from the bottom)
• 146 MCR including 18 CR of the Local Automatic Protection (LAP)
• 27 rods are provided for the local automatic control (LAC) which consists of
18 LAP CR and 9 LAR CR (LAC= LAP + LAR)
• 4 of the peripheral LARs are used for automatic control of the power level
under 10% of nominal power.
This information gives an idea of the complexity of the power control for RBMK
system.
22
2.1.2.4. The fuel cell neutronic characteristics
Several considerations on the RBMK lattice design can help in obtaining the core
neutronics characteristics. The use of graphite as moderator gives the opportunity
to use natural or (in the case of RBMK) slightly enriched uranium as fuel thanks to
its low Capture Cross Section (Sc = 2.4*10-4 cm-1). On the other hand, the weaker
slowing down power of graphite (ξSs of C is 400 smaller than ξSs of H2O) implies a
greater neutron thermalization length, resulting in the use of large lattice pitch (25
cm for the RBMK). The relevant neutronics characteristics of the graphite,
compared to the light and heavy water are given in Tab. 3.
Tab. 3 – Comparison between diffusion parameters of Graphite, Light Water and
Heavy Water
Density D Sa L tth1/2 M
Moderator ξSs ξSs/Sa
(g/cm3) (cm) (cm-1) (cm) (cm) (cm)
Graphite 1.6 0.84 0.06 175 2.4*10-4 59 19 62
D2O 1.1 0.87 0.18 6670 2.9*10-5 170 11.4 170
H2O 1.0 0.16 1.5 70 2.9*10-2 2.9 5.1 5.8
Therefore, for the graphite, the mean-squared distance that a neutron travels from
birth as fast fission until the “death” by capture as a thermal neutron is:
r 2 = 6(τ th + L2 ) ≡ 6M 2 = 173400 cm2,
resulting in an average distance of r = 416 cm.
Hence, considering the reactor lattice main parameters, diffusion parameters can
be recalculated obtaining representative values for the whole reactor core. In Tab.
4, the diffusion parameters for RBMK and LWR cores are given.
Tab. 4 – Comparison of diffusion parameters for RBMK and LWRs
Reactor L (cm) tth1/2 (cm) M (cm) Diameter (L) Diameter (M)

RBMK 14.1 14.4 20.2 100 69
BWR 2.2 7.1 7.3 180 50
PWR 1.8 6.3 6.6 190 56
From Tab. 4, it is clearly evident the reasons of the greater geometrical dimensions
of RBMK reactors compared with those of LWR of comparable thermal power.
Neutronic characteristics have some immediate consequences in the MCC
hardware design (e.g., no such large reactor pressure vessel can be constructed)
and in the containment feature (e.g., no full pressure containment). The safety
considerations about these issues are reported later in this document.
Coming back to the RBMK neutron cell characteristics, other relevant neutronics
parameters can be easily calculated considering the elementary fuel element (FA
and graphite) reproduced in Fig. 18.
23
Fig. 18 – Fuel Cell basic model
Monte Carlo analyses of the Fuel Cell (see section 4.5.5 of this report), showed the
following interesting characteristics reported in Tab. 5.
Tab. 5 – Neutronic characteristics of a RBMK fuel cell
Neutronic characteristics of a RBMK fuel cell
Percentages of fissions thermal range (E<0.625 eV): 91.67%

caused by neutrons for intermediate range (0.625 eV< E < 100 KeV): 5.39%
various energy ranges fast range (E>100 KeV): 2.94%
Average number of
neutrons produced per 2.44
fissions (n)
Average fission neutrons
produced per neutron 1.57
absorbed in the fuel (η)
Energy corresponding to
the average neutron 0.0933 eV
lethargy causing fission
24
Monte Carlo calculations shows also the distribution of the thermal, fast and total
flux into the fuel cell, see Fig. 19, Fig. 20 and Fig. 21.
Fig. 19 – Thermal neutron flux distribution
It is clear in Fig. 19 the neutron thermalization and accumulation in the graphite

brick. In Fig. 20 it is instead clear the born of the fast neutrons into the fuel rods
and their loss of energy into the zone of graphite close to the pressure tube.
Fig. 20 – Fast neutron flux
Total flux distribution is showed in the Fig. 21. The accumulation of the neutron flux
in the graphite is also here very clear.
25
Fig. 21 – Neutron flux distribution
Tab. 6 is finally summarizing the fuel performance and the core power density,
comparing it with those of a 900 MWe BWR.
Tab. 6 – Comparison between BWR and RBMK core parameters
FA FA
Thermal Average Maximum Maximum
Reactor Power Density
Power Linear Linear Burnup
Type (KW/l)
(MW) Power Power (GWd/tonHM)
(KW/m) (KW/m)
2.9
RBMK
(56 considering
(Smolensk- 3159 14.8 31 22
only the coolant
3)
volume)
BWR-6 2894 56 19.8 44 ~54
2.1.2.5. Reactivity Control

The dynamic analysis of large and strongly heterogeneous cores like those of the
RBMK reactors is quite complex. In the following sections, the most relevant
peculiarities and parameters affecting RBMK neutron kinetics are reported and
discussed.
2.1.2.5.1. Fuel Cell Overmoderation
RBMK fuel cell neutronic design is based on a slightly over-moderated lattice. This
is clearly evident in the Fig. 22, where the kinf versus the ratio of the moderator are
on fuel area is given. The kinf values were obtained performing Monte Carlo
26
MCNP5 calculations for a RBMK and for a BWR typical fuel cell, changing the pitch
of the lattice.
Operating conditions of the BWR are in the under-moderated zone while the RBMK
are slightly in the over-moderated zone. This over-moderation has, as a
consequence, that a lack of water in the RBMK FCs is causing an increase of
reactivity, with all the obvious consequences for the reactor control and safety.
1.4
OPERATING
CONDITION
1.35
1.3
Kinf
1.25
RBMK 2.0% Fuel
RBMK 2.4% Fuel
1.2 BWR 2.9% Fuel
1.15
0 10 20 30 40 50 60
Amod/Afuel
Fig. 22 – kinf versus moderation ratio of Moderator area on Fuel area
The last generation of RBMK reactor, the Kursk-5 NPP (currently under
construction), is addressing this issue, using a new design for the fuel cell graphite
brick that removes the overmoderation [2].
2.1.2.5.2. The Operating Reactivity Margin

All reactors require an ORM in order to provide reactivity control as fuel burnup
increases and to allow power maneuvering. In most reactors this reactivity reserve
is provided by either CRs or by soluble or burnable poison (e.g., boron in the
moderator or Gadolium in the fuel pins). In the case of power reduction, sufficient
ORM is also required in order to compensate the loss of reactivity caused by the
Xenon build up.
In the case of channel reactors with on-power fuelling (like the CANDU or the
RBMK), a sufficient ORM is also necessary in order to provide with the CRs the
suppression of power peaks in the vicinity of the newly fuelled channel. Removal of
CRs can also provide reactivity for continued reactor operation in the event that the
refuelling machine is unavailable.
27
Therefore the ORM has a multi purpose function and the choice of its magnitude
during the reactor core design phase has an important consequence on the fuel
burnup and consequently on the economy of the plant.
In the RBMK the magnitude of the ORM has added safety significance. The
presence of absorbers in the core reduces the relative importance of neutron
absorption in the water in the FC. In the event of loss of that water during an
accident, the reactivity addition is smaller than would be the case with no in-core
absorbers. Thus the CRs used to provide the ORM, in conjunction with the
permanent in-core absorbers (the AAs) added to RBMK cores since the Chernobyl
accident, reduce the magnitude of the void reactivity effect. The upper magnitude
of the ORM, while specified as an operating limit, has safety implications in that
enough CRs must remain available out-of-core to provide the appropriate sub-
criticality margin to ensure the reactor can be shut down for the most reactive core
configuration possible.
Consideration of the above effects has resulted in an ORM of between 43 and 48

rods for Smolensk-3 [8]. A lower safety limit of 30 rods has been established so
that the shutdown system can overcome the positive void effect. Summarizing, the
four main reasons for being within these two bounding limits are:
• reactor dynamics are improved by reducing the reactivity void coefficient
(in conjunction with the added fixed AA);
• there must be enough CRs in the core to control the radial and axial power
distribution;
• there must be sufficient reactivity holdup within the inserted rods to
compensate for Xenon buildup during power maneuvering;
• the upper limit is required to ensure there is enough negative reactivity
available to provide guaranteed cold shutdown, in most reactive core
condition, upon insertion of all CRs.
If the ORM margin drops below 30 rods, the operator must manually scram the
reactor.
2.1.2.5.3. The Reactivity Coefficients

Measurement of reactivity coefficients is a complex task for RBMK reactors. They
strongly depend by the particular reactor conditions (e.g., average fuel burnup, CR
positions, number of AAs, etc.) and some of them (e.g., the core total voiding
coefficient) can never be practically measured. Therefore, some of the values given
in Tab. 7 were derived from a posteriori calculations or simply extrapolated by
small perturbations operated at the plant. They should be considered as
representative (especially the void coefficient) of the Smolensk-3 configuration
during the ‘90s.
28
Tab. 7 – Typical Reactivity coefficient for the Smolensk-3 NPP [8].
Reactivity Coefficient Value (pcm) Notes

Power (aw) -0.23/MWth
Doppler (at) -1.2/°C
Coolant Temperature
~0 /°C
(acool)
Graphite Temperature
4-5/°C Measured at cold conditions
(at)
or +0.3 β±0.2 β for full core
voiding (assuming β = 0.0055)
Void (aφ) 1.65/%
±0.2 β is the allowed
systematic and statistical error
The most important parameter monitored and controlled in regard ot neutronic

effects is generally the void reactivity coefficient. As a result of the safety
modifications made at Smolensk 3 following the Chernobyl accident, the void
reactivity coefficient with an ORM of between 43 to 48 rods at HFP SS conditions,
with 97 AAs, is maintained at a value < 0.3 β.
The void coefficient is an integral reactor parameter and it is generally done the
assumption that it remains constant throughout the whole range of void fraction. In
order to calculate the void reactivity effect this constant value is used assuming
total core voiding (0 to 100%). This explains why the total core voiding and the void
coefficient are both quoted frequently with the same value.
2.1.3. The Balance of Plant

An overall sketch of the RBMK BOP can be found in Fig. 23. Separated steam
flows from each of the 14 SD outlet nozzles, passes through 300 mm nominal
diameter pipes with 198 mm diameter flow restrictors to two 400 mm nominal
diameter steam headers situated parallel to the drum separator axis. Connections
to each header are from alternate nozzles along the separator length to provide
uniform steam extraction from the separators and uniform steam supply to each
header. For each reactor half, outlet pipes from the midpoints of each of the four
steam headers are connected in pairs, one steam header from each drum
separator, to form 600 mm nominal diameter main steam lines, as shown in Fig.
24.
The 600 mm nominal diameter main steam lines downstream of the drum
separator are each fitted with four 250 mm internal diameter branch connections,
pairs of which lead to main safety valves.
The four main steam lines therefore provide a total of 8 main safety valves, two per
line. These eight main safety valves each open and close at one of three pressure
set-points, two opening at 7.45 MPa and closing at 7.15 MPa, four opening at 7.55
MPa and closing at 7.25 MPa and two opening at 7.64 MPa and closing at 7.35
29
MPa, relieving steam through discharge pipes with hydrolocks into the pressure
suppression pool.
Downstream the branch connections to the main safety valves, two of the 600 mm
diameter main steam lines, from different halves of the reactor, are linked by a pipe
of 300 mm nominal diameter, Fig. 25. This link is provided to promote uniformity of
steam flow and pressure between the main steam lines from the reactor halves.
Additional smoothing of the steam pressure and flow is provided by the leveling
header of, where all four main steam lines are interconnected through flow limiters.
In this area, branch connections are made to four pressure reducing turbine bypass
valves (BRU-K) each with capacity of 725 t/h which relieve excess steam to the
turbine condenser during steam pressure transients, to avoid unnecessary opening
of the main safety valves.
The opening set points of the BRU-K valves is 7.1 MPa, with an opening/closing
time of 10 s, but once open are operated in a controlled mode to reduce drum
separator pressure back to 6.86 MPa.
15 19
5
17 20 21
4
16
18
11
7 6
8 33
1
9
3
26 22
2
от НДК
10 12 13 14 32 24
23
25
27
от САОР от САОР 31
в РГК 34
29 30
28 27
в РГК
1 – reactor; 2 – fuel channel; 3 – feeding pipelines; 4 – steam-water pipelines; 5 – DS; 6 – DC; 7 – MCP suction
header; 8 –MCP; 9 – bypass; 10 – MCP header; 11 – mechanical filter; 12 – flow limiter; 13 – GDH; 14 – control-
throttling valve; 15 – mixer; 16 – feed-water valves; 17 – steam header; 18 – ГПК; 19 - BRU-K; 20 – SCV; 21 –
turbo generator; 22 – condenser; 23 – condensate pumps; 24 – condensate polishing; 25 – heater; 26 –
deaerator; 27 – AFWP; 28 – FWP; 29 – make up regenerator; 30 – cool down pump; 31 – make up cool down;
32 – bypass clearing; 33 – bubbler; 34 – water tank
Fig. 23 – Flow diagram of the BOP system of the RBMK core
30
Fig. 24 – Flow diagram of SL from the SD region till the MSVs
Fig. 25 – Flow diagram of RBMK SL from the MSIV till the turbines
When the load of either turbine is rejected, a signal to open both BRU-K valves
associated with that turbine is generated by the AZ-3 signal, which reduces reactor
power to 50%. This signal anticipates a subsequent pressure increase and
controlled flow through the valves maintains the drum separator pressure close to
31
the nominal value. The same logic is used for load rejection by both turbines or a
single operating turbine through the AZ-1 scram signal.
Branch connections are also provided to six pressure reducing valves (BRU-SN)
with a capacity of 100 t/h each, supplying steam to the unit auxiliaries. These
valves are used during depressurization and cool down to bleed steam through the
technological condenser, cooled by service water, to maintain the deaerator level
for use by the emergency feed-water pumps.
Downstream of the BRU-K and BRU-SN branch connections, the four 600 mm
nominal diameters are routed in pairs to each turbine, each dividing into two of 400
mm nominal diameter. The 400 mm diameter pipes are each fitted with two turbine
isolation valves. Of these, the upstream valve is closed for maintenance purposes
only and the downstream valve is the main stream isolation and control valve
during plant operation.
Pairs of main steam isolation and control valves are bypassed by linking pipes,
each ling fitted a slide valve and control valve. This bypass line is intended for pre-
operational testing and all bypass valves are closed during power operation.
Downstream of the main steam isolation valves two turbines are connected. The
exhaust steam is condensed, passes through the deaerators, the pre-heaters and
finally it is pumped by the feed-water pump into the SD.
2.2. The Confinement System

The RBMK confinement constitutes a very complex system. The following main
subsystems can be distinguished which contribute to mitigate or preventing
possible radioactivity releases from the MCC, Fig. 26 and Fig. 27: a) the reactor
cavity, b) the ALS, c) the reactor building, d) the fuel loading and unloading
machine hall, e) the turbine hall.
ENVIRONMENT
9
Fuel loading Reactor
Turbine
machine hall Building
ALS Hall
Top core plate 6

High energy
pipeline or
4 1Reactor
8
RC
component Cavity
Pool
Fig. 26 – Sketch of the confinement system of the RBMK NPP with main sub-
systems (numbers refer to components identified in Fig. 3.1, where applicable)
32
The RC is connected to the ALS through the pool suited for lowering the pressure
in case of accident inside the RC itself. The RC is connected to the environment
via valves that open in case of high pressure where high pressure is such to have
the potential of lifting the upper core plate. After opening these valves do not close.
The RC can be connected with the hall of the fuel loading and unloading machine
only in case of lifting of the upper core plate.
The ALS is connected with the RC and the environment. This last connection
occurs through valves that open at assigned pressure set-points. The ALS allows
the characterization of the break location and the retention of the FP in case of
mass and energy releases from the MCC. The connection via energized flow line
(SL) between ALS, RB and Turbine Hall is depicted by a dotted line in Fig. 27. The
zone of the fuel loading and un-loading machine is connected to environment by a
bare line to show that there is not pressure boundary between this zone and the
environment.
The RB and the Turbine Hall are characterized in Fig. 27 by bare lines connections
with the environment for the same reason as above (no pressure boundary
separation between each of these and the environment). However high energy flow
lines or components are located in these regions as depicted by red dotted line in
Fig. 26.
Fig. 27 – Cross sectional view of the RBMK confinement system
33
2.2.1. The Reactor Cavity
The cavity is a metal structure made up of a lateral cylindrical shroud a top and a
bottom metal structures, Fig. 28. The lateral shroud is fitted with a lenticular
compensator to compensate for longitudinal heat expansion. The pipes that are
connected with the reactor cavity (e.g. fuel channels integrity monitoring system)
are also defined as part of the reactor cavity, up to the first isolation valve.
The core with its graphite stack are enclosed into the reactor cavity that has a free
volume of about 750 m3. All the penetrations through the top (named E) and
bottom (named OR) metal structures (e.g. pressure tubes) are welded and sealed.
To avoid overpressure of the cavity in case of accident a set tubes are located at
the top and bottom of the cavity to drain water and/or steam to the Pressure
Suppression Pool.
Fig. 28 – Longitudinal view of the RC of a RBMK NPP
A He-N mixture (coming from the pressure tube integrity monitoring system)
continuously flows in the reactor cavity (flow ranging from 200 to 400 m3/h at
normal conditions). Maximum gas flow is 900 m3/h.
The gas circuit system, Fig. 29, has the function to extract the gases from the
reactor cavity which are product from the steam-graphite reaction or from leaking
34
pressure tubes and to lead them through different technical devices for their
drainage and purification.
In order to avoid leaks of the gases, all the volumes surrounding reactor cavity (e.g.
between the shroud and the lateral water shield) are filled with pure Nitrogen at
higher pressure (2-5 KPa gauge), part of which flows into the reactor cavity. The
residual flow of about 10-20 m3/h of nitrogen coming from the outer structure is
finally discharged to the high stack after having passed through filters for Iodine,
delay tanks and particulate filters.
Fig. 29 – Gas removal system of the RC of a RBMK NPP
The pressure tube (fuel channel) integrity monitoring system is designed to

continuously monitor temperature and humidity around each fuel channel. In
addition, during plant outages, ultrasonic, visual and geometry measurements
checks are performed on the pressure tubes.
2.2.2. The Accident Localization System

The Accident Localization System (ALS) is designed to avoid environment
contamination as a consequence of a LOCA in the Main Circulation Circuit (MCC).
The ALS is a seal space including the lower part of the down-comers from the
Steam Drum (SD), the inlet and the outlet headers of the Main Coolant Pumps
(MCP) and all the piping system downstream the MCP up to the cavity inlet. The
ALS does not cover some relevant parts of the plant as SD, feed-water system and
steam lines.
35
Another relevant function of the ALS is to make possible to recognize the damage
side of the core by the pressure increase in the different compartments constituting
the ALS. The main compartments of the system are:
• Leak-tight compartment (LTC): two compartments, one for each side of the
core; these compartments include lower down-comers from the SD and the
inlet header to the MCP.
• Pressure suppression pool (PSP): this part of the ALS is subdivided in
three parts connected together; the main function is condensating the
steam discharged from the break in a LOCA
• Steam corridor distribution (SDC): this part receive the steam released into
the URC.
• Under reactor compartments (URC): these compartments include the
piping down stream the MCP up to the cavity inlet.
LTC and SDC are connected with the PSP by discharge tubes in order to
condensate the steam flowing from a break in the MCC covered by the ALS
compartments. Other connections exists between the different compartments;
these connections are constituted by valves and are designed to avoid any
dangerous increase of the pressure inside a single compartment as a
consequence of a large LOCA.
Some systems capable to reduce the pressure inside the ALS are part of the ALS
itself. The main systems are:
• Sprinkler system: this system injects cold water in the pool water and in the
atmosphere of the PSP. The system is composed two identical
subsystems; each system includes pumps taking water from the pool,
coolers and distribution headers injecting in the PSP. This system works in
emergency conditions and in normal conditions to control the pressure and
the temperature in the PSP compartment.
• Surface coolers: they are located in the SDS and are actuated by ECCS
intervention signal.
• Ejector coolers: they are located in the LTC; these coolers are used to
control pressure and temperature in the LTC. The ejectors coolers are also
used in emergency conditions.
• Hydrogen removal system: this system remove the hydrogen from the
atmosphere of the ALS and provide a purification of the air of the ALS. This
system works during normal operation; in emergency condition is isolate to
realize the sealing of the ALS, but it can be put in to operation by the
operator.
The sketch of the ALS of Ignalina NPP is provided in Fig. 30, [10].
36
1. fuel channel 2. main circulation pumps 3. MCP suction header 4. MCP
pressure header 5. group distribution header 6. ECCS headers 7. hot
condensate chamber (HCC) 8. CTCS pumps and heat exchangers 9.
discharge pipes section 10. pipe from the steam relief valves 11. steam
gas mixture from the reactor cavity 12. condensing pools 13. steam
distribution headers 14. bottom steam reception chamber (BSRC) sprays.
15. water seals/S traps between HCC and BSRC 16. BSRC vacuum
breakers 17. air removal corridor sprays 18. air venting channel 19. gas
delay chamber tank 20. gas delay chamber 21. reinforced leaktight
compartments 22. Lower Water Piping compartments 23. steam relief
valves from LWP compartments to RLC 24. top steam reception chamber
25. tip up hatches 26. knock down hatches 27. main safety valve and fast
acting steam discharge valve 28. drum separators 29. Bottom Steam
Reception Chamber 30. Reactor Cavity.
Fig. 30 – Sketch of the Ignalina ALS
2.2.3. The Reactor Building and the Turbine Hall

The reactor building and the turbine hall bound high energy lines and pressurized
components. As such they are possible location of piping break. No pressure
resistant wall are provided. Therefore, steam water releases to the environment are
predicted in case of break occurrence. This does not imply significant radioactivity
release because significant part of possible FP releases can be trapped in these
confinement zones.
37
2.2.4. The fuel loading machine hall
The fuel loading machine hall is sketched in Fig. 1 (i.e. zone No. 9) and Fig. 2 (i.e.,
component No. 14). It has not a direct (or immediate) confinement role. However,
in case of displacement of the top reactor plate, steam water mixture and FP from
the reactor cavity may enter this zone.
In addition, during the (frequent) process of fuel loading and un-loading, accidents
may happen that cause releases of high pressure two-phase mixture and
eventually FP in this zone. Refueling-events were not considered during this PhD
activities.
2.3. The Engineered Safety Features and the Emergency

System
The Engineered Safety Features (ESF) include all components relevant to safety
(therefore ECCS are part of ESF). The ECCS are characterized by suitable quality
in the design and maintenance and are mostly addressed in the section below.
The pipes of ECCS trains, shown in Fig. 31, are 300 mm nominal diameter. Each
of these terminated in ECCS headers, three per core region. From each header a
pipe of 75 mm nominal leads through a 24.5 mm minimal diameter flow limiter and
a check valve to each GDH. In addition a special connecting pipe of 75 mm
nominal diameter is installed between the MCP pressure header and all the ECCS
headers per each core side.
Delivery of the cooling water to the ECCS trains is provided by three subsystems,
each of which is divided into three trains, namely: a) main (or short-term)
subsystems, b) long term subsystems, further divided into the damaged core side
and undamaged core side systems.
Flow from each train of each subsystem is delivered by a single pipe of either 200
or 300 mm nominal diameter. This pipe branches into two, and one of the branches
is connected to each core side pipe of the corresponding train. Cooling water is
delivered to the appropriate core side by an electrically powered fast-acting gate
valve situated in each branch pipe, at inlet to the ECCS train.
38
Fig. 31 – Main (or Short-term) ECCS subsystem, RBMK damaged core side
delivery system
The main delivery subsystem consists of three separate trains, to deliver water to
the trains. Trains 1 and 2 are identical, Fig. 31, each consisting of six pressurized
accumulator tanks containing a total of 80 m3 of water and 70 m3 of nitrogen at 9.8
MPa pressure, per train. Each accumulator tank is fitted with water level indication.
A low water level signal automatically closes the isolation valve downstream the
tank outlets of each train, but in addition, a float type cut-off valve avoids gas
ingress to the trains following water discharge. One tank of each train has a
pressure relief valve. The six tanks of each train deliver water at up to 40oC
through a single 300 mm nominal diameter pipe to the corresponding ECCS train.
Prior to the actuation of the two trains of the main subsystems, all valves in the flow
path downstream the tank outlets leading to the trains are open. Two valves in
each flow path, bypassed by a line fitted with a flow restrictor, are open on
actuation of the accumulator tank trains. These are closed by a signal of 27
seconds after reactor trip, to provide intermediate throttling of the water flow,
matching the reduction in decay heat of the reactor core and extending the total
flow duration to at least 2 minutes.
Train 3 of the main subsystem delivers emergency cooling through a 400 mm

nominal diameter connection from the main feed-water header. After any actuation
of this train in an emergency without loss of offsite power, water at 165 oC is drawn
from the deaerators by the main feed-water pumps for at least 10 minutes before
the deaerators are emptied. Following loss of offsite power (LOOP) the main feed-
water pumps stop. The coast-down time is approximately 55 s. In this period the
pumps are powered by the running down TG.
39
The long-term subsystem is divided in two parts, one for delivery to the damaged
core side and the other to the undamaged core side, Fig. 32 and Fig. 33,
respectively.
The damaged core side delivery is constituted by three trains, each consisting of
two damaged core side cooling pumps (DCSCP) operating in parallel, drawing
suction from the Pressure Suppression Pool (PSP) which has a capacity of 3200
m3. The six pumps of the three trains are each capable to provide 250 m3/h of
water at 7.8 MPa pressure. This system is required to operate within 2 minutes of
the main subsystem actuation signal, before the main subsystem water tanks are
exhausted.
The undamaged core side delivery is constituted by three trains, with one
undamaged core side cooling pump (UCSCP) per train drawing suction through a
common inlet line from the clean condensate tanks. There are three clean
condensate tanks (1000 m3 each) backed up by three makeup tanks, each of 750
m3 in the makeup tanks is available for ECCS use. Other systems can use this
same water source, but overridden by the ECCS actuation signal.
The three pumps of the three trains are identical to those of the damaged core side
system, delivering 250 m3/h at 7.8 MPa pressure. Water to the undamaged core
side is provided by the main circulation pumps in coast down for at least one
minute after any fault considered. The undamaged core side system is therefore
required to initiate flow within one minute of the emergency.
Fig. 32 – Long-term ECCS subsystem, RBMK damaged core side delivery system
40
Fig. 33 – Long-term ECCS subsystem, RBMK un-damaged core side delivery
system
In case of loss of offsite power, each train of power supplies is provided by a

separate diesel generator, with power available 25 s after diesel start to the
DCSCP and 35 s after diesel start to the UCSCP. Each pump can reach its rated
flow rate within 1.0 to 1.5 s from the signal.
The special connection between the pressure header and the three ECCS headers
per each core side is installed to provide a short term cooling to the affected GDH
in case of MCC break between pressure header and GDH upstream the check
valve at GDH inlet. In this emergency situation the signal for ECCS activation
comes from differential pressure between pressure header and drum separator.
41
2.4. The reference plant and the reference conditions:
Smolensk-3, 16 October 1996
2.4.1. Reference fuel load map
The map of the fuel load of the Smolensk NPP Unit-3 referring to the core status at
October 16 1996 is represented in the Fig. 34 [11].
Fig. 34 – The reference map of fuel load of the Unit-3 of Smolensk NPP (16.10.96)
In Tab. 8 is reported the legend for the Fig. 34, listing the different type of channels
present in the core, and their number.
42
Tab. 8 – Characteristics of FC and control rods
ID on Fig. 34 Description Quantity

1 Fuel rod bundle with enrichment of 2% U235 48
2 Fuel rod bundle with enrichment of 2.4% U235 1552
3 High-altitude power gauge 12
4 Control rod (design 2091) 155
6 Shortened control rod (design 2093) 32
7 Fast Acting Safety System Rod (design 2505) 24
8 Rod with additional absorber (design 1814) 90
12 Water column in the channel 1
13 Graphite reflector 604
As it can be seen from the table, 211 control rods of various design and purposes
are present in the core. The cartogram of the different type of control rods, with
their arrangement in the core is reported in Fig. 35. A detailed geometrical
description of the different control rods type can be found in the section 2.1.2 of
this document.
Fig. 35 – Control Rods Cartogram
43
In Tab. 9 is reported the initial condition of the tip of the control rods. It has to be
noted that the numeration of the control rods was done according to their
arrangement in the core reported in the cartogram of Fig. 35 and according to the
rule “from left to right, from top to bottom”. The insertion depth is measured in
centimeters (cm) and it reports the distance between the tip of the control rod and
the upper part of the top reflector. It has to be remarked that the insertion depth of
the shortened control rods is negative because they move from the bottom to the
top of the reactor core.
In Fig. 36, it is reported the map of the coolant mass flow in each of the core
channels, except for the radial reflector channels.
It is important to remark that in all of these kinds of maps here reported the colors
have always the following meaning:
• yellow is indicating a MCR channel;
• green a SR channel;
• orange a SHR channel;
• blue an Additional Absorber channel;
• grey an Axial Detector channel;
• dark blue the Water Column.
All the other colors used in such a kind of maps are indicating Radial Reflector
Channels or FC grouping.
Tab. 9 – CR Standard Position
Insert. Insert. Insert. Insert.

CR CR CR CR CR CR CR CR
depth, depth, depth, depth,
# Type # Type # Type # Type
cm (*) cm (*) cm (*) cm (*)
1 30 MCR 55 30 SR 109 30 SR 163 20 MCR
2 220 MCR 56 30 MCR 110 30 MCR 164 -270 SHR
3 60 MCR 57 30 SR 111 610 MCR 165 30 MCR
4 30 MCR 58 30 MCR 112 30 MCR 166 -290 SHR
5 -240 SHR 59 30 MCR 113 30 MCR 167 30 MCR
6 20 MCR 60 670 MCR 114 470 MCR 168 -270 SHR
7 -370 SHR 61 30 MCR 115 30 MCR 169 30 SR
8 30 SR 62 30 MCR 116 20 MCR 170 -280 SHR
9 -370 SHR 63 30 MCR 117 520 MCR 171 30 MCR
10 150 MCR 64 20 MCR 118 180 MCR 172 -310 SHR
11 -370 SHR 65 20 MCR 119 30 MCR 173 220 MCR
12 30 MCR 66 30 MCR 120 200 MCR 174 -260 SHR
13 30 SR 67 20 MCR 121 30 MCR 175 30 MCR
14 680 MCR 68 170 MCR 122 -310 SHR 176 30 MCR
15 170 MCR 69 20 MCR 123 30 MCR 177 20 MCR
16 30 MCR 70 30 MCR 124 -270 SHR 178 150 MCR
17 40 MCR 71 30 MCR 125 30 MCR 179 590 MCR
18 30 SR 72 520 MCR 126 -330 SHR 180 110 MCR
19 250 MCR 73 30 MCR 127 30 SR 181 340 MCR
20 20 MCR 74 30 MCR 128 -290 SHR 182 460 MCR
44
Tab. 9 (cont.) – CR Standard Position
Insert. Insert. Insert. CR Insert. CR

CR CR
CR depth, CR depth, CR depth, Type CR depth, Type
Type Type
# cm (*) # cm (*) # cm (*) # cm (*)
21 20 MCR 75 90 MCR 129 30 MCR 183 360 MCR
22 30 MCR 76 210 MCR 130 -290 SHR 184 280 MCR
23 30 MCR 77 30 MCR 131 210 MCR 185 30 MCR
24 30 MCR 78 -370 SHR 132 -270 SHR 186 30 MCR
25 30 MCR 79 680 MCR 133 30 MCR 187 30 MCR
26 30 MCR 80 -340 SHR 134 80 MCR 188 30 MCR
27 30 MCR 81 120 MCR 135 260 MCR 189 30 MCR
28 210 MCR 82 -290 SHR 136 20 MCR 190 30 MCR
29 600 MCR 83 30 SR 137 30 MCR 191 30 MCR
30 110 MCR 84 -290 SHR 138 20 MCR 192 30 MCR
31 20 MCR 85 30 MCR 139 550 MCR 193 30 SR
32 30 MCR 86 -270 SHR 140 30 MCR 194 640 MCR
33 140 MCR 87 30 MCR 141 20 MCR 195 80 MCR
34 680 MCR 88 -310 SHR 142 100 MCR 196 550 MCR
35 30 MCR 89 130 MCR 143 30 MCR 197 30 MCR
36 20 MCR 90 170 MCR 144 20 MCR 198 30 SR
37 -270 SHR 91 30 MCR 145 30 MCR 199 220 MCR
38 680 MCR 92 30 MCR 146 650 MCR 200 -250 SHR
39 -270 SHR 93 30 MCR 147 60 MCR 201 30 MCR
40 30 MCR 94 30 MCR 148 30 MCR 202 -300 SHR
41 -270 SHR 95 20 MCR 149 30 MCR 203 30 SR
42 30 SR 96 80 MCR 150 80 MCR 204 -310 SHR
43 -290 SHR 97 30 MCR 151 30 MCR 205 60 MCR
44 20 MCR 98 100 MCR 152 660 MCR 206 -160 SHR
45 -280 SHR 99 30 MCR 153 30 SR 207 30 MCR
46 660 MCR 100 700 MCR 154 20 MCR 208 30 MCR
47 -230 SHR 101 30 SR 155 30 SR 209 20 MCR
48 30 SR 102 150 MCR 156 680 MCR 210 110 MCR
49 30 MCR 103 30 SR 157 30 MCR 211 30 MCR
50 30 SR 104 30 SR 158 20 MCR
51 30 MCR 105 150 MCR 159 30 MCR
52 670 MCR 106 30 SR 160 30 SR
53 30 MCR 107 30 SR 161 30 MCR
54 300 MCR 108 490 MCR 162 30 SR
(*) Insertion Depth for MCR and SR is measured from the Upper Edge of the Top
Reflector. Insertion Depth for SHR is measured from Bottom of Reactor Core
(negative values because SHR are bottom to top inserted)
45
Fig. 36 – Core Channels Flow Map (Kg/s)
46
3. THE BACKGROUND FOR RBMK ACCIDENT ANALYSIS
One of the bases for the safety and the safety evaluations in NPP is constituted by
the knowledge of phenomena that are envisaged following hypothetical transients
and accidents. The present chapters introduces the safety of the RBMK starting
from the expected phenomena.
Introductory general remarks are provided in section 3.1. The safety needs focused
on the ‘binding’ acceptance criteria and the reasons for them are discussed in
section 3.2. A more comprehensive description of expected physical phenomena
and scenarios with connected relevant parameters is given in section 3.3.
3.1. The technological status for RBMK safety

3.1.1. The background
The safety of a nuclear power plant (NPP), understood as its capability to keep the
radiation exposure of personnel and population within specified limits, is ensured
by maintaining the integrity of safety barriers, which are part of the plant defense in
depth concept. A series of barriers prevents the release of radioactive fission
products from their source beyond the reactor containment and into the
environment. In analyzing the NPP safety, it is essential to assess the integrity of
these barriers and to decide to what degree the response of the whole NPP and its
systems to a certain initiating event is acceptable from the viewpoint of the plant
safety. For the sake of clarity, the integrity of the safety barriers is related to certain
threshold values, which are referred to as acceptance criteria. Essentially these are
the design limits for design basis accidents (DBA), adopted with a conservative
margin so that the safety barrier integrity is guaranteed as long as the parameters
do not exceed the relevant criteria.
Safety analysis for an RBMK NPP should assess the integrity of the following
barriers in the path of radioactivity transport and release:
• Fuel matrix.
• Fuel cladding.
• Circulation circuit pressure boundary and, in particular, the components
most susceptible to damage, namely fuel channel (pressure) tubes.
• Metal structures forming the reactor cavity.
• Structural components of the leaktight accident localization system (ALS)
compartments and other compartments of the NPP housing circulation
circuit pipelines.
Should any safety barrier fail, thus opening the pathway for the release of
radioactivity beyond the plant boundaries, the amount of radioactivity and the
population exposure should be assessed. Beyond Design Basis Accidents (BDBA)
are analyzed for the following purposes:
47
• To assess the degree of reactor protection and the time available for taking
countermeasures.
• To determine the emergency and other signals available to the operator for
identifying the plant status and to devise appropriate accident management
steps.
• To develop a package of organizational and technical measures
(management strategy) for prevention and mitigation of the accident
consequences.
• To assess the possible consequences as input information for planning
protection of the population and personnel.
Related to BDBA analyses, according to the requirements of the Russian nuclear

regulatory authority RosTechnadzor, both deterministic and probabilistic
approaches should be used. Namely, according to requirements OPB-88/97:
• The estimated probability of an event with large release should be less

than 10-7 within one reactor year.
• The estimated probability of severe reactor core degradation or melting
should be less than 10-5 within one reactor year.
However, if some initiating events can lead to severe consequences and the inside
features of reactor are not able to prevent this, accident mitigation means should
be foreseen without drawing the attention to the probability of these events.
The deterministic approach for analysis BDBA is therefore also very important: this
should be based on the method of postulating accidental conditions. The method is
based on determination of connections between the plant conditions, level of
severity of accident consequences, i.e. how many physical barriers are violated:
fuel assembly including fuel pellet and fuel cladding, pressure tube, reactor cavity,
Main Circulation Circuit (MCC), Accident Localization System (ALS) and availability
of critical safety functions. According to the above requirements (OPB-88/97), a list
of BDBA scenarios for further detailed investigations should be developed. The
analysis results will be the basis for the development of the accident management
program for RBMK reactors.
3.1.2. The relevant aspects

Because the RBMK comprises a pressure tube design with a number of unique
features, the results of much Western work, mainly done for LWR on severe
accidents, may not be directly applicable. Where work has been done on channel
type reactors, there are always differences that make direct comparisons difficult.
The main system characteristics can be summarized as follows (some of these
also constitute differences between RBMK and other LWR systems):
1. The channel design; each fuel assembly is contained in a separate

thermal-hydraulic environment.
2. The moderator is graphite; this has implications for structural performance
and heat transfer. The graphite provides a very large heat sink and source
and, as a result, slow core heating in case of hypothetical accidents
48
involving the loss of all heat removal systems and additional heat supply to
pressure tubes following core cooling.
3. RBMK reactors have a large core, where local effects may have a
significant role in safety analyses of the several initiating events (e.g.
blockage of one hydraulic channel).
4. The MCC is divided into two symmetric loops; a pipe break in one of these
directly affects the thermal-hydraulic and neutron kinetics behavior on one
side of the core. The resulting asymmetry in reactivity and power has to be
assessed by three-dimensional codes.
5. The accident localization system design has important implications for
accident management and fission product retention under accident
conditions involving radioactivity releases from the fuel.
6. Redundancy of safety system components gives very low probability of
severe accidents.
Evaluating the unique features of RBMK reactors specific phenomena should be

reflected in the modeling of transients. These phenomena are described below.
• The spatial feedback between the neutron intensity, the fuel temperature,
the graphite temperature, and the coolant void fraction (coolant density) in
the core and in separate fuel channels shall be considered. To fulfill this
condition a 3D-space dynamic neutron kinetics code shall be coupled with
a thermal-hydraulics code and properly qualified. The coupled code must
also feature high computational speed and stability.
• The specific thermal-hydraulic processes in the primary circuit and the
reactor core. At first, the thermal-hydraulic processes in the reactor
channels (pressure drop, relative motion of phases, counter-current flow,
reflood) should be considered. The abstract ‘reactor channels’ implies the
fuel channels, channels in which the reactor control rods are placed and
the instrumentation channels. The reactor core model should include the
following types of processes (some of these are typical processes for a
boiling channel in a BWR):
Heat transfer prior to the onset of critical heat flux (CHF). Heat transfer
under steady state conditions and transients.
Critical heat flux. Sharp decrease of heat transfer from fuel elements to
coolant (high L/D value typical of the RBMK channel, i.e. much higher
than in the case of BWR).
Post-CHF heat transfer (heat transfer to coolant after the CHF onset,
same note as above applies).
Radiation heat transfer. Radiant heat transfer through layer of the
saturated steam–water mixture or superheated steam between fuel
elements and fuel channel (FC) wall (the relevance of this process
upon the overall scenario should be checked, at least by sensitivity
studies).
The heat transfer in radial direction between the pressure tubes and
the graphite stack through the gaps and bushes and the heat transfer
between neighboring graphite columns should be considered.
49
• The specific natural circulation in the MCC that takes place after the main
circulation pump (MCP) shut-off and coast-down should be considered,
namely:
During transients, after reactor shutdown, natural circulation of the
coolant provides adequate cooling of the reactor.
During loss of coolant accidents (LOCA) breakdown of natural
circulation due to flashing in the down- comer piping may occur.
Analogously, during accidents involving the opening of the main safety
valve, breakdown of natural circulation might occur due to failure of the
valves to close.
• The specific phenomena in drum separators of RBMK should be
considered:
The boiling and condensation in the drum separator (DS). The
structure and dynamics of the coolant flow in the DS varies with
changes of pressure and feed-water supply.
Separation and water entrainment from DS into steam lines. The
phenomenon is typical for regimes with considerable level increase
above submerged perforated plate.
Steam entrainment into down-comers. The phenomenon is typical for
regimes with considerable decrease of total collapsed level and
pressure in DS.
• During the analysis of LOCA-type accidents the modeling of the coolant
discharge from the circulation loop break is important. Sub-cooled liquid,
two-phase mixture and steam flows are involved. Modeling of possible
internal critical sections (e.g. at the abrupt geometric discontinuities of the
long pipelines connected with the FC) is also important.
• Occurrence of thermal-hydraulic instabilities following LOCA, connected
with the (large) number of parallel channels and the (high) value of the L/D
for individual channels, should be investigated.
• The flow instability in parallel steam generating channels is relevant in
RBMK. This can take place both at forced and at natural circulation
conditions. Under such oscillations the cladding temperatures can reach
unsafe values.
• During the analysis of LOCA-type accidents the coolant through the break
is discharged into the compartments of the Accident Localization System
(ALS), unless the accident is constituted by the break of the FC (see
below). In the former case, other than the modeling of pressure and
temperature behavior inside the Reactor Cavity (RC) and the ALS, the
generation-transport-accumulation-distribution of H2, fission products and
products of the physical and chemical reaction between graphite and
primary system coolant shall be considered.
• In case of fuel channel rupture, the steam–water mixture is discharged into
the reactor cavity. The calculation of the venting capacity of a damaged
core, including the local phenomena which may have an impact on the
temperature and pressure transient in the reactor core cavity (mixing of
He-N2 mixture with flashing steam, evaporation of discharged water, heat
transfer from the graphite, liquid carry-over through the break, etc) shall be
performed.
50
Overheating of pressure tube (PT) in the case of “fuel assembly power – coolant
flow rate mismatch” at high pressure in the fuel channels is the most probable
reason for fuel channel failure. Three known single FC failures in RBMK reactors
(Leningrad NPP [LNPP] in 1975, Chernobyl NPP [ChNPP] in 1982, LNPP in 1992)
occurred as a consequence of the pressure tube overheating mentioned. The
assessment of PT rupture due to accidental overheating is the key element of
RBMK safety analysis.
Combined calculations of fuel assembly, pressure tube, and graphite stack

behavior under accident conditions with channel voiding show that, up to the
moment of pressure tube rupture, the fuel rod cladding may reach a very high
temperature, in some cases approaching the melting temperature (of the cladding).
In such conditions fuel clad oxidation (H2 production), fuel melt and rod
deformation can be expected. The PT rupture is possible owing to the combined
effect of high temperature and pressure. Following PT rupture, hydrodynamic
forces acting on the fuel assembly during the break and caused by the escaping
steam–water mixture into the reactor cavity may lead to further fuel rod failure and
to failure of the fuel assembly.
As an example of such a situation, the accident at unit 3 of LNPP in 1992 may be

considered. The decrease of coolant flow through the individual control valve at
nominal power happened and caused overheating of the channel, its rupture and
graphite stack destruction. The accident investigation showed that the pressure
tube rupture took place at the channel power of 1.94 MW, pressure about 7.0 MPa
when temperature of the channel reached 650°C. Examination of the central
stringer of the fuel assembly gave the evidence that the stringer material was
heated up to 1200°C.
In the case of heating up of a FC thermal mechanical behavior of fuel assembly

should be considered to wit elongation of fuel rods and hanger, inter action of fuel
rods having pressed out claddings with stringers and hanger, possible bowing of
assembly due to azimuthal temperature irregularity, steam-zirconium reaction,
possible formation of low temperature eutectics.
On the basis of the results obtained in the experimental study of deformation and
rupture of regular (i.e. full scale) PT and scaled PT mock-ups, as well as from
inspection results after the incidents involving PT rupture at ChNPP and LNPP, it
has been ascertained that the process of PT deformation from the start of accident
heating to the moment of rupture can be subdivided into three stages:
Stage one: regular axial-symmetric ballooning of the tube up to the moment of

closing all the gaps in the system PT – slit graphite rings – graphite blocks.
Stage two: the pressure load of the FC is partly released to the graphite rings and
graphite blocks. Joint deformation of the tube, rings and blocks and tube material
causes “pouring” into the slits of graphite rings. As a result, one or several radial
cracks appear on the internal surface of the blocks under the influence of the
ballooned PT. Their propagation results in graphite blocks destruction.
51
Stage three: the deformed PT comes in contact with fragments of the graphite rings
and blocks and under the conditions of high temperature (caused by coolant
overheating and possible contact between clad and FC walls) and high pressure
(the system pressure is, typically, not affected by an event occurring in a single FC
or limited to a small number of FC and keeps its nominal value) swells up to the
tube rupture. Hydraulic loads cause fuel fragmentation and transport across the PT
rupture. Under the influence of differential pressure loads the gap space between
adjacent graphite stacks tends to increase.
When deriving the model of stress-strain state of a pressure tube, it is assumed

that the deformation process at the first and third stages takes place with keeping
of the axial symmetry. The deformation at the second stage is assumed to be close
to zero, which is justified by the low plasticity of graphite blocks. The irradiation
level of the graphite has a role at this stage.
Possible deformation of the pressure tubes in the RBMK core will occur in
interaction with the graphite stacks. Analysis of tube ruptures taking into account
the constraints provided by the graphite block and results of inspections after the
channel rupture at LNPP and ChNPP suggest that under full pressure the graphite
blocks cannot withstand the forces caused by PT ballooning and hence break.
For a pressure exceeding 4.0 MPa the failure criteria for tubes with graphite do not
differ from the general pattern of data obtained under conditions of free-standing
tubes. Therefore, the failure criteria of PT obtained largely in the experiments
without graphite can be used for the evaluation of PT failure in case of accidental
over heating under pressures exceeding 4.0 MPa. Under FC-pressures lower than
4.0 MPa the graphite blocks can prevent the PT rupture.
When the pressure tube ruptures, coolant discharges into graphite stack where it
flashes and residual water evaporates on hot surfaces. The relationship between
steam sources and sinks determines the reactor cavity (RC) pressure history in this
case. Steam sources are steam–water mixture discharge from the ruptured FC and
liquid phase evaporation on hot surfaces of graphite and metal in the reactor
cavity. Steam sink is provided first of all by steam dumping via the reactor cavity
venting system (RCVS) pipelines to the ALS and to the environment (RC-SRV).
Some steam may condense on cold surfaces of pipes of the RCVS.
Large coolant discharge into the stack may cause essential displacement of core
components under the influence of the forming pressure field, which produces
complex deformations with gaps growth in some areas (e.g. around the broken PT)
and reduction in others (neighboring FC). The following phenomena have to be
analyzed for the pressure tube rupture accident:
• Flow regime and heat exchange in non-equilibrium two-phase flow
conditions in the circulation loop (MCC), inside the graphite stacks and
outside the stacks, before and after the PT rupture.
• Deformation and rupture of pressure tubes and fracture of graphite blocks
due to overheating.
• Oxidation of fuel pins with H2 production and possible melting of the UO2.
52
• Deformation and destruction of fuel pins and fuel assembly owing to
hydraulic and thermal loads.
• Release of fuel gap radioactive gases and fuel particles into the coolant
(FP source term into the FC): FP transport and physical and chemical
interactions between steam–water mixture and graphite.
• Displacement and deformation of graphite stack elements as a result of
loads caused by PT rupture.
• PT zirconium and graphite oxidation leading to generation of hydrogen and
hydrocarbons.
• Operation of the reactor cavity venting system and accident localization
system.
• Transport of radioactive fission products from the FC to the RC (or gaps
among the graphite stacks) and finally to the ALS.
The estimation of the possible evolution of a single FC failure (i.e. PT rupture)

toward the Multiple Pressure Tube Rupture (MPTR) should be performed. This
implies the evaluation of the structural-mechanics resistance of parallel stacks of
graphite bricks and PT, following bending loads orthogonal to the PT axis, as a
function of the PT position inside the core. The resulting ‘resistance’ loads shall be
compared with differential pressure loads caused by the release (through the PT
rupture) of MCC fluid and consequent thermal-hydraulic interactions with graphite.
3.2. The safety needs

Activity in the field of RBMK safety research including the identification of safety
limits covers the entire history of RBMK design and operation. Obviously, one can
split this activity into two parts: before 1986 and the post-Chernobyl era.
Parameters of steady state operation under nominal conditions were the main
purposes of the work during the first twenty years of RBMK technology (i.e. in the
pre-Chernobyl era). The steady state experiments concerning thermal-hydraulic
problems of the core (FC) were accomplished mainly by the Kurchatov Institute on
its KS facility (full-scale electrically heated 18-pin fuel channel model). Correlations
for heat transfer in the bundle, pressure drop, CHF, post dry-out modes, critical
discharge data, etc., were obtained at that time. Some aspects of these problems
were then investigated at the reactor vendor organization (NIKIET) on its BM
facility (seven-pin electrically heated model), including pressure drop and void
fraction distribution. In the same period, phenomena of a dynamic nature were also
addressed: experiments were performed concerning accident conditions, including
ECCS injection and FC quenching and multiple channel stability (one- and six-
channel loops of the facility #108 at EREC).
After the Chernobyl accident efforts were devoted to the investigation of the
dynamic nature of RBMK: improved 3-D codes for neutron kinetics were
developed, new facilities were constructed and instrumentation on existing ones
was essentially enhanced to provide good conditions for fine dynamic experiments.
Data were obtained related to the following problems: steam binding and counter-
current flow limit (KS semi-scale facility), flow blockage in the channel and
53
overheating (KSB facility, RRC KI), low-pressure steam condensation in the outlet
feeding pipe and FC (‘PWK tube’ facility at NIKIET) and some others. Interesting
and informative experiments were accomplished at RBMK NPP sites. New facilities
(e.g. PSB RBMK) have been constructed at EREC.
Thus the methods for accident analysis as well as the basis for acceptance criteria
have been considerably improved during the last two decades due to a better
insight into the physical phenomena through experimental research, enhancement
of the computer codes and computational capabilities. These improvements have
made it possible to switch from simplified codes to more sophisticated and
mechanistic integral system codes. The ongoing improvements in computer
capabilities have removed the main constraint of computational tools. In this
context, the use of advanced coupled 3-D neutron kinetics and thermal-hydraulic
codes is of utmost importance.
The accident analysis should demonstrate that parameters (acceptance criteria)

defined by the designer and the regulatory authority must not be violated during
accident events. Also the evaluation of the adequacy of the safety system design
involved in the occurrence shall be based on the comparison of these limits with
the calculation results.
The calculations can be provided using ‘conservative’ or ‘best estimate’

approaches. Now the trend in accident analysis has continued to move to best
estimate approach. If this approach is used, the code and model uncertainty should
be evaluated. Code predictions are uncertain due to a number of uncertainty
sources, e.g. code models, initial and boundary conditions, plant state, scaling and
numerical solutions algorithm. As it is shown in Fig. 37, [12], usually the best
estimate approach gives the lower calculated values in comparison to conservative
approach. The conservative approach does not give any indication of the actual
margins between the actual plant response and the conservatively estimated
response.
By contrast, the uncertainty estimate provided in the best estimate approach is a

direct measure of such margins. As a result, the best estimate approach may allow
for the elimination of unnecessary conservatism in the analysis and may allow the
regulatory body and plant operating organization to establish a more consistent
balance for a wide range of acceptance criteria.
54
Safety Limit (design limits for accident)
Acceptance Criterion (Regulatory Requirement)
Margin to Acceptance
Criterion
Calculated conservative
value
Actual Safety
Margin Upper Limit of Calculated
Uncertainty Range
Calculated Uncertainty Range
Real Value
Fig. 37 – Best estimate and conservative safety analysis: definitions of acceptance

criteria and safety margins, [12].
However for some cases, conservative modeling approaches could be used (to a
large extent in analysis of DBA, for example) simply to avoid the cost of developing
a more realistic model, even though conservative models are not specified by
regulation. In this case, the conservative models are selected and evaluated on a
case by case basis.
For both cases, ‘best estimate’ or ‘conservative’, it is necessary to distinguish two

groups of limits: the safety limit and the acceptance criteria. The safety limit is a
critical value of an assigned parameter associated with the failure of a system or a
component. The acceptance criterion is the quantitative limitation of selected
parameter or qualitative requirement set-up for the results of accident analysis. The
specific safety limits and acceptance criteria for RBMK plants are presented below,
together with selected (not-systematic) recommendation for evaluating safety
margins corresponding to those criteria.
3.2.1. The RBMK acceptance criteria

The IAEA guidance at [13] deals with the RBMK acceptance criteria and was
issued when the present report was completed. The content of this section reflects
and summarizes the content of the mentioned IAEA guidance.
55
3.2.1.1. Fuel clad integrity
For the integrity of fuel claddings to be confirmed it is essential to ensure that the
following maximum values of fuel rod parameters are not exceeded:
• Pellet volume – average fuel enthalpy of 710 kJ/Kg.

• Cladding temperature of 700 °C.
If the above mentioned acceptance criteria are exceeded, the safety limit on fuel
melting temperature should be checked:
• Fuel temperature of 2800 °C.
Additional analysis for design basis accidents has to demonstrate that the cladding
temperature, does not exceed the 1200 °C safety limit and the local depth of fuel
cladding oxidation should not exceed 18% of the original thickness.
The additional analysis should also confirm the compliance with the requirement
that the mass of zirconium cladding that reacted with steam should not exceed 1%
of the total mass of fuel claddings in the core. This sets a limit to the release of
hydrogen into the ALS.
3.2.1.2. Fuel channel integrity

It was agreed that for normal operation and with the re-tubing approach adopted in
Russia (complete reactor re-tubing at a certain power production level, due to
uncertainties in material properties and fabrication tolerances when local gas gap
closure occurs), no fuel channel degradation mechanisms have been identified that
could influence the fuel channel integrity.
The computational assessment of the pressure tube integrity in thermal-hydraulic

and thermal-mechanical codes uses various experimental high temperature failure
criteria: rupture temperature versus channel pressure (temperature criterion),
rupture strain versus tube temperature (strain criterion), rupture stress versus tube
temperature (stress criterion) and rupture strain power versus tube temperature
(energy criterion).
The temperature failure criterion in the form of the tube rupture temperature Twr
dependence on the channel pressure is given in Fig. 38. The data are classified by
the heating rates, the experiments with graphite blocks are marked with black
symbols. The shaded region covers all combinations of Tw and p parameters in
case of rupture. At higher heating rates the values of rupture temperature (dark-
gray region) proved to be higher than the values obtained for low rates (light-gray
region).
An energy criterion may be adopted for the thermal-mechanical code employed for
calculating the deformation and for assessing the pressure tube integrity under
accident conditions as shown in Fig. 39. This criterion is a specific rupture strain
56
power ji (W/Kg) which is determined through the stress intensity si, the material
density ρw and the strain rate intensity ζi:
σ iζ i
ji =
ρw
1300
Heat-up rate:
1) -up to 1 oC/s
1200 2) -above 1 oC/s
-tube with graphite
Tube rupture temperature (oC)
1100
2
1000
1
900
AECL data
800
700
600
0 1 2 3 4 5 6 7 8 9
Channel pressure (MPa)
Fig. 38 – Pressure tube rupture temperature versus internal pressure

2500
Heat-up rate:
-up to 1o C/s
-above 1o C/s
Rupture strain specific power (W/kg)
-tube with graphite

2000
1500
1
1000
500
600 700 800 900 1000 1100

Tube temperature (oC)
Fig. 39 – Specific rupture strain power as a function of tube temperature.
57
Simplified acceptance criteria can be used. If the following criteria are not
exceeded the integrity of the pressure tubes will be maintained (see Fig. 38):
• Maximum fuel cladding temperature is less than 700°С.

• Maximum fuel channel pressure tube temperature less than 650 С.
Multiple pressure tube rupture sequence is a BDBA. In this case, the potential
hazards are loss of the reactor cavity integrity and damage of metal structures of
the reactor. To define the scope of MPTR beyond which the threat of reactor cavity
destruction lies, it is necessary to perform analysis of the venting capacity of the
system for reactor cavity protection against overpressure. The results of this
analysis are useful in assessing the consequences of the beyond design basis
accident leading to MPTR.
Events leading to multiple pressure tube rupture have received considerable

attention in safety evaluation of RBMK, since they might develop into extended FP
releases. Studies performed so far by Russian specialists indicate that the
probability of such events is very low.
The partial break of a group distribution header was identified as a potential

precursor for the MPTR. The present Project aims, among the other things, at
establishing the realism of the MPTR as a consequence of identified precursor
events, the reference one being the single channel blockage and the consequent
(single) PT rupture.
3.2.1.3. MCC integrity

Various MCC sections are capable of withstanding different maximum (or design)
pressures.
The MCC section between the gate valves of the MCP inlet pipes and the gate
valves at the group distribution header (GDH) inlets, which can be shut off by
isolating valves, can tolerate the greatest pressure. The permissible hydraulic test
pressure at this section is 13.4 MPa.
Fuel channels are also tested under 13.4 MPa. The hydraulic test pressure
adopted for the remaining MCC components, which is determined primarily by the
strength of drum separators and steam lines, ranges from 10.1 to 10.4 MPa.
Since the MCC operates as a single system, the last of these values should be
taken as an acceptance criterion.
3.2.1.4. RC integrity
The reactor cavity is formed by three metal structures: the top plate, the bottom
plate and the barrel with a thermal expansion compensator. The barrel is
hermetically welded to the top and bottom plates. The excessive pressure of 300
KPa in the reactor cavity may be regarded as a less conservative, ‘realistic’
criterion of reactor cavity failure. But more conservative value of 214 KPa is
58
adopted in analyses of the design basis accidents (in the case of Smolensk 3 NPP
it was found that RC-SRV open at 277 KPa, [14]).
It was repeatedly shown in previous RBMK safety analyses that none of the design
basis accidents would lead to MPTR, as already mentioned. Should analysis of any
beyond design basis accident encounter a sequence of events with the ensuing
rupture of more than one fuel channel, and if the MPTR extent can be determined,
in such a case by analyzing the maximum capacity of the reactor cavity venting
system (RCVS) it is possible to assess the reactor cavity integrity, i.e. the likely
structural and mechanical consequences of such BDBA.
The methodology for assessing the venting capacity of the system involves
modeling of the system thermal-hydraulics for an MPTR with various boundary
conditions. The parameters conservative from the point of view of the maximum
venting capacity should be selected based on phenomenological analysis. The
ranges of possible variations of such parameters should be validated.
Consideration should be given to the parameters affecting the flow rate of steam
entering the stack and to the steam generation in case of pressure tube ruptures
as well as to the parameters that influence the rate of steam venting from the
reactor cavity under various boundary conditions.
Phenomenological analysis of the effect of various parameters on the maximum

venting capacity allows classifying them into key parameters and those of
secondary importance. Variation ranges are determined for the former, while the
latter have the ‘worst’ values set for them, i.e. the values detracting from the
maximum capacity of the RCVS.
The maximum venting capacity should be defined in terms of the number of broken
pressure tubes with the resultant pressure in the reactor cavity, considering the
venting capacity through the RC-SRV. The parametric results should demonstrate
compliance with the acceptance criteria. Analysis of the maximum venting capacity
should determine the dependence of the number of broken tubes on such a crucial
parameter considering the RC-SRV.
3.2.1.5. ALS integrity

Accidental release upon a break in a circulation circuit pipe will be confined in a
system of leaktight compartments equipped with devices for emergency steam
condensation (pressure suppression system). However, the ALS does not cover all
the circulation circuit pipes. Upper regions such as steam–water lines, the top part
of the down-comers, equalizing pipes of the drum separators, and all the steam
lines are found outside the ALS, in compartments designed to general building
standards and rules. This means that these compartments afford no leak-tightness
and are not nearly as strong as most of the ALS rooms. Analysis of the integrity of
the ALS and other compartments of a power plant is an essential requirement of
the safety analysis of NPP for all design basis accidents and beyond design basis
accidents.
59
The maximum permissible pressures in compartments for MCC pipes and
components and for ALS are reported in Tab. 10, [13]. Among the RBMK plants of
the second generation, two different ALS configuration can be identified. Namely:
- At Leningrad 3 & 4 and Ignalina 1 &2, the system for condensation of
accidental steam is housed in accident localization towers (ALT).
- At the other power plants, the system is found at lower elevations of the
main building. In the case of a coolant leak, pressure relief in the ALS
relies on passive condensing devices, i.e. pipes submerged under water
on two decks of the PSP.
Leaktight compartments of Smolensk 1 and 2 and Kursk 3 and 4 have safety

valves with an opening pressure difference of 270 KPa. Opening them allows the
steam and gas mixture to be vented from the ALS to the atmosphere. The ALSs at
Leningrad 3 and 4 and Ignalina 1 and 2 operate in this different manner: in the
case of MCC pipe breaks inside leaktight compartments or in DGH-LWL
compartments, the steam and gas mixture will be vented into the ALT via a steam
discharge passage. With pipe breaks in the upper part of the circuit (DS
compartment or the space above the reactor), excess steam will be vented directly
to the atmosphere.
Tab. 10 – RBMK generations.
RBMK reactors Generation

Leningrad 1 & 2
Chernobyl 1+ & 2+ I
Kursk 1 & 2
Leningrad 3 & 4
Ignalina 1+ & 2 II A
Chernobyl 3+ & 4+
Smolensk 1 & 2
II B
Kursk 3 & 4
Smolensk 3
III
Kursk 5(*)
( + ) permanently shut-down; (*) under construction
60
Tab. 11 – ALS evolution in the various RBMK generations (see Tab. 10) and
acceptance criteria
Permissible (excess) pressure (kPa)

Compartment with MCC
Generation Generation II Generation
components
I A B III
MCP pipes, suction and PH 40
GDH, LWLs 40.0 80 80 80
SWL, SD, DC & SL 25
DC 40
Leaktight compartments 300 440 (270) 440 (270)
Steam distribution passage 80 (300)* 440 440
Air space of PSP 440 440
Air space of the enclosure 440
Central Hall 2 (5)* 5 5.0
SD and space above reactor 25 25 25.0
Accident Localization Tower 80
(*) Number in brackets refer to Ignalina 1 & 2
The ALSs of these units have two major distinctions from their counterparts at
plants of the second generation: they have no enclosure to receive steam and gas
from the reactor cavity and their PSP has only one elevation.
However, the ALS may loose its function of a safety barrier at a lower excess
pressure due to opening of safety valves in the leaktight compartments, whereupon
a radioactive release may occur at upper elevations of the reactor building.
The leaktight compartments and the air space of the PSP may suffer overpressure
during LOCA with failure of the PSP cooling system (loss of ultimate heat sink),
when steam condensation in the pool water is less effective at water temperatures
exceeding about 85° C.
Therefore, either the maximum permissible water temperature (for example < 85°
C) or the operating pressure difference of the safety valves may be adopted as an
acceptance criterion for the PSP.
Hydrogen ignition is also a threat to the ALS integrity. The maximum permissible
hydrogen concentration in any single ALS compartment is taken equal to 4% by
volume in the analysis. Should this criterion value be reached, Н2 flammability must
be comprehensively assessed, taking into consideration the time dependence and
the characteristic boundary conditions of the accident scenario for the ALS.
3.2.1.6. The permissible radiation doses

According to the rules laid down by the national nuclear regulatory authority, the
consequences of a DBA should never result in any population exposure that would
61
require any countermeasures to protect the people in the early period of the
radiation accident.
The early phase (initial period) of an accident covers the time from its beginning to
the time when the atmospheric release of radioactive substances is arrested. This
period is assumed to be up to ten days. According to national regulatory
requirements, the dose limits below which no urgent decisions have to be made
during the early period of a radiation accident are:
• 0.5 cSv (rem) for the whole body.

• 5.0 cSv (rem) for the thyroid.
The design radius of the control area around an RBMK NPP is 3 km.
The main pathways of radiation effects on the population during this period are:
- External γ and β irradiation during the passage of the radioactive cloud.
- Internal irradiation through inhalation of radioactive substances.
The permissible radiation doses in DBA analysis should be confirmed with the
following conservative assumptions:
a) The radioactive release to the environment is a single event of short
duration, and the release height is equal to the source altitude above
ground level.
b) The plume rise due to its buoyancy is disregarded.
c) The radiation doses are calculated for the worst weather conditions and for
the specific elevation of the release source, with the wind speed and
atmospheric conditions producing the greatest possible near ground
concentrations of radionuclides.
According to national regulatory requirements, the probability of large radiological

release for BDBA should be less than 10-7 per reactor-year.
3.2.2. Recent requirements by RosTechnadzor pertaining to BDBA

analysis including SA
The concept of management of BDBA was finally generated in Russia after
Chernobyl accident as an additional fourth level of in-depth safety barriers of the
nuclear power plant. As a result, the concept of a ‘beyond design basis accident’
has appeared, i.e. an accident caused by initial events that have not been taken
into account or accompanied by beyond additional single failures of safety systems
in comparison with design basis accident, realization of erroneous decisions of the
personnel which can lead to severe damages or melting of the reactor core. It is
natural that consequences of such accident can be much severe, than at design
basis accidents.
Management of BDBA according to [15] constitutes the envelope of the actions

directed on prevention of development of DBA into BDBA and on mitigation of
consequences of BDBA. For this purpose any technical means available in an
62
efficient condition intended for normal operation and providing safety at design
basis accidents or for mitigation of consequences at beyond design basis accidents
can be used. All these actions and special means form the mentioned above fourth
level of in-depth safety barriers.
ОПБ-88/97 [15] — the Russian regulating document of the maximum conceptual

level of hierarchy — defines the concept of safety accepted in Russia today. It was
developed after the Chernobyl accident simultaneously with the IAEA document
INSAG-3, [16], reflecting the modern concept of safety at the international level. As
the comparative analysis of these documents performed by special consulting
IAEA group in the report [17] has shown, the concept of safety reflected in ОПБ-
88/97 basically corresponds to a modern international level.
The concept of beyond design basis accidents accepted in Russia is most

completely reflected in ОПБ-88/97. In other normative documents of Russia it is
supplemented only with some more specific requirements. The concept is based
on the requirement of restriction of radiation influence at the BDBA level by
employing of measures on their management and realization on a site of the
nuclear power plant and surrounding it area of actions on protection of the
personnel and the population. These measures are part of in-depth safety barriers
and in item 1.2.3 ОПБ-88/97 are specified into detail.
For some BDBA, the level of restriction of radiation influence is caused by criterion
of the radiation safety, established in the document [18] determining the
requirements on layout of the nuclear power plant. This criterion limits so-called
limiting emergency discharge at beyond design basis accidents so that irradiation
doses of the concerned population (critical group) on the border of the zone of
planning of protective actions and outside did not exceed 5 mSv for all body and 50
mSv for separate organs in the first year after the accident.
According to the requirement of item 1.2.17 ОПБ-88, the probability of limiting

emergency discharge should be lower 10-7 on reactor in one year. It is necessary
to avoid evacuation of the population located outside the specified zone of planning
anti-emergency actions. If the given requirement is not achieved, additional
technical measures on management of BDBA for mitigation its consequences
should be accepted (first principle based on probability).
The second principle (item 1.2.14 ОПБ-88/97) establishes for any hypothetic
(including BDBA) event that, on the basis of intrinsic features of the NPP or in case
any AM procedure is not applicable, the development of measures on management
of this event must be provided irrespective of its probability.
Additional means and strategies useful for the management of BDBA are
discussed in ОПБ-88/97.
The Russian regulatory requirements regarding the mandatory execution of

suitable probabilistic safety studies of nuclear power plants is defined by item
1.2.16 ОПБ-88 where PSA level 2 analyses are specifically mentioned. The IAEA
recommendations of INSAG-6, [19], continue to remain valid.
63
In the mentioned regulatory documents it is recognized that lists of beyond design
basis accidents cannot be developed only on the basis of probabilistic criteria. The
application of deterministic principles and approaches is necessary, as outlined
below.
Deterministic safety analysis is based upon the method of “postulated initial event”
and the principle of “single failure” with the definition, for each accident, of
meaningful sequence of events (i.e. DBA) and calculation transient scenarios and
radiological consequences. This approach in essence is systematic and provides
the necessary completeness and reliability for the safety analysis.
For beyond design basis accidents such an approach cannot be applied. BDBA
arise at beyond design initial events or at occurrence of additional failures beyond
the postulated “single failure”. As a result, the number of possible scenarios of
beyond design basis accidents is practically unlimited. Therefore the possibility of
application of the event-based approach for the management of BDBA shall be
excluded. Rather BDBA management should be based on a symptom approach.
Levels of severity may characterize symptoms of BDBA conditions and are

connected with the damage (or damage rate) of physical barriers and with the
possibility of fission product release to the environment the on a way of release of
radioactive products into an environment. For the nuclear power plants with RBMK
reactors such barriers are fuel element, including a fuel matrix and cladding, fuel
channel tube, bounds of reactor cavity, bounds of coolant circulation circuit and the
leaktight protection of the reactor (reactor building). For the nuclear power plants
with VVER reactors instead of fuel channel tube and bounds of reactor cavity, the
reactor vessel is considered.
BDBA symptoms not related to any specified scenarios shall be adopted for the
transition from DBA emergency (i.e. the domain of EOP) to the BDBA emergency
(i.e. the domain of AM). The formation of a scale of emergency conditions and the
possibility of their identification is rather essential. If any BDBA emergency
condition cannot be identified, it should be excluded from the consideration for the
planning of countermeasures.
The management of any emergency condition requires the identification and

management of certain safety functions which performance could stop the further
development of the accident (i.e. preventing transition of the given emergency
condition to another with the higher level of severity), thus making possible to
improve the starting emergency condition. Such safety functions can be named as
critical.
Therefore, alongside with levels of severity of (BDBA) emergency conditions, the

symptoms related to them and the critical safety functions should be determined.
64
In this way the entire set of considered BDBA emergency conditions is not defined
in an objective way, but depends on our choice, so the selected BDBA emergency
conditions can be considered as postulated. Making the analogy with the DBA
conditions, where the method of “postulated initial event” is adopted, in the case of
BDBA the method of “postulated emergency conditions” can be adopted.
For the development of manuals on management of BDBA it is necessary to define

time and parametrical frameworks of occurrence of postulated emergency
conditions (i.e. characterizing the concerned BDBA), in order to address the
actions of operators, to establish diagnostic and functional priorities for each level
of severity and to form a set of general functional instructions connected with the
status of the relevant (and critical) safety functions. For this purpose it is necessary
to execute detailed analyses of some characteristic BDBA scenarios. As a result
from the above task, suitable lists of BDBA postulated emergency conditions shall
be developed together with the corresponding symptoms and critical functions.
It shall be emphasized again that in this context, the development of the lists of
BDBA (including the related predicted scenarios) play an important, but an auxiliary
role. Definitely, the BDBA in the Russian normative documents is treated the same
as the DBA i.e. a combination of deterministic and probabilistic approaches that
meet the modern practice standard in the world is recommended.
3.3. Identification and characterization of selected RBMK

accident scenarios and phenomena
The detailed analysis of phenomena relevant for the transient scenarios that
constitute the objective of the PhD thesis are discussed hereafter. Then, in the
following chapters of this document, details are given about results of codes and
methods applications that have been achieved within the present framework and
are connected with the quantities that control the concerned phenomena or
transient scenarios.
Hereafter, introductive remarks are provided that also characterizes the state of the
art in the knowledge of those RBMK accident scenarios and phenomena prior to
the execution of the PhD activities. Most of the information is taken from
documents issued within TACIS Project R2.03/97, [20]. References listed in this
report are relevant for the characterization of RBMK scenarios.
3.3.1. Thermal-hydraulics of PS
The RBMK is a water cooled reactor. Therefore, transient phenomena studied for
LWR are also supposed to be applicable in RBMK conditions. However, attention
should be paid to the relevant range of parameters that can be largely different in
LWR and RBMK situations. Typical examples are constituted by the length of the
core active region, almost twice the value that characterize BWR, the presence of
long (L/D >> 100) pipes at the inlet and the outlet of FC and of large volume (> 50
m3) steam drum. All of this has large influence upon the nominal conditions
pressure drop across the loop, upon the LOCA depressurization and upon stability
65
performance. An overview of key-subject and of phenomena taken from the
validation of the Korsar code can be found in Tab. 12, [20]. The application of the
table-of-phenomena approach is well established in nuclear reactor safety (e.g.
including code validation), following the pioneering study performed by
OECD/CSNI at the end of ‘80s, [21], [22], [23]. In the present case it also gives an
idea of different relevance of thermal-hydraulic phenomena in the cases of VVER
and RBMK. More detailed RBMK thermal-hydraulic phenomena can be found in
Tab. 13, [20].
Tab. 12 – List of key-subjects and phenomena utilized for the assessment of

Korsar code and comparison between VVER and RBMK.
Key-Subject Phenomena VVER RBMK

Reactor kinetics (Point R R
model)
Reactor kinetics (Spatial R R
model)
Modeling of liquid R N/A
absorber
Modeling of non-
Non condensable gas R R
condensable gases
Critical discharge of the
Break flow R R
coolant
Core sub-critical heat transfer R R
Core critical heat transfer R R
Core supercritical heat transfer R R
Quench front formation and
R R
propagation
Flow regimes in reactor
core, core heat transfer Radiation heat transfer R R
(in the rod assemblies) Flooding of counter-current flows
of water and steam in vertical R R
channels
Reflood R R
Heat transfer in radial direction
N/A R
between different heat structures
Sub-critical, critical and
R N/A
supercritical heat transfer
Quench front formation and
R N/A
Flow regimes and heat propagation
transfer in the steam Radiation heat transfer R N/A
generating channels Counter-current flow R N/A
Reflood R N/A
Steam–water–gas mixture flow
N/A R
and heat transfer in reactor cavity
66
Tab. 12 (cont.) – List of key-subjects and phenomena utilized for the assessment of
Korsar code and comparison between VVER and RBMK.
Key-Subject Phenomena VVER RBMK

Stratification of the two-fluid flow
R R
in the horizontal channels
Natural circulation R R
Separation of phases in tees N/A R
Flow regimes in reactor
Steam condensation during ECCS
cooling circuit R R
injection
Counter-current water–air flow
R R
through perforated plates
Flashing and condensation in DS N/A R
Two-phase pump behavior R R
Fuel melting and relocation R R
Zirconium–steam reaction R R
Severe accident
propagation Steam–graphite interaction N/A R
Fuel melt – pressure tube –
N/A R
graphite interaction
Tab. 13 – List of RBMK specific thermal-hydraulic phenomena for MCC.
Processes in drum separator

• Separation, water entrainment, steam carry under:
− void fraction of ‘water’ volumes of DS.
• Dynamics of mass levels, their variation during power and/or pressure
decrease:
− Axial effects in DS:
o Heat exchange steam–water–metal during variation of pressure.
Two-phase severely non-homogeneous phenomena and natural
circulation in complex circuit during decrease of pressure, water
inventory and velocity
• Separation in T-joints, headers, W-sections.
• Steam plugs and oscillations in steam water pipelines, group distribution
headers, down-comers.
• CCF/ CCFL in parallel channels and inlet water pipelines.
• Self-sustaining flow oscillations in parallel channels.
• Degradation of natural circulation.
67
Tab. 13 (cont.) – List of RBMK specific thermal-hydraulic phenomena for MCC.
Fast thermal processes

• Water blowdown from the circuit, water blow down from fuel channels
during drop of flow rate or pressure, overflows between parallel channels.
• Stagnation, oscillations, reverse and resumption of flow in different-heated
channels.
• Fronts of flooding and quenching in the channel with fuel assembly,
distribution of ECCS water over differently heated fuel channels and GDH.
• Dynamics of temperatures in fuel assemblies and pressure tubes;
convection, radiation and heat conductivity from ‘steamed’ fuel channels to
graphite column.
Slow processes in drained part of channels
• Dynamics of formation of back flow during leak in the pressure part of the
circuit.
• Formation and movement of a physical level in fuel channels, water
entrainment, heat removal above level.
• Axial and longitudinal transport of heat to graphite and cooled channels.
• Steam–zirconium reaction.
Quasi-stationary processes
• Critical heat fluxes and critical powers of fuel channels.
• Void distribution.
• Post-dryout heat transfer under emergency conditions.
• Oscillation in parallel channels.
• Hydraulic resistance of fuel channel during sub-cooling at the inlet and
overheating at the outlet.
• Natural circulation.
• Thermal-hydraulics (dryout, post-dryout heat transfer, formation of level) in
downward two-phase flow.
Phenomena in reactor space, outside the channels
• Two-phase flow after rupture, 3-D distribution of flow and pressure,
separation in reactor space outside fuel channel during distortion of
graphite columns.
• Steam generation (quenching of graphite).
• Condensation on CPS channels, role of non-condensable gases.
Processes in CPS channels
• Hydraulic dynamics of low pressure downward flow in CPS channels.
• Formation and stability of water film during decrease of flow rate, transfer
and release of gases.
• Dynamics of emergency drainage of CPS channels, processes of water
drainage and dryout.
68
3.3.2. Three-dimensional neutron kinetics
The RBMK core consists of FC embedded into graphite stacks, subdivided into
bricks (see Chapter 2). Looking at the neutron kinetics point of view, the core of a
RBMK largely differs from the core of a LWR (namely a BWR) producing the same
thermal power owing to the following:
a. Overall dimensions and, consequently, power per unit volume, that are
respectively, much larger and much smaller (for a factor greater than 10, see
Tab. 6).
b. Presence of the graphite as moderator.
c. Presence of a large variety of ‘neutron kinetics cell types’ (i.e. association of a
graphite stack and of a different component):
1. Active fuel.
2. Additional absorber (high pressure bypass).
3. Water column, i.e. w/o inside absorber (high pressure bypass).
4. Reflector (low pressure circuit).
5. CPS or 3 CR types (low pressure circuit):
i. Manual CR,
ii. Short CR,
iii. Safety CR.
6. Axial detector (low pressure circuit).
d. Lower enrichment and, consequently, need to control a lower excess reactivity
owing to the ‘continuous-fuel-reloading’.
Specific RBMK core features for 3D neutron kinetics are given in Tab. 14, [20]. All
of this makes more stringent (related to BWR) the need to use 3D neutron kinetics
for transient safety analyses.
The two-group cross-section libraries used to predict RBMK core physics are
usually corrected for uncertainties in axial burn-up distribution, control rod positions
and thermal-hydraulic feedback. Flux reconstruction methods or other correction
procedures are applied. In the flux reconstruction method, the calculated 3-D flux
distribution is compared to the readings of the in-core detectors and cross-sections
are adjusted accordingly.
69
Tab. 14 – List of RBMK specific design features for the core neutron kinetics
Fuel cell:
• Coolant density
(0.0g/cm3 up to 1.0g/cm3).
• Graphite temperature
(300 K up to 1100 K).
• Fuel temperature
(300 K up to 1500 K).
• Xenon-concentration.
• Coolant temperature.
(300 K up to 557 K).
Non-fuel cells:
• Control rod channels.
• Additional absorber channels.
• Reflector channels.
• CPS-detector channels.
The cross-sections are typically derived as

function of:
• Fuel burn-up.
• Graphite temperature.
• Water density
(0÷0.78 g/cm3 for cells in MCC loop)
(0÷0.98 g/cm3 for cells in CPS loop for hot power
conditions)
(0÷1.0 g/cm3 for cold conditions in any cell).
70
4. THE TOOLS, THE METHODOLOGIES AND THE
QUALIFICATION OF 3D NK COUPLED TH CODES ANALYSES
The execution of detailed and realistic NPP nuclear safety analyses requires the
use of the state-of-the-art codes and methodologies. Information about these
topics is given this chapter, together with some details about the procedure for
codes and input decks qualification.
4.1. Neutron Transport codes

Performing reactor dynamic core calculations solving the neutron transport
equation is something that is still impossible nowadays, because of the limits of
current computational tools. For the RBMK, this is particularly true, considering its
enormous geometrical dimensions (14 m of diameter per 7.5 m of height).
Therefore, the use of these codes should be limited to the investigation of small
parts of the reactor (e.g., the single fuel cell or small part of the reactor lattice).
Nevertheless their use is precious in conducting in-depth calculations for assessing
the effects of some local phenomena (e.g., for the RBMK, the assessment of the
effects of a single channel voiding) and for performing the derivation of cell
homogeneous cross sections to be used for the whole reactor core calculations by
nodal diffusion codes.
During this PhD research activity, the situations for using neutron transport codes
were identified and both deterministic and stochastic transport codes were applied.
Stochastic neutron transport simulation by the MCNP5 [24] Monte Carlo based
method code was performed for obtaining detailed simulation of fuel cells during a
flow blockage event. Instead, deterministic neutron transport codes DRAGON [56]
and HELIOS [54] were applied in order to perform repetitive fuel lattice cells
calculations for the production of several fuel cell homogenized cross-section
database at several burnup steps. The description of the physical and numerical
methods of all these codes and their structure is given hereafter.
4.1.1. The Monte Carlo code MCNP5

A comprehensive description of the MCNP code can be found in [24]. We reported
here the main relevant features of the code for the use we did for the RBMK
analysis.
4.1.1.1. General features

MCNP is a general-purpose, continuous-energy, generalized-geometry, time-
dependent, coupled neutron/photon/electron Monte Carlo transport code. It can be
used in several transport modes: neutron only, photon only, electron only,
combined neutron/photon transport where the photons are produced by neutron
interactions, neutron/photon/electron, photon/electron, or electron/photon.
The neutron energy regime is from 10-11 MeV to 20 MeV for all isotopes and up to
150 MeV for some isotopes, the photon energy regime is from 1 keV to 100 GeV,
and the electron energy regime is from 1 KeV to 1 GeV. The capability to calculate
71
keff eigenvalues for fissile systems is also a standard feature and was heavily
exploited for this PhD work.
The user creates an input file that is subsequently read by MCNP. This file contains
information about the problem in areas such as: the geometry specification, the
description of materials and selection of cross-section evaluations, the location and
characteristics of the neutron, photon, or electron source, the type of answers or
tallies desired, and any variance reduction techniques used to improve efficiency.
4.1.1.2. Monte Carlo Method vs. Deterministic Method

Monte Carlo methods are very different from deterministic transport methods.
Deterministic methods, the most common of which is the discrete ordinates
method, solve the transport equation for the average particle behavior. By contrast,
Monte Carlo obtains answers by simulating individual particles and recording some
aspects (tallies) of their average behavior. The average behavior of particles in the
physical system is then inferred (using the central limit theorem) from the average
behavior of the simulated particles. Not only are Monte Carlo and deterministic
methods very different ways of solving a problem, even what constitutes a solution
is different. Deterministic methods typically give fairly complete information (for
example, flux) throughout the phase space of the problem. Monte Carlo supplies
information only about specific tallies requested by the user.
4.1.1.3. The Monte Carlo Method

Monte Carlo can be used to duplicate theoretically a statistical process (such as
the interaction of nuclear particles with materials) and is particularly useful for
complex problems that cannot be modelled by computer codes that use
deterministic methods. The individual probabilistic events that comprise a process
are simulated sequentially. The probability distributions governing these events are
statistically sampled to describe the total phenomenon. In general, the simulation is
performed on a digital computer because the number of trials necessary to
adequately describe the phenomenon is usually quite large. The statistical
sampling process is based on the selection of random numbers—analogous to
throwing dice in a gambling casino—hence the name “Monte Carlo.” In particle
transport, the Monte Carlo technique is pre-eminently realistic (a numerical
experiment). It consists of actually following each of many particles from a source
throughout its life to its death in some terminal category (absorption, escape, etc.).
Probability distributions are randomly sampled using transport data to determine
the outcome at each step of its life.
Fig. 40 represents the random history of a neutron incident on a slab of material

that can undergo fission. Numbers between 0 and 1 are selected randomly to
determine what (if any) and where interaction takes place, based on the rules
(physics) and probabilities (transport data) governing the processes and materials
involved. In this particular example, a neutron collision occurs at event 1. The
neutron is scattered in the direction shown, which is selected randomly from the
physical scattering distribution.
72
Fig. 40 – A typical neutron history simulation by MNCP5 (from [24])
A photon is also produced and is temporarily stored, or banked, for later analysis.
At event 2, fission occurs, resulting in the termination of the incoming neutron and
the birth of two outgoing neutrons and one photon. One neutron and the photon
are banked for later analysis. The first fission neutron is captured at event 3 and
terminated. The banked neutron is now retrieved and, by random sampling, leaks
out of the slab at event 4. The fission-produced photon has a collision at event 5
and leaks out at event 6. The remaining photon generated at event 1 is now
followed with a capture at event 7. Note that MCNP retrieves banked particles such
that the last particle stored in the bank is the first particle taken out.
This neutron history is now complete.
As more and more such histories are followed, the neutron and photon distributions
become better known. The quantities of interest (whatever the user requests) are
tallied, along with estimates of the statistical precision (uncertainty) of the results.
4.1.1.4. Nuclear Data and Reactions

MCNP uses continuous-energy nuclear and atomic data libraries. The primary
sources of nuclear data are evaluations from the Evaluated Nuclear Data File
(ENDF) system [61], the Evaluated Nuclear Data Library (ENDL) [62], [55],
Evaluated Photon Data Library (EPDL) [63], the Activation Library (ACTL)
compilations from Livermore, and evaluations from the Nuclear Physics (T–16)
Group at Los Alamos. Evaluated data are processed into a format appropriate for
MCNP by codes such as NJOY [64], [65]. The processed nuclear data libraries
retain as much detail from the original evaluations as is feasible to faithfully
reproduce the evaluator’s intent.
Nuclear data tables exist for neutron interactions, neutron-induced photons, photon
interactions, neutron dosimetry or activation, and thermal particle scattering S(α,β).
Each data table available to MCNP is listed on a directory file, XSDIR. Users may
select specific data tables through unique identifiers for each table, called ZAIDs.
These identifiers generally contain the atomic number Z, mass number A, and
73
library specifier ID. Over 836 neutron interaction tables are available for
approximately 100 different isotopes and elements. Multiple tables for a single
isotope are provided primarily because data have been derived from different
evaluations, but also because of different temperature regimes and different
processing tolerances. More neutron interaction tables are constantly being added
as new and revised evaluations become available.
4.1.1.5. Tallies and Output

The user can instruct MCNP to make various tallies related to particle current,
particle flux, and energy deposition. MCNP tallies are normalized to be per starting
particle except for a few special cases with criticality sources. Fluxes across any
set of surfaces, surface segments, sum of surfaces, and in cells, cell segments, or
sum of cells are also available. Tallies such as the number of fissions, the number
of absorptions, the total helium production, or any product of the flux times the
approximately 100 standard ENDF reactions plus several nonstandard ones may
be calculated with any of the MCNP tallies. In fact, any quantity of the form (1)
C= ∫ Φ( E ) f ( E )dE
(1)
can be tallied, where F(E) is the energy-dependent fluence, and f(E) is any product
or summation of the quantities in the cross-section libraries or a response function
provided by the user.
4.1.1.6. Estimation of Monte Carlo Errors

MCNP tallies are normalized to be per starting particle and are printed in the output
accompanied by a second number R, which is the estimated relative error defined
to be one estimated standard deviation of the mean Sx divided by the estimated
mean x . In MCNP, the quantities required for this error estimate −the tally and its
second moment− are computed after each complete Monte Carlo history, which
accounts for the fact that the various contributions to a tally from the same history
are correlated. For a well-behaved tally, R will be proportional to 1/N1/2 where N is
the number of histories. Thus, to halve R, we must increase the total number of
histories fourfold. For a poorly behaved tally, R may increase as the number of
histories increases.
The estimated relative error can be used to form confidence intervals about the
estimated mean, allowing one to make a statement about what the true result is.
The Central Limit Theorem states that as N approaches infinity there is a 68%
chance that the true result will be in the range x (1≤ R) and a 95% chance in the
range x (1≤ 2R).
74
4.1.2. The deterministic transport codes
As stated before, HELIOS and DRAGON deterministic neutron transport codes
were applied for the calculation of fuel cell homogenized cross sections database.
Generation of a cross sections database for RBMK reactor, composed by several
libraries, is considered a task requiring big efforts by the analysts. There are
several reasons for this statement.
One of this is the intrinsic complexity of the RBMK system due, for example, to the
very large number of different elementary cells, to the different types of CR or to
the physical separation between moderator and coolant.
Difficulties are generally due also to some lack of accuracies in the acquisition of
the basic input data (burn-up distribution, CR positions, coolant mass flow in each
channel, graphite temperatures) in the NPP. Consequently, methods were
introduced in the past for correcting the results obtained by the use of cross section
libraries, in order to overcome the big uncertainties resulting in the transient core
codes calculations. For example, in the flux reconstruction method, the calculated
3-D flux distribution is compared to the readings of the in-core detectors and cross-
sections (e.g., the thermal fission cross sections, or νΣf ) are adjusted accordingly
“a posteriori” [25]. Calculations of the cross section libraries of this PhD research
activity did not required application of “a posteriori” methods. In fact, to overcome
in-core detectors inaccuracy, a first correction of the channels burnup distribution
was performed before libraries calculation [60]. Nevertheless, local channels
correction was executed always preserving the reactor macro-areas power
distribution. Before the use of the aforementioned codes, a process for identifying
the RBMK fuel cell peculiarities was performed. Codes selection was based on
their capabilities in handling these peculiarities.
a) Geometric Capabilities
The code has to be able to model the elementary RBMK core cell made by a
square hollow block of graphite measuring 0.25 x 0.25 m and made by other
components of different shapes like fuel pins, pressure tube, additional absorber,
control rods parts, neutron detectors and graphite radial reflector. Therefore, the
capability to use Cartesian geometry is required. The code has to have the
capability to handle also a “poly-cell” model, in order to calculate the feedback
effects of the neighbor cells.
b) Physical Capabilities
An adequate range of the relevant physical parameters has to be considered by
the code in order to calculate a good transport solution and in order to include the
relevant feedback effects during the fuel cell homogenized cross sections libraries
calculations. Different requirements can be listed for fuel and non-fuel cells (see
Tab. 15) according to the family of transients that it should be analyzed.
75
Tab. 15 – Range of the physical parameters to be considered for Fuel and Non-
fuel cells
Fuel cell Non-fuel cells

• Fuel Burnup • Graphite temperature
(0 – 20 MWd/KgU) (300 K up to 1100 K);
• Coolant density • Water density
(0.0 g/cm3 up to 1.0 g/cm3) (0÷0.78 g/cm3 for cells in MCC loop)
• Graphite temperature (0÷0.98 g/cm3 for cells in CPS loop
(300 K up to 1100 K) for hot power conditions)
• Fuel temperature (0÷1.0 g/cm3 for conditions in any
(300 K up to 1500 K) neighbor cell)
• Xe135 concentration (0, 2.0E+13, • CR position
4.5E+48 atoms/cm3)
• Coolant temperature
(300 K up to 557 K)
4.1.2.1. The lattice physics code DRAGON

The computer code DRAGON [56] contains a collection of models that can
simulate the neutron behavior of a unit cell or a fuel assembly in a nuclear reactor.
It includes all of the functions that characterize a lattice cell code, namely: the
interpolation of microscopic cross sections supplied by means of standard libraries;
resonance self-shielding calculations in multidimensional geometries; multigroup
and multidimensional neutron flux calculations that can take into account neutron
leakage; transport-transport or transport-diffusion equivalence calculations as well
as editing of condensed and homogenized nuclear properties for reactor
calculations; and finally isotopic depletion calculations.
The code DRAGON contains a multigroup flux solver conceived that can use a
various algorithms to solve the neutron transport equation for the spatial and
angular distribution of the flux. Each of these algorithms is presented in the form of
a one-group solution procedure where the contributions from other energy groups
are considered as sources. The current release of DRAGON contains five such
algorithms.
The JPM option that solves the integral transport equation using the J± method,
(interface current method applied to homogeneous blocks); the SYBIL option that
solves the integral transport equation using the collision probability method for
simple 1D or 2D geometries and the interface current method for 2D Cartesian or
hexagonal assemblies; the EXCELL/NXT option to solve the integral transport
equation using the collision probability method for more general 2D geometries and
for 3D assemblies; the MOCC option to solve the transport equation using the
method of cyclic characteristics in 2D Cartesian, and finally the MCU option to
solve the transport equation using the method of characteristics (non cyclic) for 3D
Cartesian geometries.
The execution of DRAGON is managed via the GAN generalized driver. After the
collision probability or response matrices associated with a given cell have been
76
generated, the multigroup solution module can be activated. This module uses the
power iteration method and requires a number of iteration types. The thermal
iterations are carried out by DRAGON so as to rebalance the flux distribution only
in cases where neutron undergoes up-scattering.
The power iterations are performed by DRAGON to solve the fixed source or
eigenvalue problem in the cases where a multiplicative medium is analyzed. The
effective multiplication factor (keff ) is obtained during the power iterations. A search
for the critical buckling may be superimposed upon the power iterations so as to
force the multiplication factor to take on a fixed value. DRAGON can access
directly microscopic cross-section libraries defined according to the following
standard formats: DRAGLIB, MATXS, WIMS–D4 [66],[67],[68],[69], and WIMS–
AECL [70]. It has the capability of exchanging macroscopic cross-section libraries
with a codes such as TRANSX–CTR or TRANSX–2 by the use of GOXS and
ISOTXS format files. The macroscopic cross section can also be read in DRAGON
via the input data stream. The code is modular and can be interfaced easily with
other production codes.
4.1.2.2. The lattice physics code HELIOS

HELIOS [54] is a neutron and gamma transport code for lattice burnup, in general
two-dimensional geometry. It was developed by Studsvik™ ScandPower since the
1993. The code version released in April 2000, HELIOS-1.6, was used for this PhD
activity. As can be seen in Fig. 41, HELIOS is composed by several modules. In
particular, there are:
- AURORA, the input processor code module
- ZENITH, the output processor code module
The data flow between these codes is via a data base that is accessed and
maintained by the subroutine package HERMES.
Fig. 41 – HELIOS package
AURORA reads, processes and saves the User’s input. The result is a number of
so-called ‘User-arrays’. The User-arrays are written into the HERMES database for
77
retrieval by HELIOS and/or ZENITH. For each case HELIOS retrieves the input
from the database and executes calculations specified. The input, except for the
basic nuclear data in the library which are not user-specified consists of the
following data types:
1. The nuclear data library with the basic nuclear data, which also defines the
energy discretization (group structures) of the particle transport calculations.
2. Data that define the (initial) number densities of materials, and the elements of
the albedo matrix.
3. Data that define the geometry of the system, including the spatial and angular
discretization to be used in the transport calculations.
4. Data that assign one or more property sets to the geometric system, thus
defining one or more states of the system.
5. Data that define the execution sequence of the calculations
6. Data that define what output will be saved
HELIOS consists of a main module, which calls nineteen computational modules,

The first eleven modules treat the input data and they are called once per every
case. Six of the remaining modules perform physics calculations, while the last two
modules process the data for output and restart dumps. Most of these eight
modules are called at least once per reactivity calculation point. The depletion
module in only called if there is burnup or decay, while the module for the restart
dumps is called only when such dumps are requested.
Therefore HELIOS performs the lattice calculations in the following manner. The
methods described cover three areas:
• The geometric buildup of the system and its properties
• The physical methods to obtain fluxes, currents and number densities. This
consists of five parts:
o Calculating resonance-shielded microscopic Cross Sections
o Calculating fluxes and currents by the current-coupling collision
probability (CCCP) method for particle transport
o Evaluating first-flight probabilities
o Evaluating, with the B1 method, the criticality spectrum, which is used
to rebalance the spectrum of the CCCP solution
o Solving the burnup chains to obtain new number densities.
• Output processing to obtain the output data arrays.
Almost all data in the library are based on the ENDF/B-VI data. The exceptions are
the cross-sections of erbium and thulium isotopes and the (n,γ) matrices of many
fission products. In the thermal and resolved-resonance energy regions, below 2
keV the cross-sections of the erbium and thulium isotopes are constructed by the
RABBLE code.
Fig. 42 shows the flow scheme of library generation. The ENDF/B-VI files are
processed with a version of NJOY91.13 [65] – with upgrades through version
91.105 that includes code RABBLE as a module. The main task of NJOY is to
generate infinite-dilution neutron and gamma Cross Sections in 190 neutron and
48 gamma groups. Another NJOY task is to generate PENDF’s which are files with
78
Cross Sections in many thousands of energy points that are input to RABBLE.
While all isotopes must be processed with NJOY, only those to be treated as
resonance isotopes by HELIOS have to be processed by RABBLE.
The GENDF group Cross Sections and the resonance-shielded Cross Sections
produced by RABBLE are collected in the HERMES file, the master database. The
creation of the database is one of the activities of HEBE code. Another activity of
the HEBE code is to construct a master library for HELIOS from selected data in
the database. With the master library HELIOS runs are made to generate flux
spectra in different regions, at different temperatures and at different burnups.
In the third activity HEBE adds these spectra to the master database. In the fourth
activity they are used to group-collapse the database into a condensed database,
from which the condensed library is made.
Fig. 42 – Flow scheme of library generation.
Presently, three libraries exist: the master library with 190/48 neutron/gamma
groups: a fast reactor library with 112/18 neutron/gamma groups and a production
library with 45/18 neutron/gamma groups. For the cross-section generation the
79
master library with 190/48 neutron/gamma groups were used. A comparison with
the results from the 45/18 group library showed a good agreement to the master
library.
286 isotopes, elements and mixtures are included in the HELIOS database, 28
heavy isotopes and 121 fission products. The code also contains resonance tables
for 39 resonance isotopes (11 heavy isotopes, 28 fission products) and 23
burnable absorber and control isotopes. Isotopic compositions are automatically
saved if they are to be used for branch-off calculations. Compositions (e.g. boron
contents, water density, temperatures and material densities) can be changed in
branch-offs.
One of the advantages of HELIOS is its geometry flexibility, which allows

representing fully the RBMK cell geometry. The HELIOS modeling capabilities
permit also to represent exactly non-homogeneous assembles (i.e. each rod in the
assembly). The HELIOS method is the CCCP technique. HELIOS was qualified for
RBMK applications based on cross code comparisons and using measured data of
critical facilities [26].
First, in summary the important issues in the RBMK cross-section generation are:
• Input library – evaluated nuclear data files,
• Resonance treatment,
• Energy condensation,
• Main transport calculations,
• Homogenization,
• Depletion.
Interface to run the lattice physics code HELIOS in an automated manner to

perform depletion and branch calculations, extract macroscopic cross-section and
neutron kinetics data, and assemble them in the format of cross-section libraries
was developed at the Reactor Dynamics and Fuel Management Group (RDFMG)
of the Pennsylvania State University (PSU).
4.2. The RELAP5-3D system code

4.2.1. Introduction
The RELAP5 series of codes has been developed at the Idaho National Laboratory
(INL) under sponsorship by the U.S. Nuclear Regulatory Commission (NRC), the
U.S. Department of Energy (DOE), and a consortium of several countries and
domestic organizations that were members of the International Code Assessment
and Applications Programme (ICAP) and its successor, the Code Applications and
Maintenance Programme. Specific applications of the code have included
simulations of transients in LWR systems, such as loss of coolant, anticipated
transients without scram (ATWS), and operational transients such as loss of
feedwater, loss of offsite power, station blackout, and turbine trip.
RELAP5-3D, the latest code version in the series of RELAP5 codes, is a highly
generic code that, in addition to calculating the behavior of a reactor coolant
system during a transient, can be used for simulation of a wide variety of hydraulic
80
and thermal transients in both nuclear and nonnuclear systems involving mixtures
of steam, water, non-condensable, and solute. Development of RELAP5-3D The
RELAP5-3D code version is a successor to the RELAP5/MOD3 code that was
developed at the INL for the NRC. Development of the RELAP5 code series
began at the INL under NRC sponsorship in 1975 and continued through several
released versions, ending in October 1997 with the soon to be released
RELAP5/MOD3.3.
The U.S. DOE began sponsoring additional RELAP5 development in the early
1980s to meet its own reactor safety assessment needs. Following the accident at
Chernobyl, DOE undertook a re-assessment of the safety of all of its test and
production reactors throughout the United States. The RELAP5 code was chosen
as the thermal-hydraulic analysis tool because of its widespread acceptance.
Systematic safety analyses were carried out for the DOE that included the N
reactor at Hanford, the K and L reactors at Savannah River, the Advanced Test
Reactor (ATR) at INL, the High Flux Isotope Reactor (HFIR) and Advanced
Neutron Source (ANS) at Oak Ridge, and the High Flux Beam Reactor (HFBR) at
Brookhaven. DOE also chose RELAP5 for the independent safety analysis of the
New Production Reactor (NPR) proposed for Savannah River before that
programme was cancelled in the wake of the end of the cold war.
The application of RELAP5 to these various reactor designs created the need for
new modelling capabilities. For example, the analysis of the Savannah River
reactors necessitated the adding three-dimensional flow model and heavy water
properties to the code. ATR required a new critical heat flux correlation applicable
to its unique fuel design. All together, DOE sponsored improvements and
enhancements have amounted to a multimillion-dollar investment in the code.
Toward the end of 1995, it became clear that the efficiencies realized by the
maintenance of a single source code for use by both NRC and DOE were being
overcome by the extra effort required to accommodate sometimes conflicting
requirements. The code was therefore “split” into two versions, one for NRC and
the other for DOE. The DOE version maintained all of had been sponsored by the
DOE before and after the split.
4.2.2. Relationship of RELAP5-3D to prior versions

At the outset of the decision to split the code into NRC and DOE versions, the INL
recognized the importance of retaining the pedigree stemming from the extensive
validation history of RELAP5/MOD3. Consequently, the developmental activities
with respect to RELAP5-3D since the split have been carefully integrated so as not
to compromise this legacy validation. In fact, virtually all of the enhancements in
RELAP5-3D are optional and supplemental to the proven performance of
RELAP5/MOD3.2. Consequently, users of RELAP5-3D can confidently apply the
code using existing, one-dimensional RELAP5/MOD3.2 input decks and expect
their results to be the same or improved.
81
4.2.3. The Thermal-hydraulic part: the RELAP5-3D code
RELAP5-3D is a highly generic code that, in addition to calculating the behavior of
a reactor coolant system during a transient, can be used for simulation of a wide
variety of hydraulic and thermal transients in both nuclear and nonnuclear systems
involving mixtures of vapor, liquid, non-condensable gases, and non-volatile solute.
The mission of the RELAP5-3D development program was to develop a code

version suitable for the analysis of all transients and postulated accidents in LWR
systems, including both large- and small-break loss-of-coolant accidents (LOCAs)
as well as the full range of operational and fusion reactor transient applications.
Additional capabilities include space reactor simulations, gas cooled reactor

applications, fast breeder reactor modeling, and cardiovascular blood flow
simulations.
The RELAP5-3D code is based on a non-homogeneous and non-equilibrium model

for the two-phase system that is solved by a fast, partially implicit numerical
scheme to permit economical calculation of system transients. The objective of the
RELAP5-3D development effort from the outset was to produce a code that
included important first-order effects necessary for accurate prediction of system
transients but that was sufficiently simple and cost effective so that parametric or
sensitivity studies were possible.
The code includes many generic component models from which general systems
can be simulated. The component models include pumps, valves, pipes, heat
releasing or absorbing structures, reactor kinetics, electric heaters, jet pumps,
turbines, separators, annuli, pressurizers, feedwater heaters, ECC mixers,
accumulators, and control system components. In addition, special process models
are included for effects such as form loss, flow at an abrupt area change,
branching, choked flow, boron tracking, and non-condensable gas transport.
The system mathematical models are coupled into an efficient code structure. The
code includes extensive input checking capability to help the user discover input
errors and inconsistencies. Also included are free-format input, restart,
renodalization, and variable output edit features. These user conveniences were
developed in recognition that generally the major cost associated with the use of a
system transient code is in the engineering labor and time involved in accumulating
system data and developing system models, while the computer cost associated
with generation of the final result is usually small.
The development of the models and code versions that constitute RELAP5-3D has
spanned more than two decades from the early stages of RELAP5 numerical
scheme development (circa 1976) to the present. RELAP5-3D represents the
aggregate accumulation of experience in modeling reactor core behavior during
accidents, two-phase flow processes, and LWR systems. The code development
has benefited from extensive application and comparison to experimental data in
the LOFT, PBF, Semiscale, ACRR, NRU, and other experimental programs.
82
The RELAP5-3D version contains several important enhancements over previous
versions of the code. The most prominent attribute that distinguishes the RELAP5-
3D code from the previous versions is the fully integrated, multi-dimensional
thermal- hydraulic and kinetic modeling capability. This removes any restrictions on
the applicability of the code to the full range of postulated reactor accidents.
Enhancements include a new matrix solver for 3D problems, new thermodynamic

properties for water, and improved time advancement for greater robustness. The
multi-dimensional component in RELAP5-3D was developed to allow the user to
more accurately model the multi-dimensional flow behavior that can be exhibited in
any component or region of a LWR system. Typically, this will be the lower plenum,
core, upper plenum and downcomer regions of an LWR. However, the model is
general, and is not restricted to use in the reactor vessel. The component defines a
one, two, or three- dimensional array of volumes and the internal junctions
connecting them. The geometry can be either Cartesian (x, y, z) or cylindrical (r, θ,
z). An orthogonal, three-dimensional grid is defined by mesh interval input data in
each of the three coordinate directions.
The new Border-Profile Lower Upper (BPLU) matrix solver is used to efficiently
solve sparse linear systems of the form AX = B. BPLU is designed to take
advantage of pipelines, vector hardware, and shared-memory parallel architecture
to run fast. BPLU is most efficient for solving systems that correspond to networks,
such as pipes, but is efficient for any system that it can permute into border-banded
form. Speed-ups over the previously used sparse matrix solver are achieved in
RELAP5-3D running with BPLU on multi-dimensional problems, for which it was
intended. For one-dimensional problems, the BPLU solver runs as fast or faster
than the previously used sparse matrix solver.
4.2.4. The three dimensional neutron kinetics routine: the NESTLE

code
4.2.4.1. The Steady-State and the Transient Problem
The RELAP5-3D code allows to calculate a 3D core power distribution thanks to
multi-dimensional neutron kinetics models based on NESTLE code, developed by
Paul Turinsky and his co-workers at the North Carolina State University. This
feature allows to compute the reactor fission power in either Cartesian or
hexagonal geometry [24].
The subroutines taken from the NESTLE source code used to solve these
problems were modified to be compatible with the coding standards and data
storage methodology used in RELAP5-3D, and were inserted into RELAP5-3D.
RELAP5-3D was modified to call the appropriate NESTLE subroutines depending
upon the options chosen by the user.
The Nodal Expansion Method (NEM) is used to solve the neutron diffusion
equations for the neutron flux in either two or four neutron energy groups. Quartic
or quadratic polynomial expansions for the transverse integrated fluxes are
employed for Cartesian or hexagonal geometries, respectively. E.g., the one
83
dimensional flux obtained by the transverse integration can be written (in the case
of Cartesian geometry) as:
(2)
where φ1g is the node average flux, implying for Equation (2) to be true that fn(x)
must be chosen such that the basis functions satisfy
, N=1 to 4
(3)
Transverse leakage terms are represented by a quadratic polynomial or constant
for Cartesian or hexagonal geometry, respectively. E.g., the x-direction spatial
dependence of the y-direction transverse leakage is approximated by:
(4)
where L1gy is the average y-directed leakage in node l, and the coefficients ρ1gy1
and ρ1gy2 can be expressed in terms of average y-directed leakages of the two
nearest-neighbor nodes along the x-direction (i.e. nodes l-1 and l+1) so as to
preserve the node average leakages of these three nodes. Discontinuity factors
(DFs) are utilized to correct for homogenization errors.
Transient problems utilize a user specified number of delayed neutron precursor

groups. Time discretization is done in a fully implicit manner utilizing a first-order
difference operator for the diffusion equation, or:
(5)
The precursor equations are analytically solved assuming the fission rate behaves
linearly over a time-step.
Independent of problem type, an outer-inner iterative strategy is employed to solve

the resulting matrix system. Outer iterations can employ Chebychev acceleration
and the Fixed Source Scaling Technique to accelerate convergence [27], [28], [29],
[30]. Inner iterations employ either a Red-Black Point or Line SOR iteration
schemes, dependent upon problem geometry. Values of the energy group
dependent optimum relaxation parameter and the number of inner iterations per
outer iteration to achieve a specified L2 relative error reduction are determined a
84
priori. The non-linear iterative strategy associated with NEM is utilized. This
technique was developed by Smith [31], [32], [33], and successfully implemented
into the Studsvik QPANDA and SIMULATE code packages.
The basic idea is applicable to the standard FDM solution algorithm of the multi-
group diffusion equation. Solving the FDM based equation utilizing an outer-inner
iterative strategy, every ∆N0 outer iterations (where ∆N0 is somewhat arbitrary but
can be optimized) the so-called “two-node problem” calculation (a spatially-
decoupled NEM calculation spanning two adjoining nodes) is performed for every
interface (for all nodes and in all directions) to provide an improved estimate of the
net surface current at that particular interface. Subsequently, the NEM estimated
net surface currents are used to update (i.e. change) the original FDM diffusion
coupling coefficients. Outer iterations of the FDM based equation are then
continued utilizing the updated FDM coupling coefficients for ∆N0 outer iterations.
The entire process is then repeated. This procedure of updating the FDM couplings
is a convergent technique which progressively forces the FDM equation to yield the
higher-order NEM predicted values of the net surface currents while satisfying the
nodal balance equation, thus yielding the NEM results for the node-average flux
and fundamental mode eigenvalue. The advantages of this technique come in
many forms;
• the storage requirements are minimal because the two-node problem
arrays are re-usable (disposable) at each interface,
• the rate of convergence is nearly comparable to that of the base FDM
algorithm being used,
• the number of iteratively determined unknowns is reduced by a factor of 6
(node flux vs. partial surface current),
• and the simplicity of the algorithm and ease of implementation, compared
to any other nodal technique, is far superior.
In addition, by selecting to not update the coupling coefficients in the non-linear

iterative strategy, the finite difference method (FDM) representation, utilizing the
box scheme, of the few-group neutron diffusion equation results. The implication is
that the model can be utilized to solve either the nodal or FDM representation of
the few-group neutron diffusion equation. In Fig. 43 the overall nested iterative
solution strategy used within RELAP5-3D code for each attempted advancement is
showed.
Two or four energy groups can be utilized, with all groups being thermal groups
(i.e. upscatter exits) if desired. Three, two and one dimensional models can be
utilized. Various core symmetry options are available, including quarter, half and
full core for Cartesian geometry and one-sixth, one-third and full core for hexagonal
geometry. Zero flux, non-reentrant current, reflective and cyclic boundary
conditions are treated.
85
Fig. 43 – Overview of the nested iterative solution strategy
The neutron kinetics subroutines require as input the neutron cross-sections in the
computational nodes of the kinetics mesh. A neutron cross-section model has been
implemented that allows the neutron cross-sections to be parameterized as
functions of RELAP5-3D heat structure temperatures, fluid void fraction or fluid
density, poison concentration, and fluid temperatures. Thus this feature allows
RELAP5-3D to be suitable for RBMK studies, allowing to calculate the graphite
moderator temperature, the coolant density and of course the fuel temperature
feed-back effects.
A flexible coupling scheme between the neutron kinetics mesh and the thermal-
hydraulics spatial mesh has been developed to minimize the input data needed to
specify the neutron cross-sections in terms of RELAP5-3D thermal-hydraulic
variables.
A control rod model has been implemented so that the effect of the initial position
and subsequent movement of the control rods during transients may be taken into
account in the computation of the neutron cross-sections. The control system has
been modified to allow the movement of control rods by control variables.
4.2.4.2. The Cross Section and the feedbacks

The inputs to the kinetics modules in RELAP5-3D consist of neutron cross
sections, boundary conditions, control flags, control data, etc. Neutron cross
sections are needed for each neutron energy group and kinetics node and consist
of the diffusion coefficient (D), absorption cross section (Σa), fission cross section
(Σf), the product of the mean number of secondary neutrons per fission and the
fission cross section (νΣf), and the scattering cross section for scattering into the
neutron energy group from the other neutron energy groups (Σsg‘g). Discontinuity
factors (DFs) for each face of the kinetics nodes are also needed for each energy
group. The user supplies the control information and the boundary conditions as
part of the required input data. The neutron cross sections are computed from a
function selected from a set of built-in functions or a user supplied function whose
86
independent variables are weighted averages of RELAP5-3D hydraulic or heat
structure variables.
There are four built-in neutron cross section functions from which the user can
select or the user may supply his own function in the form of an external
subroutine. The same user specified function is used for the computation of the
neutron cross sections for all kinetics nodes. The coefficients in the first three built-
in cross section functions for each of the cross sections are supplied through user
input and a set of coefficients for the neutron cross sections is called a
composition.
The user defines a number of compositions and specifies which composition is to

be used for the computation of the neutron cross sections in each kinetics node in
the reactor core model. The mapping of compositions to nodes is accomplished
through the use of composition figures. A composition figure specifies the
composition for all of the kinetics nodes in a single axial plane in the reactor model
and composition figures are required for each axial plane in the reactor model. The
user may specify as many composition figures as needed to describe the reactor
and a composition figure may be assigned to more than one axial plane in the
reactor model.
The user specifies the volumes and heat structures to be used in computing the
weighted averages of thermal-hydraulic variables as well as the values of the
weighting factors. A set of volume averaged properties consists of an average fluid
density (or fluid vapor fraction depending on which neutron cross section function is
selected by the user), an average fluid temperature, and an average poison
density, while the average heat structure property is simply the average heat
structure temperature.
The group of volumes and heat structures used to define sets of average thermal-
hydraulic properties is called a zone. One of the built-in cross section functions
uses a single set of average properties in a zone while the other two built-in
neutron cross section functions use multiple sets of average volume and average
heat structure properties in a zone. The user defines a number of zones and
specifies which set of zone average thermal-hydraulic properties is to be used for
the computation of the neutron cross sections in each node of the reactor core
through the use of zone figures. A zone figure specifies which zone is to be used
for the computation of the neutron cross sections for each node in an axial plane of
the reactor. Zone figures are required for each axial plane in the reactor model and
the same zone figure may be specified for more than one axial level. In addition to
the weighted averages of hydraulic and heat structure variables, the neutron cross
section function uses the position of the control rods as another independent
variable.
A control rod model is used to determine the positions of the control rods and to
compute the control rod variable needed by the neutron cross section function on
each axial level in the reactor core. The three neutron cross section functions that
the user may select to compute the neutron cross sections and the control rod
model are described in the following section.
87
4.2.4.3. The GEN neutron cross section model
The last built-in cross section model is the most general built-in cross section
model and was developed [34] as part of the RELAP5-3D code development
project. The GEN cross section function uses a form similar to the HWR
formulation except that the variable portion of the cross section is defined for three
control states, active controlled, driver controlled and uncontrolled states. The GEN
cross section function is given by
(6)
where:
Cfla, d = control fraction for active and driver portions of the control rods in
node l
Σxca,d,u = base cross section of reaction type x for composition c for active
controlled, driver controlled, or uncontrolled state
δΣxca,d,u = variation of cross section for reaction type x for composition c

due the changes in the thermal-hydraulic variables from the base thermal-
hydraulic state for active controlled, driver controlled, and uncontrolled
state
and composition c has been specified for node l.

The variations for the active controlled, driver controlled, and uncontrolled states
are given by:
(7)
where the coefficients a, b, c, d, and e are input separately for the active
controlled, driver controlled, and uncontrolled states.
Therefore, each neutron cross section is the sum of the base cross section and the
change in the cross section:
88
(8)
Where:
Σxlbase = base cross section of type x in node l
δΣxl = change in cross section x in node l due to changes in the thermal-hydraulic

state of the zone to which node l corresponds
Cfl = control fraction in node l
axin = coefficient for change in cross section x of composition n due to changes

in average moderator temperature of volume region i
∆Tmik = change in average moderator temperature in volume region i of zone k,

(Tmik - Tminbase)
Tmik = average moderator temperature in volume region i of zone k
Tminbase = average moderator temperature in volume region i for composition n at

base thermal-hydraulic conditions, e.g. full power steady state
NV = number of volume regions in each zone
bxin = linear coefficient in change of cross section x of composition n due to

changes in average moderator density in volume region i
∆ρmik = change in average moderator density in volume region i of zone k, (ρmik -

ρminbase)
ρminbase = average moderator density in volume region i for composition n at base

thermal-hydraulic state
cxin = quadratic coefficient in change of cross section x of composition n due to

changes in average moderator density in volume region i
dxin = coefficient for change in cross section x of composition n due to changes

in average poison density in volume region i
∆Bik = change in average poison concentration in volume region i of zone k, (Bik -

Binbase)
Binbase = average poison concentration in volume region i for composition n at base

thermal-hydraulic condition
exin = coefficient for change in cross section x of composition n due to changes
89
in average structure temperature of structure region i
∆Tsik = change in the average structure temperature in structure region i of zone k

(Tsik - Tsin base)
∆Tsin base= average structure temperature in structure region i of composition n for

base thermal-hydraulic state
NS = number of structure regions in each zone.
The nodal discontinuity factors are also computed from Equation (6) except that the
thermal-hydraulic variations are identically zero and are therefore not included.
The GEN model subdivides each zone into a number of regions and define
averages of volume and heat structure quantities for each region of a zone. The
number of regions in a zone for the computation of volume average properties may
be different from the number of regions in a zone for the computation of the heat
structure average properties, but the number of volume regions and the number of
heat structure regions is the same for each zone. The region average moderator
temperature, the region average moderator density, the region average poison
density, and the region structure temperature are computed using equations like,
e.g., equation (9):
(9)
where:
Tmik = average moderator temperature in volume region I of zone k
Wmijk = moderator temperature weighting factor for volume j in volume region i of

zone k.
Other average quantities are calculated in a similar manner. The GEN cross
section model can also use either the moderator density or the moderator void
fraction depending upon which option the user specifies.
90
4.3. The Cross Section Processing methodology
4.3.1. Introduction
The execution of the 3D NK calculations requires the development and the use of a
methodology for the cross section data processing. Each 3D NK code has its own
input formalism for the use of these data and different approaches are possible.
The GRNSPG was since, its institution, performing calculations with the most
advanced of these methodologies and was actively participating in International
Research Programs dealing with researches in such field, e.g. [43]. Thus, the PhD
activities for the cross section processing were performed in such a framework.
4.3.2. The State of the Art methodologies
The need for a more accurate method of modelling cross-section variations for off-
nominal core conditions is becoming an important issue with the increased use of
coupled three-dimensional (3-D) neutronics/thermal-hydraulic simulations. In
traditional reactor core analysis for both steady-state and transient calculations of
LWR conventional nuclear power plants, condensed few-group two-dimensional (2-
D) cross-section sets are used as input data.
These cross-section sets are generated by separate database calculations using

characteristic weighting spectra and are parameterized in terms of burn-up and
thermal-hydraulic feedback parameters. Under the real reactor conditions,
especially in transient situations, these spectra change and the 2-D cross-section
modeling based on a parameterization model only approximately describes the
effects of neutron flux distributions, which change in space, time and energy. This
so-called 2-D off-line cross-section generation and modeling constitutes a basic
input data uncertainty affecting the results of coupled 3-D neutronics/thermal-
hydraulic calculations.
Historically, a two-step process is applied in traditional reactor core analysis for

both steady-state and transient applications. The first step in the process is to
calculate few-group cross-sections with different dependencies (i.e. as a function of
burn-up and local feedback parameters) for various regions of a reactor core in 2-D
geometry, employing lattice physics codes such as CASMO [35], [36] and HELIOS
[54]. The second step is to use this cross-section data in a 3-D nodal diffusion code
for determination of different parameters throughout the reactor core. There are
several shortcomings in this approach associated with both cross-section
generation and cross-section modeling. In regard to the cross-section generation
two shortcomings are addressed:
• The use of 2-D lattice physics codes for cross-section generation, based
on the fact that the majority of current lattice physics codes use the
collision probability method (CPM), becomes cumbersome and impractical
in 3-D geometries.
• Current methodology homogenizes representative assemblies assuming
symmetry (reflective) boundary conditions. This approach introduces
significant errors in the determination of neutron flow among assemblies in
91
a real reactor core configuration. The errors are somewhat mitigated by the
use of ad hoc assembly discontinuity factors for conventional reactor core
analysis.
The amount of few-group cross-section data necessary for steady-state, depletion

and transient analysis, is significant. Standard cross-section modeling for coupled
3-D steady state and transient simulations are based on the data generated in the
so-called base and branch calculations using a lattice physics code. The cross-
section history and instantaneous dependence models developed in this way are
based on burn-up and local feedback parameters (i.e. fuel temperature, pressure,
moderator temperature, void, boron concentration). The thermal-hydraulics model
coupled with the neutronics simulator calculates these feedback parameters.
Changing each of the parameters one at a time develops the instantaneous cross-
section dependencies.
A typical dependence of a cross-section on a particular parameter is displayed in
Fig. 44. This shows that over a large range of values the cross-section does not
behave linearly.
1.98E-03
1.96E-03
1.94E-03
1.92E-03
Fission X-Section, (cm-1)
1.90E-03
1.88E-03 Roc=0,555 g/cm3, B=11 ppm

Roc=0,555 g/cm3, B=500 ppm
Roc=0,833 g/cm3, B=11 ppm
1.86E-03 Roc=0,833 g/cm3, B=500 ppm
1.84E-03
1.82E-03
1.80E-03
1.78E-03
0 200 400 600 800 1000 1200 1400 1600 1800
Fuel Temperature, °C
Fig. 44 – Cross Section dependence
More interesting cross-section behavior is shown in Fig. 45. The figure shows the
interdependence of cross-sections when two parameters are varied at once. The
cross-sections generated in this way are called cross-terms, since they are not
dependent on just one parameter, but on all parameters. Such cross-term cross-
sections have to be taken into account in transient analysis since they are actual
points on the curved surface.
92
Fig. 45 – Cross-section interdependence between fuel temperature and moderator
temperature
However, standard methods (currently used in core steady-state, depletion and

transient analysis) such as the polynomial fitting procedure (usually based on
Taylor expansions) do not take these cross-terms into account. Since these
methods utilize no cross-term dependencies on local feedback parameters they are
especially inaccurate for transients in which large departures from nominal
conditions exist.
The typical calculation points necessary to develop the cross-section derivatives

used in the polynomial fitting procedure can be seen from the schematic given in
Fig. 46. For simplicity only two parameters are shown, fuel temperature (TF) and
moderator temperature (TM).
This method uses a cross-section calculation at average conditions, shown as the

black dot, as a reference value. From the reference value parameter perturbations
are performed to develop cross-sections at different conditions, shown by the red
dots. In this method only one parameter is varied at a time, all other parameters
remain at average conditions.
Once the new cross-section is established along with the magnitude of the
individual parameter variation, a derivative can be constructed which is used
directly in the polynomial equation. Using these derivatives along with the average
cross-sections, a cross-section can be calculated at any reactor condition.
93
Fig. 46 – Cross-section calculation points for the polynomial fitting procedure
The most significant problem with this procedure is that it becomes more
inaccurate as the parameter variations get farther away from average conditions.
The areas where the polynomial fitting procedure becomes inaccurate are shown
as the blue hashes in Fig. 47. The increased inaccuracy of the cross-sections
calculated in these regions is more important in transient analysis where parameter
variations extend into this region during a typical transient calculation.
Fig. 47 – Areas of inaccurate cross-section calculation using the polynomial fitting

procedure
The widely used CASMO/SIMULATE cross-section parameterization model

attempts to model the cross-section cross-term dependence involving an
approximate type of cross-section representation [35].
Each cross-section can be evaluated as a summation of base and partial values.

The base cross-sections represent the burn-up dependence (exposure, spectral
94
history and control history) while the partial cross-sections represent the
instantaneous dependence on local feedback parameters. Performing branch
calculations generates the partial cross-sections, where again as with the
polynomial fitting procedure only one feedback parameter is changed for a given
perturbation. The model tries to account for the cross-term dependence by using
separate partial cross-sections for different feedback effects. While the model is an
improvement over the polynomial fitting procedure it is limited to small
perturbations.
To remedy the inaccuracies of the procedures described above, PSU has

developed a sophisticated, unique cross-section representation methodology for 3-
D coupled transient calculations [37]. The method developed at PSU employs not
only the cross-section at average conditions, but also the cross-term cross-
sections, as can be seen in Fig. 48. Cross-term cross-sections are cross-sections
calculated by varying two or more parameters at the same time. The cross-sections
are then tabulated in N-dimensional tables. The N-dimensional tables are then
interpolated for the appropriate cross-section value. The tabulated cross-sections
completely encompass the full range of conditions that may be present during the
initial steady state and the transient. This method is called the Adaptive High-order
Table Look-up Method (AHTLM) [38]. In the AHTLM the user develops an
operating condition box-envelope. This box bounds the expected range of change
of the feedback parameters during both steady-state and transient operation. The
cross-sections are then calculated for the bounding box edges and within the box.
These reference cross-section values are placed in multi-dimensional tables that
are used in a sophisticated table interpolation method that is employed to calculate
the cross-sections.
Fig. 48 – The PSU transient cross-section representation
The advantage of this method is that there is no chance that the calculation of the
cross-section can be outside the bounds set by the user and to involve
extrapolation procedure. In addition, this approach helps to improve the accuracy
of modeling the cross-section variations by avoiding user-calculated coefficients
that could contain errors. This method takes into account the non-linear thermal-
95
hydraulic feedback parameter phenomena that are critical for the accurate
prediction of cross-section behavior.
4.3.3. The Methodology for the RBMK calculations

The RBMK cross section libraries were developed according to the AHTLM
procedures. As reported in section 4.5.2, data were parameterized considering a
relevant set of variables (fuel temperature, moderator temperature, coolant
density), and a their suitable range. Nevertheless, due to the fact that the RELAP5-
3D (NESTLE) did not allowed on-line cross section interpolation (see section
4.2.4.2), several relevant modifications were performed.
A 4D linear interpolation routine was used for calculating at the beginning of the
transient the reference cross section. Cross section variation coefficients were
instead calculated using a Least Square Methods interpolation routine. Thus, the
resulting cross section representation was a standard “polynomial fitting” procedure
(see section 4.5.2.2 for more details). A check of trend of the main cross sections
as function of the main physical parameters confirmed their linear behavior,
allowing the NESTLE GEN Cross Section representation without major
approximations. An example of the main cross sections trends is given in Fig. 49,
Fig. 50.
3.99E-03
3.99E-03
3.99E-03
Thermal Absorption Xsec (cm-1)
3.99E-03
3.98E-03
3.98E-03
3.98E-03
3.98E-03
3.98E-03
3.97E-03
400 500 600 700 800 900 1000 1100
Fuel Temp (K)
Fig. 49 – RBMK Libraries Cross Section – Thermal absorption trend for

composition 29 (see section 4.5.2)
96
1.90E-03
1.88E-03
1.86E-03
Fission Thermal XSec (cm-1)
1.84E-03
1.82E-03
1.80E-03
1.78E-03
1.76E-03
1.74E-03
0 200 400 600 800 1000 1200
Coolant Density (Kg/m3)
Fig. 50 – RBMK Libraries Cross Section – Thermal fission trend for composition 29
(see section 4.5.2)
Nevertheless, the calculated cross section libraries structure, allows a future use of
an on-line interpolation as soon as RELAP5-3D will include this option.
4.4. The coupling methodology

Certain requirements with regard to the coupling of thermal-hydraulic system codes
and neutron kinetics codes should be considered, e.g., [39], [40], [41], [42]. The
objective of these requirements is to provide accurate solutions in a reasonable
amount of CPU time in coupled simulations of detailed operational transient and
accident scenarios.
These requirements are met by the development and implementation of six basic
components of the coupling methodologies:
1) Coupling approach – integration algorithm or parallel processing.
2) Ways of coupling – internal or external coupling.
3) Spatial mesh overlays.
4) Coupled time-step algorithms.
5) Coupling numerics – explicit, semi-implicit and implicit schemes.
6) Coupled convergence schemes
Detailed explanation of the previous points can be found in [43]. It should be noted
that for the execution of the plant SS and transient calculation, RELAP5-3D code
exploits an integration algorithm (i.e., NESTLE code is fully integrated into the
system code) performing an internal coupling. Mesh overlays between NESTLE
and RELAP5-3D TH module are managed through the use of dedicated input deck
cards. The time steps are instead governed by the TH module.
97
4.5. The procedure for code application including the
nodalization
4.5.1. Introduction
The calculation of 3D NK TH transients by a system code like RELAP5-3D code
requires several “auxiliary” steps, or development of cross section libraries and
validation calculations. In Fig. 51 is given the main steps followed for the
calculations performed during this PhD thesis.
Fig. 51 – Codes and Models developed for 3D NK TH RBMK analyses
The firs step concerned the developed of a Smolensk-3 TH nodalization for the
RELAP5 TH code. A very detailed model of the plant was developed,
comprehensive of both the MCC, the SD, part of the SL and with a core TH
description based on 50 FC. Active and passive heat structures were also taken
into account. This model is described in section 4.5.3.
The second step concerned the development of a 3D NK model. This was then
coupled with the full TH hydraulic model of the previous step and with a simplified
one. The 3D NK model and its coupling scheme are described in section 4.5.4 and
4.5.4.1 respectively.
A third (parallel to the second one) step was the development, in collaboration with
the RDFMG of PSU, of suitable RBMK cross section libraries by the HELIOS code.
This model is described in section 4.5.2.
Coupled codes calculations of a specific transient, the FC blockage, prompted the
necessity to perform detailed Monte Carlo simulation of the involved Fuel Cells.
This constituted the fourth step and it was performed developing a MCNP5 model,
described in section 4.5.5.
The fifth and last step was the development by the lattice physics code DRAGON
of fuel cell spectrum-weighted microscopic cross sections libraries for Xenon, in
98
order to simulate by the 3D NK TH RELAP5-3D code the low power transient, i.e.,
a Chernobyl like accident. The DRAGON model is described in section 4.5.6
4.5.2. The HELIOS code model

4.5.2.1. Core geometry
The developed HELIOS RBMK fuel assembly (cell) model was a single cell with
reflective boundary conditions, with which depletion and branch calculations were
performed. In addition a HELIOS RBMK non-fuel assembly (cell) model was
developed. In this case the cell was surrounded by eight fuel cells (color set
calculations).
All of the fuel cells were assumed to have the same burnup. The graphite
temperature was assumed to be uniform in the whole color set (macro cell). Only
branch calculations were performed with this model by varying coolant density in
the non-fuel cell, coolant density in surrounding fuel cells, and graphite
temperature in the non-fuel cell
The HELIOS sophisticated methods address properly all of these issues and
HELIOS can be used for the RBMK cross-section generation and modeling. There
was no need for additional ad-hoc corrections of the macroscopic cross-sections or
burnup as it is generally done for RBMK with the cross-sections generated by other
codes (e.g., WIMS, see section 4.1.2).
Radially, the core was divided into square cells with a pitch 25.0 cm, each
corresponding to one channel, plus a four rows of radial reflector blocks of the
same size (shaded area in Fig. 35, Fig. 55 and Fig. 56).
As reported in Chapter 2, there are a total of 2488 channels, 1570 fuel channels,
314 non-fuel channels, and 604 reflector channels. Axially, the reactor core was
divided into 10 layers with a height (starting from the bottom) of 70.0 cm, adding up
to a total active core height of 700 cm. Both upper and lower axial reflectors have a
thickness of 30 cm.
The axial nodalization scheme accounts for material changes in the fuel design
and for the exposure variations. Zero flux boundary conditions were specified on
outer reflector surface for both radial and axial reflectors. The geometric model for
the developed elementary cells is respectively shown in and in Fig. 52 for the Fuel-
Cell and Fig. 53 for the Non-Fuel Cell.
99
Graphite
Zirconium Alloy (Э125)
UO2
Coolant
Fig. 52 – HELIOS modeling of RBMK FC cell
Graphite
Aluminum Alloy
CR Channel
Graphite
UO2
Coolant
Zirconium Alloy
Absorber
Fig. 53 – HELIOS modeling of RBMK MCR cell
The number of axial mesh used in the HELIOS core modeling is given in Fig. 54. A
detailed description of the TH coupled part can be found further below in this
document.
100
12 Top Reflector
CORE MESHING
(10 x 0.7 m)
2 to 11
1 Bottom Reflector
Fig. 54 – Core neutron kinetic mesh
Fuel channels with different 235U enrichments are present in the core. 48 fuel
channels have 2.0% enriched fuel and 1522 fuel channels have 2.4% enriched
fuel. HELIOS model took into account these differences and the reactor fuel
burnup map. The three different types of control rods were also modeled, or the
155 manual control rods with its three parts – absorber, connecting rod, and
displacer (made of graphite) –, the shortened control rods (32 in the core) with its
two parts – absorber and displacer – and the safety control rod type (24 in the
core) with their absorber. The rest of the non-fuel channels modeled were axial
detector channels (12 in the core), additional absorber (90 in the core), and water
column (1 in the core). Each of these different fuel channels, were identified by a
number. Their physical characteristics (i.e., enrichment, burnup, fuel/non fuel
channel) are reported in Tab. 16, Tab. 17. Their location in the reactor core lattice
are instead given in Fig. 55, Fig. 56.
Tab. 16 – Definition of fuel channel types
Channel
Enrichment, % w/o Exposure, MWd/kgU
type
1 2.0 6.75
2 2.0 14.25
3 2.0 14.75
4 2.0 15.25
5 2.0 15.75
6 2.0 16.25
101
Tab. 16 (cont.) – Definition of fuel channel types
Channel
type
7 2.0 16.75
8 2.0 17.25
9 2.0 17.75
10 2.0 18.25
11 2.0 19.75
12 2.0 20.25
13 2.4 0.25
14 2.4 0.75
15 2.4 1.25
16 2.4 1.75
17 2.4 2.25
18 2.4 2.75
19 2.4 3.25
20 2.4 3.75
21 2.4 4.25
22 2.4 4.75
23 2.4 5.25
24 2.4 5.75
25 2.4 6.25
26 2.4 6.75
27 2.4 7.25
28 2.4 7.75
29 2.4 8.25
30 2.4 8.75
31 2.4 9.25
32 2.4 9.75
33 2.4 10.25
34 2.4 10.75
35 2.4 11.25
36 2.4 11.75
37 2.4 12.25
38 2.4 12.75
39 2.4 13.25
40 2.4 13.75
41 2.4 14.25
42 2.4 14.75
43 2.4 15.25
44 2.4 15.75
45 2.4 16.25
46 2.4 16.75
47 2.4 17.25
48 2.4 17.75
102
Tab. 16 (cont.) – Definition of fuel channel types
Channel
type
49 2.4 18.25
50 2.4 18.75
51 2.4 19.25
52 2.4 19.75
53 2.4 20.25
54 2.4 20.75
55 2.4 21.25
56 2.4 21.75
57 2.4 22.25
58 2.4 22.75
59 2.4 23.25
60 2.4 23.75
61 2.4 24.25
Tab. 17 – Definition of non-fuel channel types
Exposure of
Enrichment of
Channel surrounding
Description surrounding Fuel
type Fuel Channels,
Channels, % w/o
MWd/kgU
62 Axial Detector Channel 2.4 10.00
63 Manual Control Rods 2.4 10.00
64 Shortened Control Rods 2.4 10.00
65 Safety Control Rods 2.4 10.00
66 Additional Absorber 2.4 10.00
67 Water Column 2.4 10.00
68 Radial Reflector 2.4 10.00
103
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
1 70 70 70 70 70 1
2 70 70 70 70 70 69 69 69 69 69 2
3 70 70 69 69 69 69 69 69 69 69 69 69 3
4 70 70 69 69 69 69 69 68 68 68 68 68 68 68 4
5 70 70 69 69 69 69 68 68 68 27 22 39 34 38 34 28 5
6 70 69 69 69 69 68 68 34 42 3 63 40 52 31 63 26 66 6
7 70 69 69 69 68 68 31 28 51 13 29 18 45 36 46 22 13 36 7
8 70 69 69 69 68 20 64 35 15 26 63 51 33 47 64 15 52 34 65 8
9 70 69 69 69 68 25 30 26 19 55 33 16 20 24 16 21 31 38 16 36 9
10 70 69 69 69 68 33 63 13 66 47 65 51 66 42 63 39 66 48 63 25 66 10
11 70 69 69 69 68 34 27 50 19 51 32 23 40 31 30 36 21 13 26 20 34 52 11
12 70 69 69 69 68 29 63 16 40 33 63 30 42 45 62 14 35 30 63 46 39 50 63 12
13 70 69 69 69 68 33 21 5 13 47 16 47 24 13 27 46 44 27 20 14 33 19 33 29 13
14 70 69 69 68 2 63 46 37 19 63 54 66 40 63 35 66 51 63 48 66 5 63 48 66 14
15 70 69 69 68 20 30 15 51 9 52 26 39 28 17 28 23 29 25 41 20 16 24 31 34 19 15
16 70 69 69 68 64 39 48 22 63 39 37 21 64 13 10 46 63 55 61 28 64 46 54 40 65 16
17 70 69 69 68 26 5 19 43 29 43 24 45 28 43 25 21 37 27 15 37 50 22 31 39 18 34 17
18 70 69 69 68 34 22 7 65 21 66 14 63 26 66 16 65 49 66 51 63 42 66 47 63 23 66 18
19 70 69 69 68 27 4 14 52 40 3 50 43 24 18 51 36 41 19 5 29 25 18 27 31 35 34 47 19
20 70 69 69 68 23 19 63 29 33 48 62 29 34 40 63 44 7 29 63 14 43 37 63 16 47 22 62 20
21 70 69 69 68 34 49 15 22 53 36 19 54 16 27 19 25 56 35 44 7 33 60 26 30 5 49 42 21
22 70 69 68 28 63 29 66 44 63 46 66 23 63 44 66 47 63 23 60 19 63 53 66 39 63 22 66 22
23 70 69 68 15 50 24 34 18 33 25 19 30 43 49 38 23 45 17 53 33 41 51 36 45 14 53 50 23
24 70 69 69 68 32 41 18 64 49 52 15 63 55 38 20 64 47 39 34 63 21 45 47 64 19 37 27 65 24
25 70 69 69 68 27 13 48 21 13 36 31 20 32 28 46 24 30 53 55 27 16 36 22 33 29 47 20 55 25
26 70 69 69 68 19 63 25 66 17 63 51 66 47 63 31 66 47 63 23 66 56 63 50 66 22 63 26 66 26
27 70 69 69 68 29 49 18 22 36 5 27 19 29 22 15 42 37 47 18 33 29 47 34 53 38 50 36 45 27
28 70 69 69 68 18 36 38 65 37 33 13 63 15 38 47 65 44 43 52 62 41 22 47 65 29 29 13 63 28
29 70 69 69 68 51 51 45 23 16 47 40 26 7 44 24 46 54 25 29 20 25 34 67 50 17 49 32 46 29
30 70 69 69 68 17 63 18 66 32 63 56 38 59 63 45 66 16 63 46 66 53 63 19 66 24 63 52 66 30
31 70 69 69 68 27 48 24 40 28 48 28 55 37 47 16 22 55 40 18 25 53 25 45 29 47 30 36 42 31
32 70 69 69 68 34 44 19 64 21 13 47 63 38 8 28 64 48 44 36 63 14 31 18 64 20 14 45 65 32
33 70 69 69 68 48 25 14 41 26 38 34 49 32 13 24 18 26 56 32 26 36 53 27 15 27 36 46 35 33
34 70 69 68 18 63 45 66 48 63 17 66 21 63 27 66 30 63 21 66 22 63 54 66 53 63 47 66 34
35 70 69 68 2 13 24 50 21 45 44 27 6 42 23 45 20 47 36 29 17 42 24 45 26 20 34 43 35
36 70 69 69 68 33 44 63 30 39 52 62 39 14 38 63 55 32 15 63 36 41 34 63 31 48 44 62 36
37 70 69 69 68 20 6 52 59 48 41 21 27 18 27 31 38 24 40 45 20 14 29 54 22 36 14 47 37
38 70 69 69 68 36 16 49 45 65 28 66 50 63 47 66 53 65 55 66 27 63 55 66 43 63 31 66 38
39 70 69 69 68 21 33 46 15 22 34 25 29 22 31 29 18 7 36 23 42 35 16 33 21 38 26 39
40 70 69 69 68 1 64 49 43 50 63 19 33 51 64 14 26 51 63 15 53 47 64 45 51 35 65 40
41 70 69 69 68 27 49 26 16 49 28 40 38 26 20 52 36 29 25 37 26 51 25 34 20 44 41
42 70 69 69 68 33 41 63 53 66 39 63 40 66 51 63 24 66 42 63 45 66 20 63 51 66 42
43 70 69 69 68 30 26 22 28 45 42 43 41 51 29 18 33 37 15 24 49 33 14 32 44 43
44 70 69 69 69 68 16 53 63 21 47 36 63 34 45 39 62 20 43 36 63 38 25 54 63 44
45 70 69 69 69 68 19 47 31 45 29 47 25 50 41 50 28 24 19 52 30 19 38 56 45
46 70 69 69 69 68 40 55 63 52 66 53 65 20 66 50 63 14 66 44 63 46 60 46
47 70 69 69 69 68 31 18 38 31 24 16 58 49 26 37 53 20 16 39 34 42 47
48 70 69 69 69 68 24 28 64 37 31 28 63 36 53 31 64 50 24 43 65 48
49 70 69 69 69 68 42 29 44 44 50 18 31 24 60 20 14 34 45 20 49
50 70 69 69 69 68 68 38 51 5 37 48 63 9 31 49 63 51 66 50
51 70 69 69 69 69 68 68 44 23 34 18 45 37 39 32 44 17 51
52 70 70 69 69 69 69 68 68 68 22 32 40 35 44 38 42 52
53 70 70 69 69 69 69 69 68 68 68 68 68 68 68 53
54 70 70 69 69 69 69 69 69 69 69 69 69 54
55 70 70 70 70 70 69 69 69 69 69 55
56 70 70 70 70 70 56
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Fig. 55 – Two-dimensional channel type map (core left part)
104
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
1 70 70 70 70 70 1
2 69 69 69 69 69 70 70 70 70 70 2
3 69 69 69 69 69 69 69 69 69 69 70 70 3
4 68 68 68 68 68 68 68 69 69 69 69 69 70 70 4
5 35 13 29 45 29 22 26 68 68 68 69 69 69 69 70 70 5
6 49 63 18 36 39 63 44 37 32 36 68 68 69 69 69 69 70 6
7 51 39 47 37 48 15 19 54 30 40 34 25 68 68 69 69 69 70 7
8 40 44 22 64 26 37 51 63 18 46 13 64 30 26 68 69 69 69 70 8
9 17 54 31 46 52 34 24 13 36 34 25 37 39 47 16 68 69 69 69 70 9
10 21 63 44 66 14 63 45 66 39 65 21 66 18 63 46 26 68 69 69 69 70 10
11 31 26 34 23 53 22 30 34 47 47 44 47 52 24 13 40 30 68 69 69 69 70 11
12 36 44 48 63 7 39 50 62 47 51 17 63 32 37 49 63 18 38 68 69 69 69 70 12
13 53 50 30 17 25 15 42 39 5 54 25 45 52 16 45 53 24 44 36 68 69 69 69 70 13
14 22 63 47 66 42 63 26 44 40 63 20 66 19 63 50 37 50 63 51 20 68 69 69 70 14
15 44 36 44 4 26 21 35 56 22 40 15 41 27 35 31 43 18 36 31 34 28 68 69 69 70 15
16 40 29 19 64 17 40 9 63 52 48 53 64 50 16 54 63 25 50 52 64 23 68 69 69 70 16
17 14 46 51 22 32 28 32 54 39 35 33 27 24 32 23 29 13 42 43 31 17 30 68 69 69 70 17
18 25 63 47 66 55 63 18 66 33 65 24 66 46 63 40 66 22 65 52 66 3 2 68 69 69 70 18
19 44 23 53 24 42 30 25 36 23 40 36 42 19 36 46 49 41 48 25 20 32 15 21 68 69 69 70 19
20 15 35 32 63 56 40 54 63 49 46 54 63 33 34 13 62 28 40 44 63 27 40 47 68 69 69 70 20
21 49 38 15 21 31 21 25 21 28 38 27 40 23 50 18 43 16 47 30 45 20 49 42 68 69 69 70 21
22 26 63 32 66 45 63 49 66 54 63 48 66 44 63 26 66 24 63 50 66 24 63 21 33 68 69 70 22
23 18 13 21 38 14 43 17 37 32 45 18 23 14 33 53 20 13 30 44 14 20 15 51 15 68 69 70 23
24 43 44 46 64 46 36 29 63 49 38 50 64 48 37 30 63 52 47 24 64 48 27 37 6 68 69 69 70 24
25 33 18 25 50 30 56 25 50 32 23 28 16 27 47 24 52 16 31 35 19 3 55 22 34 68 69 69 70 25
26 50 63 46 66 24 63 15 66 43 63 54 66 32 63 32 66 43 63 26 66 45 63 42 2 68 69 69 70 26
27 31 26 20 46 17 50 20 49 25 41 28 40 17 49 18 45 34 42 4 17 27 34 20 37 68 69 69 70 27
28 22 40 35 65 43 25 31 62 42 44 43 65 40 44 35 63 23 41 13 65 43 13 51 8 68 69 69 70 28
29 37 7 17 22 36 60 32 43 16 53 37 41 45 19 31 50 28 38 25 41 37 27 34 17 68 69 69 70 29
30 25 63 48 66 53 63 47 60 37 63 22 66 48 63 25 66 22 63 48 66 48 63 52 37 68 69 69 70 30
31 32 14 44 37 23 38 22 58 49 28 17 38 27 44 20 13 34 6 23 3 25 48 21 36 68 69 69 70 31
32 22 39 42 64 30 36 56 63 16 53 47 64 21 33 38 63 50 18 13 64 18 31 5 40 68 69 69 70 32
33 16 55 44 21 45 20 26 45 33 55 23 15 29 54 15 53 36 40 29 23 28 53 37 4 68 69 69 70 33
34 47 63 45 66 52 63 51 66 26 63 41 66 44 63 20 66 22 63 16 66 17 63 19 37 68 69 70 34
35 23 5 17 21 28 50 35 23 19 42 28 20 25 51 41 34 17 28 34 47 30 51 6 30 68 69 70 35
36 14 38 31 63 35 40 18 63 33 51 56 63 40 40 45 62 51 12 24 63 25 40 36 68 69 69 70 36
37 19 52 45 15 27 55 30 43 39 38 19 45 26 54 41 38 7 15 46 47 13 30 23 68 69 69 70 37
38 27 63 39 66 46 63 21 66 20 65 47 66 16 63 49 44 39 65 21 66 37 63 31 68 69 69 70 38
39 35 22 46 18 25 42 33 37 49 44 48 33 50 26 36 53 18 43 46 24 33 28 68 69 69 70 39
40 54 40 26 64 15 52 57 63 45 54 30 64 22 44 20 63 52 44 32 64 18 23 68 69 69 70 40
41 25 17 45 35 22 27 34 18 14 24 48 39 29 13 30 49 32 13 52 26 29 68 69 69 70 41
42 49 63 24 66 44 63 52 66 46 63 44 66 32 63 20 66 21 63 46 14 38 68 69 69 70 42
43 14 40 19 57 33 41 37 34 29 34 26 52 29 45 27 43 15 52 40 40 68 69 69 70 43
44 25 35 36 63 54 7 17 62 50 14 19 63 39 33 52 63 22 30 34 68 69 69 69 70 44
45 31 55 26 32 25 32 35 22 38 33 23 52 17 45 47 14 50 39 68 69 69 69 70 45
46 19 63 43 66 47 63 51 66 6 65 27 66 23 63 53 32 30 68 69 69 69 70 46
47 38 42 21 16 22 18 27 31 19 40 49 38 51 35 40 35 68 69 69 69 70 47
48 42 7 52 64 50 38 51 63 47 15 54 64 24 37 17 68 69 69 69 70 48
49 49 33 23 31 23 26 11 24 52 27 45 32 28 31 68 69 69 69 70 49
50 49 63 47 50 46 63 46 48 41 63 40 19 68 68 69 69 69 70 50
51 43 23 41 15 40 35 49 36 16 32 68 68 69 69 69 69 70 51
52 32 42 45 48 44 44 43 68 68 68 69 69 69 69 70 70 52
53 68 68 68 68 68 68 68 69 69 69 69 69 70 70 53
54 69 69 69 69 69 69 69 69 69 69 70 70 54
55 69 69 69 69 69 70 70 70 70 70 55
56 70 70 70 70 70 56
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
Fig. 56 – Two-dimensional channel type map (core right part)
105
Thus, seventy channel types were identified within the core geometry. Each
channel type is then defined by several “compositions”, according to its material
properties (due to changes in the fuel design) and burn-up. The axial locations of
compositions for each fuel type is shown in Tab. 18. The normalized axial exposure
distribution for the different fuel types is given in Tab. 19.
4.5.2.2. The Cross-section library

The required set of 2-group macroscopic cross sections for fuel channels consists
of the following seven types of cross-sections:
• diffusion coefficient, fast group (D1),
• diffusion coefficient, thermal group (D2),
• removal cross-section, fast group (Sa1+Ss12),
• absorption cross-section, thermal group (Sa2),
• neutron generation cross-section, fast group (nSf1),
• neutron generation cross-section, thermal group (nSf2),
• neutron transfer cross-section from fast to thermal group (Ss12).
Separate cross-sections were calculated for upper and lower parts of fuel
assemblies. For the non-fuel channels the following types of 2-group cross-
sections were obtained:
• diffusion coefficient, fast group (D1),
• diffusion coefficient, thermal group (D2),
• removal cross-section, fast group (Sa1+Ss12),
• absorption cross-section, thermal group (Sa2),
• neutron transfer cross-section from fast to thermal group (Ss12).
Also, a complete set of assembly discontinuity factors (ADF) were calculated. All
these cell parameters were expressed as a function of the fuel temperature,
moderator temperature, and coolant density for each composition. The group
inverse neutron velocities were also provided for each composition.
The expected range of the transients was covered by the selection of an adequate
range for the independent variables shown in Tab. 20.
Dependence of the cross-sections on the above variables was specified through a

three-dimensional look-up table. Each composition was assigned to a cross-section
set containing separate tables for the diffusion coefficient, cross-sections, and
ADF, with each point in the table representing a possible core state. A linear
interpolation scheme could be used to obtain the appropriate total cross-sections
from the tabulated ones based on the reactor conditions being modeled (see Fig.
57).
106
Fig. 57 – Interpolation Scheme for Cross Sections parameters
For this purpose, a 4D Linear Interpolation Routine (lint4D) implemented in a

FORTRAN 90 program was used. Finally, after the calculation of the Cross Section
Variation Coefficients (i.e. with the Least Square Method) too, it was possible to
implement in the 3D NK code all the Cross Section related information, e.g. using a
Cross Section representation like that one reported in Equation (10):
∂Σ ∂Σ ∂Σ
Σ(Tf , Tgr , D) = Σ ref + ∆Tf + ∆Tgr + ∆D
∂Tf ∂Tgr ∂D
( 10 )
where:
Sref = Reference Cross Section
Tf = Fuel Temperature
Tgr = Graphite Temperature
D = coolant density
Tab. 21 and Fig. 58 shows in details the definition of a cross-section table

associated with a composition. All cross-section sets are assembled into a cross-
section library. The cross-sections were provided in separate libraries for rodded
(nemtabr) (absorber part of the control rods), unrodded compositions (nemtab)
(fuel channels, displacer parts of the control rods, water column, additional
absorber, and axial detector channels) and nemtab (connecting rod of the manual
control rods). The format of each library was as follows:
• The first line of data was used to show the number of data points used for
the independent thermal-hydraulic (T-H) parameters. These parameters
included fuel temperature, moderator temperature, coolant density
• Each cross-section set is in the order shown in Tab. 21. Each table is in
the format described in Fig. 58. First, the values of the independent
thermal-hydraulic parameters used to specify that particular set of cross-
107
sections are listed, followed by the values of the cross-sections. Finally,
the group inverse neutron velocities complete the data for a given cross-
section set.
• The dependence of the reflector and of CPS cross-section on thermal-
hydraulic parameters is also modeled. This is because the reflector and
CPS cross-sections are generated by performing lattice physics transport
calculations, including the next fuel region. Therefore, the feedback
parameters should be taken from the neighboring fuel region.
Tab. 18 – Composition numbers in axial layers for each channel type
Bottom
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
1 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671
2 1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151 161 171 181 191 201 211 221
3 2 12 22 32 42 52 62 72 82 92 102 112 122 132 142 152 162 172 182 192 202 212 222
4 3 13 23 33 43 53 63 73 83 93 103 113 123 133 143 153 163 173 183 193 203 213 223
5 4 14 24 34 44 54 64 74 84 94 104 114 124 134 144 154 164 174 184 194 204 214 224
6 5 15 25 35 45 55 65 75 85 95 105 115 125 135 145 155 165 175 185 195 205 215 225
7 6 16 26 36 46 56 66 76 86 96 106 116 126 136 146 156 166 176 186 196 206 216 226
8 7 17 27 37 47 57 67 77 87 97 107 117 127 137 147 157 167 177 187 197 207 217 227
9 8 18 28 38 48 58 68 78 88 98 108 118 128 138 148 158 168 178 188 198 208 218 228
10 9 19 29 39 49 59 69 79 89 99 109 119 129 139 149 159 169 179 189 199 209 219 229
11 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230
12 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675
Top
Bottom
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46
671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671
231 241 251 261 271 281 291 301 311 321 331 341 351 361 371 381 391 401 411 421 431 441 451
232 242 252 262 272 282 292 302 312 322 332 342 352 362 372 382 392 402 412 422 432 442 452
233 243 253 263 273 283 293 303 313 323 333 343 353 363 373 383 393 403 413 423 433 443 453
234 244 254 264 274 284 294 304 314 324 334 344 354 364 374 384 394 404 414 424 434 444 454
235 245 255 265 275 285 295 305 315 325 335 345 355 365 375 385 395 405 415 425 435 445 455
236 246 256 266 276 286 296 306 316 326 336 346 356 366 376 386 396 406 416 426 436 446 456
237 247 257 267 277 287 297 307 317 327 337 347 357 367 377 387 397 407 417 427 437 447 457
238 248 258 268 278 288 298 308 318 328 338 348 358 368 378 388 398 408 418 428 438 448 458
239 249 259 269 279 289 299 309 319 329 339 349 359 369 379 389 399 409 419 429 439 449 459
240 250 260 270 280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460
675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675
Top
Bottom
47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70
671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671 671
461 471 481 491 501 511 521 531 541 551 561 571 581 591 601 611 621 631 641 651 661 672 673 674
462 472 482 492 502 512 522 532 542 552 562 572 582 592 602 612 622 632 642 652 662 672 673 674
463 473 483 493 503 513 523 533 543 553 563 573 583 593 603 613 623 633 643 653 663 672 673 674
464 474 484 494 504 514 524 534 544 554 564 574 584 594 604 614 624 634 644 654 664 672 673 674
465 475 485 495 505 515 525 535 545 555 565 575 585 595 605 615 625 635 645 655 665 672 673 674
466 476 486 496 506 516 526 536 546 556 566 576 586 596 606 616 626 636 646 656 666 672 673 674
467 477 487 497 507 517 527 537 547 557 567 577 587 597 607 617 627 637 647 657 667 672 673 674
468 478 488 498 508 518 528 538 548 558 568 578 588 598 608 618 628 638 648 658 668 672 673 674
469 479 489 499 509 519 529 539 549 559 569 579 589 599 609 619 629 639 649 659 669 672 673 674
470 480 490 500 510 520 530 540 550 560 570 580 590 600 610 620 630 640 650 660 670 672 673 674
675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675 675
Top
108
Tab. 19 – Normalized axial exposure
Axial
layer 2 3 4 5 6 7 8 9 10 11
#
Fuel
0.758 1.036 1.087 1.118 1.154 1.160 1.113 1.029 0.915 0.629
2.0%
Fuel
0.756 1.076 1.126 1.143 1.170 1.164 1.103 1.006 0.878 0.577
2.4%
Tab. 20 – Range of independent variables
Rho Rho
T fuel T moderator T fuel T moderator
coolant coolant
(K) (K) (K) (K)
(kg/m3) (kg/m3)
557.00 557.00 10.00 557.00 557.00 500.00
850.00 557.00 10.00 850.00 557.00 500.00
1030.00 557.00 10.00 1030.00 557.00 500.00
557.00 700.00 10.00 557.00 700.00 500.00
850.00 700.00 10.00 850.00 700.00 500.00
1030.00 700.00 10.00 1030.00 700.00 500.00
557.00 856.00 10.00 557.00 856.00 500.00
850.00 856.00 10.00 850.00 856.00 500.00
1030.00 856.00 10.00 1030.00 856.00 500.00
557.00 1060.00 10.00 557.00 1060.00 500.00
850.00 1060.00 10.00 850.00 1060.00 500.00
1030.00 1060.00 10.00 1030.00 1060.00 500.00
557.00 557.00 100.00 557.00 557.00 744.00
850.00 557.00 100.00 850.00 557.00 744.00
1030.00 557.00 100.00 1030.00 557.00 744.00
557.00 700.00 100.00 557.00 700.00 744.00
850.00 700.00 100.00 850.00 700.00 744.00
1030.00 700.00 100.00 1030.00 700.00 744.00
557.00 856.00 100.00 557.00 856.00 744.00
850.00 856.00 100.00 850.00 856.00 744.00
1030.00 856.00 100.00 1030.00 856.00 744.00
557.00 1060.00 100.00 557.00 1060.00 744.00
850.00 1060.00 100.00 850.00 1060.00 744.00
1030.00 1060.00 100.00 1030.00 1060.00 744.00
557.00 557.00 300.00 557.00 557.00 998.00
850.00 557.00 300.00 850.00 557.00 998.00
1030.00 557.00 300.00 1030.00 557.00 998.00
557.00 700.00 300.00 557.00 700.00 998.00
850.00 700.00 300.00 850.00 700.00 998.00
1030.00 700.00 300.00 1030.00 700.00 998.00
557.00 856.00 300.00 557.00 856.00 998.00
850.00 856.00 300.00 850.00 856.00 998.00
1030.00 856.00 300.00 1030.00 856.00 998.00
557.00 1060.00 300.00 557.00 1060.00 998.00
850.00 1060.00 300.00 850.00 1060.00 998.00
1030.00 1060.00 300.00 1030.00 1060.00 998.00
109
Tab. 21 – Key to macroscopic cross-section tables
TF1 TF2 TF3 TM1 TM2 Where:

TM3 TM4 ρC1 ρC2 ρC3 TF – fuel temp.
ρC4 ρC5 ρC6 Σ1(f1,m1,c1) Σ2(f2,m1,c1) [K]
Σ3(f3,m1,c1) Σ4(f1,m2,c1) Σ5(f2,m2,c1) Σ6(f3,m2,c1) Σ7(f1,m3,c1) TM – moderator
temp. [K]
Σ8(f2,m3,c1) Σ9(f3,m3,c1) Σ10(f1,m4,c1) Σ11(f2,m4,c1) Σ12(f3,m4,c1)
ρC – coolant
Σ13(f1,m1,c2) Σ14(f2,m1,c2) Σ15(f3,m1,c2) Σ16(f1,m2,c2) Σ17(f2,m2,c2)
density [kg/m3]
Σ18(f3,m2,c2) Σ19(f1,m3,c2) …… f – same as TF
…… Σ71(f2,m4,c6) Σ72(f3,m4,c6) m – same as TM
c – same as ρC
*************************************************************************
NEM – Cross-Section Table Input
*
* T Fuel T Mod. Rho Cool. CXe
3 4 6 0
*
*********** X-Section Set #
#
*************************************************************************
Group No. 1
*
************* Diffusion Coefficient Table
*
************* Absorption X-Section Table
*
************* Fission X-Section Table
*
************* Nu-Fission X-Section Table
*
************* Scattering From Group 1 to 2 X-Section Table
*
************* Assembly Discontinuity Factors Table
*
*************************************************************************
Group No. 2
*
************* Diffusion Coefficient Table
*
************* Absorption X-Section Table
*
************* Fission X-Section Table
*
************* Nu-Fission X-Section Table
*
************* Assembly Discontinuity Factors Table
*
*************************************************************************
************* Inv. Neutron Velocities
*
Fig. 58 – Macroscopic cross-section table’s structure
110
4.5.3. The RELAP5-3D TH nodalization
The detailed description of the RELAP5 nodalization developed for Smolensk-3
NPP is presented in [44]. An idea of the nodalization can be derived from Fig. 59 to
Fig. 63. Two models of the NPP were developed, one representing the NPP up to
the turbine, the other one (simplified), representing only the MCC from the MCPs
exit up to the SD inlet. The latter one was used for 3D NK TH analyses were it was
not necessary to model the whole primary circuit (e.g., CR / CR group withdrawal).
All the nodalizations are divided into two parts identified as right and left side. Both
sides are modeled in the same way, i.e., the main components (pumps, pipes,
steam drums) are identical. An higher resolution of the core is achieved for the right
side (47 channels modeled versus 8 for the left side), thus the right side is
considered as the damaged one during the safety analyses.
As it can be seen in Fig. 59, all the main hydraulics components of the plant are
modeled. Starting from the Steam Drums (SD), both two are represented by a set
of branches and pipes, Fig. 61 in order to reproduce the stratification that ensures
the gravity driven steam separation. The water (in the lower part) and the steam (in
the upper part) zones of the SD are linked each other by a water and steam bridge.
Two pipes joined to the top of each SD drive the steam to the steam ring where the
steam lines are connected. All the safety valves (e.g. BRU-K, MSV and ICV) are
considered in the input deck. The nozzle of the feed-water system is represented
by the jet mixer component to simulate the suction effect of these devices.
The two feed-water systems (one per each SD) are simulated by a time dependent
volume and a time dependent junction with a nominal imposed mass flow rate. The
jet mixer components are joined to two pipes that simulate the downcomer. These
two pipes are connected to four branches that simulate the suction header, Fig. 62.
Four pipes supply the water from the suction header to four pumps connected via
other pipe components to a pressure header schematized by eight branches. 22
GDH are simulated by 22 pipes (8 pipes for the left side) and, at the end of each of
them, one branch is placed. 47 pipes (8 pipes for the left side) divided into 59
volumes represent the water lines and the fuel channels, up to the connection with
the steam water line, with different scaling factor ranging from 1 up to 34 (from 62
to 134, for the left side). Additional 47 pipes (8 pipes for the left side) simulate the
steam water lines connected to the lower part of the SD, Fig. 63.
The additional absorbers are simulated with a unique pipe (the same subdivision of
the core channel has been used) together with its water line and steam water line.
One separate circuit is reproduced to simulate the control rod and radial reflector
cooling channels (one hydraulic channel for the CR cooling and one for the
reflector).
All the passive components have the relative mono-dimensional heat structure
divided into five radial mesh points and they are considered insulated on one side.
The active heat structures that simulate the fuel are divided into 12 radial mesh
points and into 20 axial parts. They are, on the left side, thermally coupled with the
corresponding hydraulic volume of the core channel, on the right side insulated.
111
Fig. 59 – Sketch of the reference RELAP5-3D TH nodalization of Smolensk-3 (only left MCC showed)
112
Fig. 60 – Sketch of the simplified RELAP5-3D TH nodalization (Left and Right MCC)
113
Fig. 61 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of FW and
SL systems.
114
Fig. 62 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of down-
comers up to the water lines.
Fig. 63 – RELAP5-3D model of Smolensk 3 NPP: nodalization sketch of core, FC

exit lines and CPS cooling system.
115
The thermal structures used to simulate the pressure tube, the graphite rings and
the graphite bricks are divided into 12 radial meshes and in 24 axial parts in order
to consider the axial reflector. They are thermally coupled with the core channel at
the inner side and with the corresponding pipe of the reactor cavity at the outer
side.
The ECCS included in the model of the Smolensk 3 NPP are, Fig. 64:
• Main-subsystem 1 and 2 including two trains of accumulators. Each train is
composed by 6 accumulators. The two trains are connected to the headers
to inject water in the damaged side of the core.
• Main-subsystem 3: this system injects water in the damaged side of the
core by connecting to the main feed water pumps. The feed water pumps
take water from the deareators.
• Long term sub-system damaged core side 1, 2 and 3: the system is
constituted by three trains. Each train includes two pumps injecting in the
damaged side of the core the water from the PSP in the ASL.
• Long term sub-system undamaged core side 1, 2 and 3: the system is
constituted by three trains. Each train includes one pump. The pumps
inject in the undamaged side of the core the water from the clean
condensate tank.
The twelve accumulators of the Main sub-system 1 and 2 are schematized. In the
model the accumulators injection is stopped when the level is low to avoid nitrogen
injection in the MCC. The model makes possible to manage separately injection
capability of any accumulators, if needed, simply changing the intervention logic
that is separately implemented for each accumulator.
The Main sub-system 3 is schematized with a “tmdpjun” component simulating the

main feed water pumps and a “tmdpvol” component simulating the thermal
hydraulic condition of the water in the deaerators. The “tmdpjun” component makes
possible a specific simulation of the mass flow-rate of the main feed water pumps
for this system separated by the simulation of the feed water pumps of the MCC;
this schematization has been preferred to avoid any “complicated interaction” with
the model of the pumps used in the MCC. In addition this model makes available to
change easily the mass flow-rate trend of the pumps to take into account the
effects of LOOP and to control the integral mass flow-rate through the pumps: this
value is used to check if the deaerators contain water to be injected.
The three trains of the Long term sub-system for the damaged core side are
separately modeled. Each model includes a “tmdpjun” component connected whit
the PSP simulating the two pumps of each train. Each train can be isolated if
needed to investigate transient involving ECCS failure. The three train of the Long
term sub-system for the undamaged core side are separately modeled. Each
model includes a “tmdpjun” component simulating the pump of each train. Each
train can be isolated if needed to investigate transient involving ECCS failure. The
clean condensate tanks are schematized by a suitable “tmdpvol” component
reproducing the thermal hydraulic condition of the water into the tank. The integral
mass flow-rate exiting from the “tmdpvol” component is computed by the model to
check the amount of water available to be injected by the system.
116
The Main sub-system 1, 2 and 3 and the Long term sub-system for the damaged
core side 1, 2 and 3 are connected to three main distribution headers; each header
is connected to all the 22 GDH schematized for the right side of the core; this side
is supposed to be the damaged side of the core. The three trains of the long term
sub-system for the undamaged core side 1, 2 and 3 are connected together in a
single common distribution header because only one GDH is schematized for the
right side of the core (suppose to be the undamaged side).
The logic of actuation of the systems is implemented especially in relation to

accident involving break occurrence in component in the ALS (e.g. rupture in the
piping downstream the MCP up to core inlet). The ECCS intervention logic for other
kind of accident (i.e. feed water pipe break) can be implemented adding some
specific conditions because the model does not include some special rooms where
ECCS actuation signals are generated in some transients (e.g. no feed water
compartment is modeled and the pressure increase in the this compartment as a
consequence of a feed water piping break is not reproduced).
Schematization
of ECCS
Main subsystem train 1 (accumulators)
Main subsystem train 2 (accumulators)
Main subsystem train 3 (from deaerators)
Long term damaged subsystem

train 1-2-3 (from PSP)
Long term undamaged subsystem

train 1-2-3 (from clean condensate tanks)
ECCS headers
DGH
Fig. 64 – Relap5 UNIPI model of Smolensk 3 NPP: nodalization sketch of ECCS.
117
In Tab. 22 and Tab. 23 is reported, as already mentioned at the beginning of the
present section, the list of all the components modeled identified by the RELAP5-
3D card numbers and components.

nodes and NPP components (left side).
LEFT SIDE
ID of
Component Node type
node
010, 012,
Pipe
Downcomer 014, 016
011, 015 Branch
Suction 020, 021,
Branch
headers 022, 023
030, 040,
Pipe
050, 060
Suction
031 Branch
piping
032, 042,
Pipe
044
035, 045,
Pipe
055, 065
036, 046,
Bypass line Valve
056, 066
037, 047,
Pipe
057, 067
033, 043,
MCP Pump component
053, 063
Pressure
034, 044,
pipes after Pipe component
054, 064
the pumps
070, 071,
Branch
Pressure 072, 073
headers 074, 075,
Branch
076, 077
80÷86,
Group 87÷91,
Pipe
distribution 92÷96,
header 97÷101
102÷123 Pipe
141 Branch
Additional
142 Pipe
absorber
143 Branch
channel
144 Pipe
Core 145, 148,
Pipe
channels …, 285
118
Tab. 22 (cont.) – Relap5 UNIPI model of Smolensk 3 NPP: correspondence
between nodes and NPP components (left side).
LEFT SIDE
ID of
Component Node type
node
146, 149,
Branch
Core …, 286
channels 147, 150,
Pipe
…, 287
300÷348
for the
Pipe
core
Steam channels
water pipes 329, 330
for the
Pipe
additional
absorber
354÷357 Pipe
376÷379 Branch
Steam
358, 363 Branch
Drum 1
359, 361 Pipe
360 Branch
Jet spray
system for 349÷353 Jet mixer
SD 1
489 Branch
490 Pipe
Water
491 Valve
bridge
492 Pipe
488 Branch
495 Pipe
Steam
496 Valve
bridge
497 Pipe
Steam Line
366÷367 Pipe
after SD 1
404÷407 Pipe
426÷429 Branch
Steam drum
408, 413 Branch
2
409, 411 Pipe
410 Branch
Jet spray
SD2
Steam Line
416÷417 Branch
after SD2
119
between nodes and NPP components (left side).
LEFT SIDE
ID of
Component Node type
node
395, 445 Branch
396, 446 Pipe
Steam rings 897 Branch
steam lines 898, 947 Pipe
and turbines 397, 447 Trip valve
Time dependent
398, 448
volume
Time dependent
381
volume
Feed-water Time dependent
382
system 1 junctions
383 Pipe
384 Branch
Time dependent
431
volume
432
system 2 junctions
433 Pipe
434 Branch

nodes and NPP components (right side).
RIGHT SIDE
ID of
Component Node type
node
510,
512, Pipe
Downcomer
514, 516
511, 515 Branch
520,
Suction
521, Branch
headers
522, 523
530,
Suction
540, Pipe
piping
550, 560
120
between nodes and NPP components (right side).
RIGHT SIDE
ID of
Component Node type
node
531,
541, Branch
551, 561
532,
Pipe
542, 544
535,
545,
555,
565, Pipe
537,
Bypass line
547,
557, 567
536,
546, Valve
556, 566
533,
MCP 543, Pump component
553, 563
Pressure 534,
pipes after 544, Pipe component
the pumps 554, 564
570,
571,
572,
Pressure
573, Branch
headers
574,
575,
576, 577
Group 580 Pipe
distribution
602 Branch
header
641 Branch
Additional
642 Pipe
absorber
643 Branch
channel
644 Pipe
645,
Core
648, …, Pipe
channels
666
121
RIGHT SIDE
ID of
Component Node type
node
646,
649, …, Branch
667
647,
650, …, Pipe
668
805÷812
for the
Pipe
core
Steam water
channels
pipes
829, 830
for the Pipe
AA
854÷857 Pipe
876÷879 Branch
Steam Drum
858, 863 Branch
1
859, 861 Pipe
860 Branch
Jet spray
SD 1
989 Branch
990 Pipe
Water bridge 991 Valve
992 Pipe
988 Branch
995 Pipe
Steam bridge 996 Valve
997 Pipe
Steam Line
866, 867 Pipe
after SD 1
904÷907 Pipe
926÷929 Branch
Steam drum
908, 913 Branch
2
909, 911 Pipe
910 Branch
Jet spray
SD2
122
RIGHT SIDE
ID of
Component Node type
node
Steam Line
916, 917 Branch
after SD2
Steam rings 895, 945 Branch
and turbines 896, 946 Pipe
Time dependent
881
volume
882
system 1 junctions
883 Pipe
884 Branch
Time dependent
931
volume
932
system 2 junctions
933 Pipe
934 Branch
4.5.4. The NESTLE 3D NK nodalization and the coupling

RELAP5-3D NK model was developed according to data reported in Chapter 2 of
this document, so the main parameters (power, burnup, CR position, channels
flow-rate) were referring to the Smolensk-3 NPP core status of 16 October 1996.
RELAP5-3D reactor core model is composed by 2488 channels. 1570 Fuel
Channel, 314 Non-Fuel Channels (CRs, Axial Detectors, Additional Absorber,
Water Column), 604 Radial Reflector Channels are modeled.
Radially the core was divided into square cells with a pitch of 0.25 m, each one
corresponding to a core channel. Thus, the model resulted in a 56 per 56 squared
arrays (see Fig. 68).
Axially the core was divided by 12 planes. First and twelfth plane have an axial
dimension of 0.3 m (i.e. corresponding to the graphite bottom and top reflectors),
planes from 2 to 11 have a dimension of 0.7 m. A sketch of the Fuel Channel
Neutronic Mesh is reported in Fig. 65. It has to be noted that the axial mesh utilized
is equal to the axial mesh used for the HELIOS code Cross Sections calculations.
123
12 Top Reflector
CORE MESHING
(10 x 0.7 m)
2 to 11
1 Bottom Reflector
Fig. 65 – RELAP5-3D Core neutron kinetic mesh
All different CR types are implemented in the RELAP5-3D code, with the correct
insertion direction and the appropriate physical structure (absorber + connecting
rod+ graphite displacer for MCR, only absorber for FASS, absorber + graphite
displacer for SHR). Their numeration with their arrangement is shown in Fig. 69;
their insertion depth is reported in Tab. 9 as indicated by the plant data in the
chapter 2 of this document. This CR configuration has to be intended for all
calculations here reported, where it is not specified a different one.
4.5.4.1. 3D NK-TH Coupling Scheme

Because of the different mesh dimension between the core TH nodalization (0.35
m) and the 3D NK axial mesh (0.7 m), the coupling procedures connected two
thermal-hydraulic mesh points to one neutron kinetic mesh point. Instead the
dimension of the TH node for the bottom and top reflector was the same one of the
neutron kinetic node (i.e. 0.3 m). The heat structures of the fuel and of the graphite
have the same axial dimension of the corresponding TH node and are, for this
reason, coupled in the same way with the NK nodes. In Fig. 66 it can be viewed the
coupling scheme between the HELIOS, NESTLE and RELAP5 nodalization.
124
22 22 12 Top Reflector
20
CORE MESHING
(10 x 0.7 m)
2 to 11
1 1 Bottom Reflector
1 1
Fuel TH Graphite
HS MESH HS
HELIOS 2D Transport Code

RELAP5 TH Code
NESTLE 3D NK Code
Fig. 66 – TH / 3D NK & Lattice Codes meshes coupling
Neutronic nodes were coupled to a total of 56 thermal-hydraulic channels

according to the maps showed in Fig. 68. Left part of the core was modeled with:
o 8 TH channels (grouped in NK Zone # 46 to # 53)
o 1 GDH
Right part was instead modeled with:
o 48 TH channels (grouped in NK Zone # 2 to 45, 56, 57, 58,60)
o 22 GDH
All the three types of Control Rods (Manual, Shortened, Safety) and the Axial
Detectors were modeled with an equivalent TH channel (equivalent TH channel
#288, Zone 54). Instead the Additional Absorber and the Water Column were
coupled to two TH channel, one for both halves (equivalent TH channel #142,
#642, NK Zone #55 and #59). Radial Reflector neutron kinetics nodes were
coupled with an equivalent TH channel too (equivalent TH channel #292, Zone #1)
simulating the cooling circuit.
125
Fig. 67 – Radial Arrangement of TH Channels for General Purpose Nodalization
126
Fig. 68 – Correlation map between TH channels, TH Zones and NK Nodes for General Purpose Nodalization
127
Fig. 69 – CR Numbers & Types – RELAP5-3D Numeration
128
Finally, all the neutron kinetics nodes, grouped in 60 Macro – Zones, were coupled
to a total of 60 TH channels according to the map showed in Fig. 67.
4.5.4.2. Some considerations on the axial meshing scheme
Some considerations on numerics and FD methods can help to derive the

maximum size of the node to be used by the NESTLE code. As reported in section
4.2.4, NESTLE calculations are by default employing NEM, providing sufficient
robustness and accuracy in the neutron flux distribution calculations. Nevertheless,
the implemented non-linear strategy allows to result in a FD calculation by
choosing to not update each n outer iterations the coupling coefficients between
two nodes (see section 4.2.4.1). It is well known that, using a FD method the error
on the flux calculation is proportional to the size of the node and to the migration
length. For example, for a 1D mono-energetic FD discretization it results that [71]:
3 ∆l 2
εR ≅ ( ) + ...
4 L
( 11 )
where:
∆l = dimension of node l
εR = error on the flux calculation
L = cell diffusion length
The RBMK neutronic characteristics, with a greater diffusion (and migration) length
allows the use of a relatively large (compared, e.g. to a BWR) node dimension.
From Tab. 4 of this report it results that 1/L2 for RBMK is 5.03*10-3 while for BWR
2.07*10-1. Therefore, a reasonable accuracy can be get with a RBMK node size
also of the order of 100 cm size [89]. Several tests were performed for a SS
configuration, halving the axial discretization length and changing the numerical
solution scheme. Some results of these sensitivities are showed in Fig. 70 and Fig.
71.
1.4
1.2
1
FD - 10 Axial Nodes
Relative Power
0.8 FD - 20 Axial Nodes
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7 8
Distance from the Core Bottom (m)
Fig. 70 – Relative Axial Power: SS sensitivity, FD method calculations
129
1.6
1.4 NEM - 20 axial nodes

NEM - 10 Axial Nodes
1.2
1
Relative Power
0.8
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7 8
Distance from the Core Bottom (m)
Fig. 71 – Relative Axial Power: SS sensitivity, NEM calculations
It was found that the use of a 22 axial nodes (20 for the active zone plus 2 for the
top and bottom reflectors) scheme (node length for the active zone = 35 cm) did
not improve the calculation accuracy in a significant manner. On the other hand,
the computational time increased dramatically (more than doubling in the cases of
both FD and NEM). Thus, the use of NEM with a 12 axial nodes (10 nodes of 70
cm length for the active zone plus 2 for the top and bottom reflectors) was chosen
for all the SS and transient calculations.
4.5.5. The MCNP5 code model

4.5.5.1. Reference Geometry Material
Cell modeling by MCNP5 code was done according to the following geometrical
data (see Fig. 72 and Tab. 25). These data were supplied by Russian Research
Center Kurchatov Institute (RRC KI) [73].
Fig. 72 – Fuel Channel Cross-View
130
• Distance inner row fuel rod axis - center of fuel assembly axis: 1.6 cm
• Distance outer row fuel rod axis - center of fuel assembly axis: 3.1 cm
4.5.5.2. Reference Material data

Material compositions for MCNP5 input decks were also supplied by RRC KI and
are reported in Tab. 24 and Tab. 25.
Tab. 24 – Nuclear densities in fresh fuel, 1024 atoms/cm3
Fuel type 2,0 % 2,4 %

235
U 4.3386E-4 5.2063E-4
238
U 2.0991E-2 2.0905E-2
16
O 4.4283E-2 4.4283E-2
Tab. 25 – Geometry and material composition of fuel cell
Nuclear density,
Element Size, cm
1024/см3
Zr = 4,287*10-2
Central rod Router=0,75 Hf= 7,06*10-6
Nb= 1,069*10-3
Fuel Router=0,5850 See Tab. 24
Zr= 4,338*10-2
Rinner=0,5850
Clad Hf= 1,35*10-5
Router=0,6815
Nb= 8,135*10-4
H= 6.6177*10-2
O= 3.3088*10-2
Water with mixed holding greed
Inside tube Fe= 4,774*10-4
(Water density 0.998 g/cm3)
Cr= 1,3395*10-4
Ni= 6,593*10-5
Zr= 4,287*10-2
Rinner=4,0
Tube Hf= 7,06*10-6
Router=4,4
Nb=1,069*10-3
Graphite C= 0,084233
25х25
B10= 1.8*10-8
For voiding calculation, concentrations of hydrogen and oxygen were taken as

zero, the other one were unchanged. Temperature of all cell components was fixed
at 300 K. These geometric and composition data were used for the all reference
criticality calculations given in this document.
131
4.5.5.3. MCNP5 Reference Model Geometry
Geometric dimensions and materials data used for developing the following model
are reported in the previous paragraph. An infinite – along z-axis – geometry was
considered. Model pictures are given in Fig. 73 and Fig. 74.
Fig. 73 – Fuel Cell in XY plane (at Z=10)
Fig. 74 – Fuel Cell in YZ and XZ plane (at X=0 and Y=0)
132
4.5.5.4. MCNP5 Cross Section Libraries
ENDL92 [62] and ENDF-B/VI.6 [61] continuous cross section data library was used
for reference calculations. Temperature was set up at 300 K. Thermal S(a,b)
tables were used to correctly represent the role of molecular forces acting on the
hydrogen and carbon atoms in the water and in the moderator, respectively.
4.5.5.5. Boundary Conditions

Specular reflective boundary conditions (J(m) = J(-m)) were implemented on the XZ
and YZ planes delimiting the graphite block.
4.5.6. The DRAGON code model

DRAGON is a 2D-3D deterministic transport code, so the implemented model was
similar to the MCNP and HELIOS ones and it consisted of a 2D simulation of a
single fuel channel with the relative graphite block. A picture of the model is given
in Fig. 75.
Fig. 75 – DRAGON RBMK FC model
The source data for the model set up were taken from [73] and are consistent with
the data reported in Tab. 24 and Tab. 25.
133
4.5.6.1. Cross Section Libraries
The DRAGON code can use different format of multi-group cross sections libraries.
The WIMS-D libraries, developed by the IAEA in the framework of the WIMS-D
Library Update Project were used. They have are composed by 69 and 172 energy
groups and an extensive description of them can be found in [58].
4.5.6.2. Boundary Conditions

Reflective boundary conditions on the model periphery were imposed for all the
DRAGON calculations.
4.5.6.3. Calculation modules

DRAGON code is a modular code and can perform the same type of calculation
using different solution algorithms. Several options were tested, resulting in the
selection of the modules reported in Fig. 76 [59].
EXCELT
(GEOMETRY TRACKING)
Tracking File Tracking File
SHI IAEA 69 (or 172) Groups

(SELF-SHIELDING) WIMS-
WIMS-D Libraries
Self Shielded Libraries
EVO
(Burnup calculations)
ASM
(Collision Probabilities)
FLU
(Flux & Criticality calculation)
EDI
(Editing)
Fig. 76 – DRAGON code: calculation scheme used for RBMK calculations
After the geometry and the material definition, DRAGON code effectuated the
geometry tracking using the EXCELL (Isotropic) tracking module. This information
was then used for the self shielding calculation and for the collision probabilities
matrix calculations. The use of the previous data (self-shielded libraries plus
collision probabilities matrix) allow DRAGON to calculate the multigroup transport
solution for the considered model. EDI module is then used for printing results and
for performing eventual homogenizations. EVO module was also used iteratively
for the burnup calculations.
134
4.6. The Qualification
The execution of nuclear reactor calculations should be performed according to the
best international practice guidelines (e.g., [45], [46]), and with a level of quality
comparable to the one used in other fields of the nuclear technology. Therefore the
most rigorous methodologies for calculations should be developed and submitted
to the review of the scientific and technical community (e.g., through participation to
international benchmarks and/or thorough review of a Safety Authority). The
calculation methodologies developed at the GRNSPG/UNIPI and applied here for
these RBMK analyses were discussed and tested in an international framework
during several Projects and publications (e.g., see [43], [47], etc.)
The GRNSPG/UNIPI calculation methodologies are mainly based on the respect of
the followings criteria:
• Use of the most advanced codes. Codes versions used for calculations are
“frozen” (no beta or testing versions of codes are used). Code should not
modified by the user.
• Code user should be “qualified” (see [48] for the meaning of “qualified
user”)
• Nodalizations and input decks developed should pass a “steady-state” and
“on transient” qualification process [49].
In the following paragraphs, information about the codes and the models
qualification process is reported.
4.6.1. Codes qualification

4.6.1.1. The RELAP5-3D validation for RBMK analyses
RELAP5-3D code was validated for pressure tube graphite moderated reactors in
the framework of several Projects managed by the US-DOE for the study of the
American production reactors and for the study of the RBMK reactors in the
aftermath of the Chernobyl accident (see the code description paragraph).
Example of code modifications for considering RBMK peculiarities of RBMK are,

e.g., the implementation of Osmachkin TH correlation, the implementation of a
special RBMK option for handle Kurchatov Institute’ generated Cross Section
Libraries.
The multidimensional neutron kinetic model in RELAP5-3D based on NESTLE

code, was extensively validated [75],[76]. Since the 1990 it was adapted and used
for the RBMK analyses too. In the literature it can be found recently published
papers [77], [78], [79], focused to develop RELAP5-3D models for RBMK system
(more precisely for Ignalina and Kursk NPPs). All these analyses use the RELAP5-
3D code developed by INL.
E.g., a validation of the nodal kinetics part of the RELAP5-3D was made using
steady-state data and using reactor power reduction and feedwater flow rate
perturbation transients data from the Ignalina NPP in [78], [79].
135
4.6.1.2. The HELIOS code validation for RBMK analyses
HELIOS is an industrial world-wide recognized lattice physics code, with
capabilities to model the most complex geometries. This characteristic and a
several RBMK related activities like [26] and [50], suggested its application to
RBMK calculations.
4.6.1.3. The DRAGON code validation for RBMK analyses

DRAGON code is a lattice physics code developed by the Politecnique of Montreal
with the support of the Canadian nuclear industry and utilities. Its main applications
are for heavy water reactor studies. Nevertheless the geometry modeling
capabilities as well as the IAEA multi-group libraries developed during an
International Research Project [69] allows its applications to RBMK reactors too.
DRAGON validation works for this reactor technology was not found in the
literature before the execution of the present activities.
4.6.1.4. The MCNP validation for RBMK analyses

MCNP5 is an extensively validated code, used world-wide as a code for reference
calculations. ENDFB-VI continuous energy cross section libraries were validated
for RBMK analyses [74]. Comparison of MCNP5 results with Russian Monte Carlo
codes like MCU code showed comparable agreements with experimental data [51].
136
4.6.2. Model and Nodalization qualification
4.6.2.1. RELAP5-3D TH nodalization qualification
The nodalization qualification procedure proposed at GRNSPG has been applied.
This consisted of two main steps: steady state level qualification and on-transient
level qualification. The first part of the procedure is considered. A comprehensive
documentation of the activity can be found in [44].
A nodalization is qualified at the steady state level if it has a geometrical fidelity of

the simulated plant and if it shows a stable time trend of the main thermal-hydraulic
parameters.
Relevant geometrical data are compared with the design data and a set of
acceptable errors in Tab. 26. A demonstration of the geometrical fidelity of the
nodalization can be derived from Fig. 77. Key thermal-hydraulic parameters are
considered in Tab. 27 and in Fig. 78 to Fig. 82. Emphasis is given, when reporting
selected curves, to the void fraction in the core and to the initial temperature
distribution inside graphite stacks, including radial and axial distributions.
The completed nodalization qualification process shows that the Smolensk 3

nodalization is qualified at the steady state level.
Tab. 26 – RELAP5-3D Smolensk 3 NPP: nodalization: qualification at steady state

level, key MCC geometric parameters.
RELAP5-3D Acceptab
Uni
QUANTITY Design Right side le Error Notes
t
Left side Error +
667.67 667.78
Circuit volume m3 1% 0.01%
667.67 667.78
External
12134.42
Active radius
(fuel & PT)
structures 2 12135.66 -0.01% calculated
m 19904 0.1%
heat transfer 17476.36 -12.2% to
(Graphite-
area (overall) preserve
Gap)
the mass
Active
structures
14.4 14.4 0.02%
heat transfer m3 0.2%
1088.2 1088.8 -0.05%
volume
(overall)
Graphite
blocks total Ton 1850 1851 0.05%
mass
+ The % error is defined as the ration (reference or measured value – calculated value)/(reference or
measured value); “The dimensional error” is the numerator of the above expression
137
700.0
Nodalization BAF TAF SD in

Smolensk 3 NPP
600.0
500.0
Volume (m3)
400.0
300.0
200.0
100.0
0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0
Height (m)

level, comparison between adopted and reference (design) volume vs. height curve
of MCC.
Tab. 27 – RELAP5-3D Smolensk 3 NPP: nodalization qualification at steady state

level, key MCC thermal-hydraulic parameters.
RELAP5-3D Acceptab
Quantity Unit Design Right side le Error Error° (%) Notes
left side (%)+
Thermal 1606 1594 -0.75
MW 2
power 1565 1567 0.13
6.86
6.865 0.07 0.04 On all
3
DS pressure MPa 6.86 0.1 SD
6.86
6.862 0.03 0.04 dome
3
Core inlet 269.5 -0.19
°C 270 0.5
temperature 269.1 -0.33
DS inlet 284.66 0.23
°C 284 0.5
temperature 284.65 0.23
Void 82.5 3.13
Fraction at % 80 10.0
DS Inlet 82.4 3.00
DS feed
Impose
water °C 165 165 0.5 0.0
d Value
temperature
104.71 0.11
MCP speed Rad/s 104.6 1
104.71 0.11
138
Tab. 27(cont.) – RELAP5-3D Smolensk 3 NPP: nodalization qualification at steady
state level, key MCC thermal-hydraulic parameters.
RELAP5-3D Acceptabl
Quantity Unit Design Right side e Error Error° (%) Notes
left side (%)+
Pressure 0.8 N/A
drop on the 0.5*
water line N/A
Ref. for
Pressure 0.475 N/A 2.9
drop on the 0.657*
MPa 10 N/A MW
core
channe
Steam-
0.249 N/A l
water line
0.392*
pressure
N/A
drop
Total
pressure
MPa 1.549 1.524 10 1.6
drop on the
FC
1.51 1.507 -0.20
MCP head MPa 10
1.51 1.518 0.53
85.2
84.89 -0.3 0.1
7 Liquid
DS mass 3
m 85 2 Fractio
inventory
84.4 n Vol
85.49 0.6 -0.5
7
Total loop 5157.9 5196.6 0.75

coolant flow Kg/s 2
rate 5103.9 5112.5 0.17
Technologic 5064.9 5099.2 0.68

al channel Kg/s 2
flow rate 4987.9 5001.0 0.26
Additional 93 97.4 4.74

absorber Kg/s 10
flow rate 116 111.5 -3.89
* Reference conditions refer to the maximum power channel (3.2 MW).
+ The % error is defined as the ration (reference or measured value – calculated value)/(reference or
measured value); “The dimensional error” is the numerator of the above expression
139
DC SH Inlet PH GDH Outlet CORE Inlet CORE Outlet SWL Inlet SD Inlet
9.1E+06
References
Calculated
8.6E+06
8.1E+06
Pressure (Pa)
Pump Head
7.6E+06
7.1E+06
6.6E+06
0 20 40 60 80 100 120 140
Volumes

pressure distribution along the MCC.
RBMK_SS voidg305210000
0.85 RBMK_SS voidg306210000
0.79
Void Fraction
0.77
0.75
0.73

0.69 Reference
0 20 40 60 80 100 120 140 160 180 200
Time (s)

level, comparison between steady state and reference (nominal conditions) void
fraction at SD inlet, right part, channels 1-24.
140
600
Reference
Calculated
550
500
450
Temperature (°C)
400
350
300
250
200
-1 0 1 2 3 4 5 6 7 8
Distance from core bottom (m)

graphite axial temperature at outer surface.
6
Core Height from the Bottom (m)
2
Calculation
Reference
1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Void Fraction

level, comparison between steady state and reference (nominal conditions) FC
void fraction (2.01 MW)
141
PT Graphite Graphite brick
700
670
640
610
580
550
520
Temperature (°C)
490
460
430
400
370 plane 3
plane 4
340 plane 6
plane 9
310
plane 13
280 plane 18
250
0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11 0.12 0.13 0.14
Radius (m)

level, radial temperature distribution of graphite at various elevations along the FC
stack.
4.6.2.2. RELAP5-3D 3D NK nodalization qualification

Steady state calculations for the RELAP5-3D code neutron kinetics model
validation were performed. When available, the results were compared with the
NPP data, otherwise only calculated results were reported. As can be seen in Fig.
83 and Fig. 84, the calculated trends of the axial power are similar to those
measured in the NPP.
1.40
1.20
1.00
0.80
NESTLE_DETECTORS
SMOLENSK_3_DETECTORS
0.60
0.40
0.20
0.00
0 100 200 300 400 500 600 700
Distance from the Bottom (cm)
Fig. 83 –RELAP5-3D Smolensk 3 Axial Power
142
1.4
1.2
0.8
DETECTOR 40-47
0.6
NESTLE 40-47
0.4
0.2
0
0 100 200 300 400 500 600 700
Distance from the Bottom (cm)
Fig. 84 – Smolensk 3 nodalization qualification for 3D NK-TH Relap5/Nestle code:

axial power distribution in selected FC (coordinate 40-47)
Tab. 28 – RELAP5-3D Axial Peaking Factor
Axial Peaking Factor (Fz) (from Bottom of the Core)

Maximum 1.21 at ( at 3.15 m)
Minimum 0.522 ( at 6.478 m)
1.200
NORMALIZED POWER
1.000
0.800
Normalized Power
0.600
0.400
0.200
0.000
0.000 2.000 4.000 6.000 8.000 10.000 12.000 14.000
X-COORD (m)
Fig. 85 – RELAP5-3D Radially Averaged Power
143
Fig. 86 – RELAP5-3D Relative Radial Power
Fig. 87 – 3D Thermal Neutron Flux Distribution (n/(cm2*sec))
144
Tab. 29 – RELAP5-3D Channels Radial Averaged Power
Normalized Thermal Power Channel Position

Power (MW) (xx-yy)
Maximum 1.6942 3.4091 43-20
Minimum 0.53410 1.0813 15-49
Tab. 30 – RELAP5-3D Reactor core Neutron Kinetics Parameters
Parameter Value
K-eff 1.01359 (at HFP)
Void Reactivity Feedback
CPS + 0.125 β/s (*)
FC + AA - 0.16 β
CR Worth
MCR - 12 β
MCR & FASS - 14 β
FASS - 1.7 β
(*) Imposed Value, [8].
Tab. 31 – RELAP5-3D adopted delayed neutron fraction data
Group Decay Constants (s-1) Delayed Neutron Fraction, %

1 0.0125 0.0167
2 0.0305 0.0987
3 0.111 0.0912
4 0.305 0.2094
5 1.13 0.1013
6 3.0 0.0399
TOTAL β: 0.5572
In Fig. 85, the radially averaged Power profile of the core is reported; the peaking
trend is due to the presence in the averaging columns of a different number of
active channels. From this picture is resulting that in the right side of the core there
is a higher power production.
In Fig. 87 the 3D thermal neutron flux distribution calculated by RELAP5-3D is

reported. From Fig. 86 and Tab. 29 it can be seen that the radial peak factor is
ranging from a minimum of 0.53410, localized in a channel in the bottom left
quadrant of the core, to a maximum of 1.6942, localized in the upper right
quadrant. These values can be connected to the fuel burnup distribution in that
zone and to the adopted CRs position (see Tab. 9).
In Tab. 30 are reported the reactivity coefficients calculated for the insertion of the
all type of CR (MCR, FASS and MCR and FASS together). It should be noted that
during these calculations, the Shortened CR where always supposed to be in the
same position (see Tab. 9). In the same table are also reported the values of the
reactivity inserted in case of total and partial MCC voiding and in case of CPS
145
cooling system voiding. It should be noted, also for its implications for the safety of
the RBMK system that only in the last case there was a positive reactivity feedback
during a LOCA.
In Tab. 31 are reported the delayed neutron fraction data used for the neutron
kinetics calculation. They refer to a fuel with an average burnup of 20 MWd/KgU.
4.6.2.3. HELIOS model qualification

Calculations comparison performed with MCNP and HELIOS about the keff for an
RBMK-1000 infinite lattice was executed. The results showed that the greater
deviations between the two codes are in the FC with lower enrichment FC (see
Tab. 32).
Tab. 32 – Comparison between HELIOS and MCNP codes of the kinf for a RBMK
lattice (Temp. 300 K, fresh fuel)
Presence kinf
Enrichment, Relative
of coolant in the MCNP
% HELIOS Difference,%
channel (NIKIET)
YES 2.0 1.27332 1.27838 0.40
YES 2.4 1.33144 1.33594 0.34
NO 2.0 1.32256 1.31549 -0.53
NO 2.4 1.36991 1.36475 -0.38
A comparison of the difference value between the two kinf showed that HELIOS
model underestimates the effect of a FC voiding more than MCNP model. HELIOS
error in this case is around 25% in absolute value for both different FC (see Tab.
33).
Tab. 33 – Void Effects by HELIOS and MCNP Codes, absolute units
Fuel enrichment, ε, (HELIOS/MCNP)

MCNP (NIKIET) HELIOS
% %
2.0 0.04924 0.03711 - 24.6

2.4 0.03847 0.02881 - 25.1
146
4.6.2.4. The MCNP5 model qualification
MCNP5 criticality calculation were run for estimating the kinf of the system. 10050
neutron cycles were calculated, using 104 neutron per cycle, thus totaling 100.5
Million of Histories (MH). The first 50 neutron cycles calculation (i.e., 0.5 MH) were
skipped for statistical purposes. Both ENDL and ENDF/B-VI [61] were used for
reference calculations. An example of the convergence of the kinf and of the fission
source distribution through the Shannon Entropy measurement is given in Fig. 88,
Fig. 89.
Fig. 88 – kinf trend versus cycle number
Fig. 89 – Shannon Entropy trend versus cycle number
147
Tab. 34 – Results for the reference calculations
Presence
UO2 Standard
of coolant
Run ID Enrich kinf Deviation
in the
ment, % (pcm)
channel
RBMK20 (ENDL) 2.0 YES 1.22864 5
RBMK20A (ENDL) 2.0 NO 1.25134 5
RBMK40 (ENDL) 2.4 YES 1.28352 5
RBMK40A (ENDL) 2.4 NO 1.29895 5
RBMK20 (ENDF-B-VI) 2.0 YES 1.27449 5
RBMK20A (ENDF-B-VI) 2.0 NO 1.31911 5
RBMK40 (ENDF-B-VI) 2.4 YES 1.33159 5
RBMK40A (ENDF-B-VI) 2.4 NO 1.36892 5
4.6.2.4.1. Sensitivity analyses

A series of sensitivity analyses were performed in order to:
• assess the model developed;
• evaluate the effects of the main input parameters on the calculations
results.
For the sake of simplicity, sensitivities were performed using results from the ENDL
2.0% fuel calculations only. In the followings paragraphs, the description of how the
main parameters were changed is reported.
4.6.2.4.2. Geometry Variation

Two calculations were performed with a finite-along-the-z-axis model, using 35 and
70 cm tall models respectively. Reflection condition were applied to the bottom and
the top XY surfaces. Finite 3x3 and 5x5 lattices of identical RBMK FC cell were
considered too (see Fig. 90, Fig. 91).
4.6.2.4.3. Cross Section Libraries Variation

Different cross section libraries were used in order to assess their effects on the
calculation results. The following two cases were investigated:
• Multi-group Cross Section Libraries MGXSNP (at 300 K) [80].
4.6.2.4.4. Boundary Conditions

A “white” reflective boundary condition was used instead of the normal reflection
one.
4.6.2.4.5. Calculation Parameters Variation

Two different calculation parameters were changed in order to assess their effects
on the results:
• Number of neutron cycles: 250;
• Total number of neutron histories: 1.5 MH.
148
Fig. 90 – 3 x 3 Lattice
Fig. 91 – 5 x 5 Lattice
149
4.6.2.4.6. Results of Sensitivity Analyses
Results from sensitivity analyses reported in Tab. 35 showed that:
• geometry limitation along z-axis does not influence significantly the results;
• use of lattice models (3x3 or 5x5) does not influence significantly the results;
• Multi-group cross section libraries MGXNSP as well as activation of thermal
model S(a,b) have the strongest influence on the kinf (some percent);
• decreasing of number of cycles/total number of neutron histories cause the
increase of the results variance.
Tab. 35 – Results from sensitivity calculation
Variation
Standard
Sensitivity with
Run ID # kinf Deviation
Parameter reference
(pcm)
case, %
RBMK16L Geometry: H=70 cm 1.22863 5 -0.8e-3
RBMK16K Geometry: H=35 cm 1.22882 5 +0.015
Geometry:
RBMK26 1.22870 5 -0.004
5 x 5 lattice
Geometry:
RBMK28 1.22860 5 -0.003
3 x 3 lattice
XSec:
RBMK24 1.24122 5 +1.0
No S(α, β) model
XSec:
RBMK22 1.27339 5 +3.6
Multigroup MGXSNP
Boundary:
RBMK19 1.22850 5 -0.011
White reflection
RBMK20I Calculation: 1.5 MH 1.22855 39 N/A
Calculation:
RBMK21 1.22842 34 N/A
250 cycles
NIKIET MCNP results for criticality calculations of an RBMK cell were supplied to
GRNSPG/UNIPI during the framework of the TACIS Project R2.03/97 by RRC KI.
They are reported in Tab. 36; in the fifth column, a comparison with the UNIPI
results presented in this document is given.
Tab. 36 – Comparison between NIKIET and GRNSPG criticality calculations
Deviatio
n GRNSPG Deviation
GRNSPG
Water Fuel NIKIET NIKIET/ MCNP NIKIET/GRNS
MCNP
in enrich., MCNP GRNSP (ENDF/B- PG (ENDF/B-
(ENDL)
channel % (kinf) G VI) VI)
(kinf)
(ENDL), (kinf) %
%
YES 2.0 1.27332 1.22864 +3.6 1.27449 +0.09
YES 2.4 1.33144 1.28352 +3.7 1.33159 +0.01
NO 2.0 1.32256 1.25134 +5.7 1.31911 -0.26
NO 2.4 1.36991 1.29895 +5.4 1.36892 -0.07
150
Tab. 37 – Comparison between RRC-KI and GRNSPG results for voiding effect
Deviation, Voiding Deviation,

Fuel Voiding Voiding
(NIKIET/GR GRNSPG/E (NIKIET/GRNS
enrichme NIKIET GRNSPG/END
NSPG- NDF/B-VI PG-ENDF/B-
nt (%) (∆kinf) L (∆kinf)
ENDL), % (∆kinf) VI), %
2.0 0.04924 0.02270 +117 0.04462 -9.3
2.4 0.03847 0.01543 +150 0.03733 -2.9
From the comparison, a large deviation was observed when using for calculation
the ENDL libraries. This could be explained by the presence of relevant
uncertainties in some of the cross sections used for calculation. Therefore all the
successive MCNP calculations were executed using the ENDF-B/VI libraries.
4.6.2.5. The DRAGON model qualification
Comparison of the reference calculations performed by different codes and input

decks are given in Tab. 38. A good agreement resulted in the prediction of the kinf
between the independent NIKIET MCNP4C calculations and DRAGON
calculations. This agreement increased when considering the DRAGON
calculations performed with a 172 groups cross section libraries.
Tab. 38 - Comparison of the reference criticality calculations Monte Carlo versus

DRAGON code
GRNSPG
NIKIET DRAGON
Deviation DRAGON/ Deviation
Fuel MCNP4C /EXCELL
Water in NIKIET/GR EXCELL NIKIET/GR
Enr. – IAEA 69
channel NSPG, IAEA 172 NSPG,
,% ENDF/B Groups
% Groups libr. %
data (kinf) libr. (kinf)
(kinf)
YES 2.0 1.27332 1.278768 -0.4 1.277029 -0.3
YES 2.4 1.33144 1.335680 -0.3 1.334003 -0.2
NO 2.0 1.32256 1.312085 +0.8 1.315861 +0.5
NO 2.4 1.36991 1.359907 +0.7 1.363961 +0.4
151
The comparison of the calculated FC voiding effect (Tab. 39) showed an
underestimation by DRAGON code. This underestimation is greater when using
few energy groups (69) and when considering higher enrichment of the fuel. It is
interesting to note that also the other deterministic lattice physics code, HELIOS,
showed the same problem when considering the voiding effect (underestimation of
roughly -25% was found).
The roots of this underestimation should lie both in the data libraries as well as in
the neutron transport solution methods.
Tab. 39 – Comparison between RRC-KI and DRAGON results for voiding effect
Fuel Voiding ε, Voiding

Voiding Deviation,
enric DRAGON (DRAGON DRAGON
NIKIET (NIKIET/DRA
hmen 69 69/NIKIET), 172
(∆kinf) GON 172), %
t (%) groups(∆kinf) % groups(∆kinf)
2.0 0.04924 0.03332 -32.3% 0.03883 -21.1%
2.4 0.03847 0.02423 -37.0% 0.02996 -22.1%
152
5. THE APPLICATION TO REALISTIC TRANSIENTS
ANALYSES
5.1. Hot Full Power Analyses

In this chapter, the results from a series of transient calculations are reported. All
transients were calculated considering the plant at Hot Full Power (HFP), or 3159
MW. This allowed to maximize the power and temperature excursions of the fuel
element calculated in a realistic safety analyses by the best estimate (BE) code
RELAP5-3D.
The selection of these transients was performed in order to investigate the main
relevant physical phenomena for the RBMK technology (see chapter 3 of this
document), to test the developed coupled codes model capabilities, to asses the
safety level of the RBMK system.
Six transients are here presented. They are considered the most significant to be
analyzed by coupled codes technology [52]. They are:
• CR withdrawal: Reactivity Initiated Event (RIA), transient involving
asymmetric power and flux perturbation;
• CR group withdrawal: as the previous one, but with a more significant
reactivity insertion;
• GDH blockage: decrease of coolant flow event, transient involving possible
coupled flow and power oscillations;
• Single FC blockage: decrease of coolant flow event, local transient,
influences from and on neighbor channels to be assessed;
• GDH rupture: loss of coolant event (LOCA), Design Basis Accident (DBA);
• CPS cooling circuit LOCA: Beyond Design Basis Accident (BDBA) LOCA,
accident involving strong reactivity insertion
All these transients, except the GDH rupture, were calculated using the simplified
RELAP5-3D model (see Fig. 60 in the previous chapter), simulating only the
reactor core with imposed boundary conditions. The GDH rupture, since it involved
the feedback from more NPP components (e.g., MCPs and SD), was instead
calculated using the full scale nodalization (see Fig. 59).
5.1.1. Reactivity Initiated Accidents

In these two transients, the CR/CRs were selected among the ones having the
highest worth. Their positions in the core is given in the Fig. 92.
Since in the RBMK, the CR are designed to operate in a low pressure circuit (the
CPS cooling circuit, operating at atmospheric pressure), they cannot be ejected
(e.g., like could happen in a PWR). Therefore, it has to be assumed that the cause
of this events is a malfunction of the CPS or a wrong action by the reactor operator.
153
CR WITHDRAWAL
CR GROUP WITHDRAWAL
Fig. 92 – CR Cartogram for RELAP5-3D RIA calculations
5.1.1.1. CR Withdrawal
The main purpose of the analysis of the CR withdrawal event is the evaluation of
the overall core power (even in relation to the achievement of the scram set-point)
and the demonstration of the capabilities of the developed code model to deal with
this event. After a preliminary calculation of the worth of the all MCR, the CR with
higher worth (resulted to be the #151 according to map of Fig. 35) was selected to
be withdrawn during this transient. An ‘ad-hoc’ TH and 3D NK nodalization was
developed in order to investigate the effects on the neighbor channels too (Fig. 93
and Fig. 94).
Selected results of the event originated by a spurious single CR withdrawal are

given from Fig. 95 to Fig. 101. The CR withdrawal event resulted to be:
• 0.0 s: onset of control rod #151 withdrawal with the speed 40 cm/s (1 rod
with initial insertion depth of 6.70 m from bottom of top reflector);
154
• 17 sec.: reactor peak power (3350 MW – 106% of Nominal Power, see Fig.
95)
The scram signal is generated if the reactor power reaches the level of 110% of
nominal power (3475 MW). The scram delay time is 0.25 s and the speed of CR
insertion is 3.5 and 0.5 m/s for safety and manual CR, respectively.
It can be noted that the power surge in not large enough to generate the AZ scram
signal, which set point is 110% of nominal power. The extraction of this MCR
control rod from the core results in a mild increase of power in all channels, Fig. 96.
In Fig. 96, Fig. 97 and Fig. 98 are reported the neighbor channels power trends
and their equivalent mass flow and void fraction. All these parameters were slightly
affected by the transient. In Fig. 99, Fig. 100 and Fig. 101 there are also reported
the temperatures trends of clad, of fuel centerline and of Graphite hot spot.
All the investigated parameters did not exceed the safety thresholds.
155
Fig. 93 – Radial Arrangement of TH Channels for CR Withdrawal Nodalization
156
Fig. 94 – Correlation map between TH channels, TH Zones and NK Nodes for CR Withdrawal Nodalization
157
x 10 9 WinGraf 4.1 - 08-17-2005
3.4
3.35 X
X
X
X
3.3 X X X X X X X X X
X X X X
X X
X
Power (W)
3.25
XXX RBMK_CR_Withdrawal rkotpow0
X
3.2
X
X
3.15
3.1
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 95 – Total Reactor Power
2.40E+06
2.35E+06
2.30E+06
214-01
259_EQ
Power (W)
2.25E+06 265_EQ
2.20E+06
2.15E+06
2.10E+06
2.05E+06
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 96 – 1 FC equivalent Power
158
WinGraf 4.1 - 08-17-2005
6.16
Z
6.14 Z
Z Z Z Z Z Z Z
Z Z Z Z Z Z Z Z Z Z Z Z
6.12
Mass Flowrate (Kg/s)
X X X X X X X X X X X X X X X X X
X X X X X X
6.10
XXX RBMK_CR_Withdrawal mflowj214
YYY RBMK_CR_Withdrawal mflowj259
ZZZ RBMK_CR_Withdrawal mflowj265
6.08
Y
Y
Y Y
Y Y Y Y Y Y Y Y Y Y
Y Y Y Y Y Y Y Y
6.06
6.04
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 97 – 1FC Equivalent Coolant Mass Flow – 214, 259, 265 channel
WinGraf 4.1 - 08-17-2005
.87
.87
.86 Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z
Z
Z
Z
Void Fraction
.86 Z XXX RBMK_CR_Withdrawal voidg214560000

Z
YYY RBMK_CR_Withdrawal voidg265560000
ZZZ RBMK_CR_Withdrawal voidg259560000
.85
X X X X X X
X X X X X X X X X X X
.85 X Y Y Y Y Y
Y Y Y Y Y Y Y Y Y Y Y Y
X Y
X
X X Y
.84 X
Y
Y
Y
.84
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 98 – Void Fraction at the exit of FC 214, 259, 265
159
WinGraf 4.1 - 08-17-2005
576.20
576.00
Y Y Y Y
Y Y Y Y Y Y Y Y Y Y Y Y Y
Y
575.80
Y
Temperature (K)
Y XXX RBMK_CR_Withdrawal httemp265500712

575.60 YYY RBMK_CR_Withdrawal httemp259500712
Y ZZZ RBMK_CR_Withdrawal httemp214500712
Y
575.40 Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z
Z X X X X
X X X X X X X X X X X X X
X
Z
575.20 X
Z
Z X
575.00
X
X
X
574.80
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 99 – Clad Temperature at Hot Spot
WinGraf 4.1 - 08-17-2005
1380
1360
Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
Y
Y
1340
Y
Temperature (K)
XXX RBMK_CR_Withdrawal httemp265500601

Y YYY RBMK_CR_Withdrawal httemp259500601
Y ZZZ RBMK_CR_Withdrawal httemp214500601
1320 Y
Z X Z X ZX ZX ZX ZX ZX ZX ZX ZX ZX ZX ZX ZX ZX ZX
1300 ZX
ZX
ZX
ZX
1280 Z
X
X X
1260
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 100 – Fuel CL temperature
160
WinGraf 4.1 - 08-17-2005
932.5
930.0 Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
Y
927.5
925.0
Temperature (K)
922.5 XXX RBMK_CR_Withdrawal httemp265400812

YYY RBMK_CR_Withdrawal httemp259400912
ZZZ RBMK_CR_Withdrawal httemp214400812
920.0
917.5
915.0 Z Z Z Z Z Z
Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z
X X X X X X X X X X X X X X X X X X X X
X X X
912.5
910.0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Time (s)
Fig. 101 – Graphite Temperatures
5.1.1.2. CR Group Withdrawal

The main purpose of the analysis of the CR-Group-Withdrawal event is the
evaluation of the overall core power and the demonstration of the capabilities of the
adopted code system to deal with this event. The event is originated by a spurious
signal for CR bank (or group) withdrawal. The uncontrolled withdrawal of a
peripheral control rod bank occurs (3 CR, the #179, #194 and the #196 with initial
insertion depth of 5.80 m from bottom of top reflector, see Fig. 92).
An ‘ad-hoc’ TH and 3D NK nodalization was developed in order to investigate the

effects on the neighbor channels too (Fig. 102 and Fig. 103).
The temporal development of the transient resulted to be:

1) 0 sec. – onset of CR Group withdrawal with the speed of 40 cm/s;
2) 7 sec – generation of AZ signal (Reactor power: 110% of Nominal Power,
see Fig. 104)
a. Insertion of all MCR with a speed of 50 cm/s
b. Insertion of all SR with a speed of 350 cm/s
The withdrawal speed is 0.40 m/s. The scram signal is generated when the reactor
power reaches the level of 110% of nominal power (3475 MW). The scram delay
time is 0.25 s.
Key results are given in Fig. 104 to Fig. 111. The set-point for scram signal is
achieved in a few seconds and total core power decreases as reported in Fig. 104.
Fission power in equivalent FC (per individual channel) is given in Fig. 105 and
161
axial power distribution can be found in Fig. 111. Dangerous fuel rod temperature
excursions are not predicted (Fig. 108), therefore reactor safety limits are
preserved.
162
Fig. 102 – Radial Arrangement of TH Channels for CR Bank Withdrawal Nodalization
163
Fig. 103 – Correlation map between TH channels, TH Zones and NK Nodes for CR Bank Withdrawal Nodalization
164
x 10 9 WinGraf 4.1 - 08-16-2005
4.5
3.5
X XXX RBMK_CR_BANK_W rkotpow0
3
2.5
Power (W)
1.5
X
.5 X
X X X X X X X X X X X X X X X X X
0
-.5
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 104 – Reactor Power
3.00E+06
2.50E+06
214_EQ
669_EQ
2.00E+06 259_EQ
265_EQ
Power (W)
1.50E+06
1.00E+06
5.00E+05
0.00E+00
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
Fig. 105 – 1 FC Equivalent Power
165
WinGraf 4.1 - 08-16-2005
6.90
6.80
6.70 XXX RBMK_CR_BANK_W mflowj265

X YYY RBMK_CR_BANK_W mflowj214
ZZZ RBMK_CR_BANK_W mflowj669
6.60 VVV RBMK_CR_BANK_W mflowj259

Y
6.50
Y ZY ZY Z
Y Z Y ZY ZY ZY ZY ZY ZY ZY ZY ZY ZY ZY
Y
YX XZ
6.40 V X
X Z V V X X
V X
V V
X V
X V
X V
X V
X V
X V
X V
X V
X V
X V
X V
X
V
6.30
Z
6.20
6.10
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 106 – Equivalent 1FC mass flow rate
WinGraf 4.1 - 08-16-2005
1.00
.80 Z
XY XXX RBMK_CR_BANK_W voidg214560000
XY
X YYY RBMK_CR_BANK_W voidg265560000
V
ZZZ RBMK_CR_BANK_W voidg669560000
ZY VVV RBMK_CR_BANK_W voidg259560000
.60
VZ
X
Void Fraction
.40 Y
V
XZ
.20
Y
V
X
0 ZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
VZY X
V
-.20
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
Fig. 107 – Top FC void fraction
166
WinGraf 4.1 - 08-16-2005
580.0
575.0
X XXX RBMK_CR_BANK_W httemp669500812
YYY RBMK_CR_BANK_W httemp265501112
X Y Z
570.0 ZZZ RBMK_CR_BANK_W httemp259500812
YX VVV RBMK_CR_BANK_W httemp214501012
V
ZY
565.0
Temperature (K)
V
X
560.0
ZY
555.0 V
X
ZY
V
X Y
550.0 Z
V
XZY
V Y
XZ V Y Y Y
XZ V
XZ V Y Y Y Y Y Y Y Y
XZ VXZ VXZ VXZ VXZ VXZ VXZ VXZ V
545.0 XZ VX
540.0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 108 – Hot Spot Clad Temperatures
WinGraf 4.1 - 08-16-2005
1200
1100
Z XXX RBMK_CR_BANK_W httemp265501301
X X ZZZ RBMK_CR_BANK_W httemp669500801
1000 Y Y VVV RBMK_CR_BANK_W httemp214501301
Temperature (K)
Z
900 V
X
Y
800 Z
V
X
Y
700 Z
V
X
Y
V
XZ
Y V
XZ V
600 Y XZY V
XZY V
XZY V
XZY V
XZY V
XZY V
XZY V VZY X
XZY X VZY X
VZY X
VZY V
X
500
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 109 – Hot Spot Fuel CL temperature
167
WinGraf 4.1 - 08-16-2005
840.0
835.0
XXX RBMK_CR_BANK_W httemp669401312
830.0 X X X ZZZ RBMK_CR_BANK_W httemp259401312
X X X X VVV RBMK_CR_BANK_W httemp214401512
X X X X X X
825.0 V V X X
V V X X
V X
Temperature (K)
V V X X
V V V V
820.0 V V V V V V V
815.0 Y Y Y Y Y Y Y Y Y Y Y
810.0 Y Y Y Y Y Y Y Y
805.0 Z Z Z Z Z Z Z Z Z Z Z
800.0 Z Z Z Z Z Z Z
795.0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 110 – Hot Spot Graphite Outer Temperature
3.50E+05
Time_0.0
Time_5.0
3.00E+05 Time_10.0
Time_15.0
Time_20.0
Time_30.0
2.50E+05 Time_40.0
Time_80.0
Power (W)
2.00E+05
1.50E+05
1.00E+05
5.00E+04
0.00E+00
0 1 2 3 4 5 6 7 8
Distance from the core bottom (m)
Fig. 111 – 1 FC - Equivalent axial Power
168
5.1.2. Decrease of coolant flow event
5.1.2.1. GDH blockage
The main purpose of the analysis of the GDH-blockage event is the evaluation of
the power inside the affected FCs and the demonstration of the capabilities of the
adopted code system to deal with this event. The event is supposed to be
originated by the blockage of the 11th GDH on the right half of the MCC, Fig. 67.
The available bypass mass flow rate is 60 Kg/s for the affected GDH. The
emergency protection signal is generated when the reactor power reaches the level
of 110% of nominal power (i.e. 3475 MW) or if one PT rupture occurs. The scram
delay time is 0.25 s and the speed of CR insertion is 3.5 and 0.5 m/s for safety and
manual CR, respectively.
Key results are given in Fig. 112 to Fig. 121. The total core power decreases soon
after the accident start, Fig. 112. Flow oscillations are predicted in the FC of the
affected GDH (see Fig. 113 and Fig. 114) that cause rod surface temperature
excursions in some of the channels early into the transient, Fig. 118. This is one
reason for global core power decrease (e.g. Fig. 112). The PT temperature is
reported in Fig. 120. During the calculated transient duration, the PT temperature
does overpass the safety limit.
In connection with FC thermo-hydraulic instabilities (e.g. flow-rates, Fig. 113, Fig.

114) it must be noted that a physical response of the system can be obtained when
each FC of the affected GDH is modeled. This implies modeling accurately the
length and the geometrical discontinuities of each of the 43 individual pipes
connected with the affected GDH (i.e. from the GDH to the FC inlet in the core
region) and from the individual FC outlet and the SD. This is well beyond the scope
of the present activity.
Conditions for PT rupture are reached for Smolensk at different times (these could
not be reached in Ignalina NPP because of the special-additional scram signal
considered that causes an early core power reduction following scram). The
present results, i.e. PT rupture in 300 s of calculation, show the importance of the
3D NK coupled TH model.
The close connection between thermal-hydraulic instabilities and occurrence of

CHF in the rod bundle and consequent rod surface temperature rise can be noted
from Fig. 118. The further consequence is the PT wall temperature above the
acceptable threshold, as already emphasized.
169
x 10 9 WinGraf 4.1 - 08-17-2005
3.18
3.16 XXX RBMK_GDH_BLOCKAGE rkotpow0
3.14
Power (W)
3.12
X X
X
X X X X
X X X X X X
3.1 X X X X X
X
3.08
X
3.06
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
WinGraf 4.1 - 08-19-2005
7.00
6.00
XXX RBMK_GDH_BLOCKAGE mflowj214

5.00
YYY RBMK_GDH_BLOCKAGE mflowj259
ZZZ RBMK_GDH_BLOCKAGE mflowj265
4.00
Z
Z
3.00 Z
Y Y Z Y
ZY Y
Z Z
2.00 Y Z XZ Y X X
X X X X X XZ X
X Y Z X X X
X
1.00 X Z Z X
Z Y
Z Y Z
X
X Z Z
0 Y X Y
Y Y
Y Y
Y Y
-1.00 Y
-2.00
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 113 – 1FC Equivalent Mass Flow Rate (channels 265, 259, 214)
170
WinGraf 4.1 - 08-19-2005
7.00
6.00
YYY RBMK_GDH_BLOCKAGE mflowj175
5.00 ZZZ RBMK_GDH_BLOCKAGE mflowj244
4.00
X X
3.00 X Y Y
X Y X
Y Y Y Y
Z X
Z ZY Z Z XZ Z Z Z
Y
Z
2.00 Z Z Z Z Z Z Z
Y X Z
X X X
1.00 Y YX
Y X
X Y Y Y
X X X X
0 X Y Y
X Y
-1.00
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 114 – 1FC Equivalent Mass Flow Rate (channels 244, 175, 669)
WinGraf 4.1 - 08-17-2005
5100
5050
5000
4950
XXX RBMK_GDH_BLOCKAGE mtotal_unaff_GDH
4900
4850
4800
4750
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 115 – Total Mass Flow in not-affected GDHs (Right Half)
171
WinGraf 4.1 - 08-17-2005
300
275
250
225
200
175
150
125
100
75
50
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 116 – Mass Flowrate in the affected GDH
WinGraf 4.1 - 08-17-2005
1.05
1.00 X YX YV Z JYV Z JYV Z JYX

V Z JYX
V Z JYX
V Z JYV Z JYV Z JYX
V Z JYX
V Z JYX
V Z JYV
X Z JYV Y Y Y Y Y
Z X X X XZ J V
XZ J V X JX JX
XZ J V
X X Z VZ V
.95 H H H H H H H H H H H H H H H H
Void Fraction
.90
XXX RBMK_GDH_BLOCKAGE voidg214560000
YYY RBMK_GDH_BLOCKAGE voidg259560000
ZZZ RBMK_GDH_BLOCKAGE voidg265560000
.85 VVV RBMK_GDH_BLOCKAGE voidg669560000
JJJ RBMK_GDH_BLOCKAGE voidg175560000
HHH RBMK_GDH_BLOCKAGE voidg244560000
.80
.75
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 117 – Void Fraction at top of affected GDH FCs
172
WinGraf 4.1 - 08-17-2005
1100
1000
Y Y Y
Y
Y Y Y
Y Y Y
Y Y
900 Y Y
Y Y Y
Temperature (K)
XXX RBMK_GDH_BLOCKAGE httemp214501512

Y YYY RBMK_GDH_BLOCKAGE httemp259501812
ZZZ RBMK_GDH_BLOCKAGE httemp265501612
800 VVV RBMK_GDH_BLOCKAGE httemp669501412
Y JJJ RBMK_GDH_BLOCKAGE httemp175501612
HHH V V
RBMK_GDH_BLOCKAGE
V httemp244501012
Z
VZ Z
Z V
700 VZ VZ VZ VZ
V VZ VZ Z V
V J J J J J V V V V
Z Z J J J J Z
Z J Z Z Z
J
X X J J
600
X J
X H X H X H X H X H X H X H X H X H X H X H X H X H X HJ X HJ X HJ X
500
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 118 – Hot Spot Clad Temperatures - FC 214, 259, 265, 175, 244
WinGraf 4.1 - 08-17-2005
1700
1600
Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
Y
Y
1500
Y
XXX RBMK_GDH_BLOCKAGE httemp214500601
1400
Temperature (K)
YYY RBMK_GDH_BLOCKAGE httemp259500901

Y ZZZ RBMK_GDH_BLOCKAGE httemp265500601
VVV RBMK_GDH_BLOCKAGE httemp669500601
1300 JJJ RBMK_GDH_BLOCKAGE httemp175500601
HHH RBMK_GDH_BLOCKAGE httemp244500801
X
1200
XZ XZ XZ Z Z Z Z Z Z Z Z Z Z Z Z Z Z Z
V JV JV
X JV
X JX
V JX X JV
X JV X JV
X JV
V JV V JX
X JX V JX
V JX
V JX
V JX
V JX
V
1100
1000
H H H H H H H H H H H H H H H H
900
0 50.0 100.0 150.0 200.0 250.0 300.0 350.0
Time (s)
Fig. 119 – Hot Spot Fuel CL Temperatures - FC 214, 259, 265, 175, 244
173
WinGraf 4.1 - 08-17-2005
900
850
Y Y Y Y
Y Y Y Y
Y Y Y Y
800 Y Y
Y
Y
Y
750 XXX RBMK_GDH_BLOCKAGE httemp214402101
Temperature (K)

Y ZZZ RBMK_GDH_BLOCKAGE httemp265402001
JJJ RBMK_GDH_BLOCKAGE httemp175401801
HHH RBMK_GDH_BLOCKAGE httemp244401201
Y
650
VZ VZ V
600 Z VZ Z VZ V
Z J VZ J VZ J V J J J Z V
Z V J J J Z V
X XZ V V
XZ J XHZ X H X H X H X H X H X H X H X H X H X H X HJ X HJ X H
J V J VH
ZJ X VH
ZJ X V
ZJ X
550
500
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 120 – Pressure Tubes temperatures at Hot Spot – FC 214, 259, 265, 175
WinGraf 4.1 - 08-17-2005
980
960
Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y
940 Z Z Z Z Z Z
X X X X X X X Z Z Z Z Z Z Z Z
X X X X X X X Z Z Z Z
J J J J J J X X X X X X
V V V V V V J J J J J J J J J
920 V V V V V V V V V J J
V V V
Temperature (K)
900
880 XXX RBMK_GDH_BLOCKAGE httemp214400812

ZZZ RBMK_GDH_BLOCKAGE httemp265400812
JJJ RBMK_GDH_BLOCKAGE httemp175400812
840 HHH RBMK_GDH_BLOCKAGE httemp244401212
H H H H H H H H H H H H H H H H
820
800
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 121 – Graphite Temperatures
174
5.1.2.2. Fuel Channel Blockage: the System Code Analysis
The main purpose of the analysis of the Fuel Channel (FC) Blockage event is the
evaluation of the affected FC power. This is relevant to calculate the time of the PT
failure and, therefore, the time available for any corrective action. Following FC
blockage, two counteracting main phenomena contribute to the variation of the
local (i.e. of the affected FC) fission power: a) the voiding that causes an easier
way for graphite moderated neutron to reach the fuel bundle; b) the fuel
temperature increase consequence of lack of cooling, that through the Doppler
effect tends to cause a greater parasitic neutron capture, thus introducing the
potential for power decrease. Both phenomena depend upon burn-up and, to a
lower extent, upon the position of the FC in the core region.
The analysis has been conducted, assuming no scram, with reference to the FC
indicated in the bottom left part of Fig. 67 (channel ID is 214 and surrounding group
of channel ID is 259). Key results are given in Fig. 122 to Fig. 128. It can be seen
that the affected FC power substantially decreases soon after the event, Fig. 126,
with overall core power slightly influenced by the event, Fig. 123. Conditions for
fuel rod damage and PT failure are reached in a few tens of seconds after the
transient start, Fig. 127 and Fig. 128, respectively.
WinGraf 4.1 - 08-19-2005
8.00
7.00
6.00
5.00
XXX RBMK_FC_BLOCKAGE mflowj214260000
4.00
3.00
2.00
1.00
0 X X X X X X X X X X X X
-1.00
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Time (s)

3 RBMK NPP: flow-rate in the affected FC
175
x 10 9 WinGraf 4.1 - 08-19-2005
3.16
3.16
3.16
XXX RBMK_FC_BLOCKAGE rkotpow0
X
3.16
Power (W)
X
3.15
X
3.15 X
X
X X X X X X
3.15 X
3.15
3.14
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5
Time (s)

3 RBMK NPP: core power
WinGraf 4.1 - 08-19-2005
7.00
6.80
6.60
J J J J J J J J J J J
XXX RBMK_FC_BLOCKAGE mflwoj259
YYY RBMK_FC_BLOCKAGE mflowj265

6.40
ZZZ RBMK_FC_BLOCKAGE mflowj669
VVV RBMK_FC_BLOCKAGE mflowj175
JJJ RBMK_FC_BLOCKAGE mflowj244
6.20
V V V V V V V V V V VV
6.00
Z Z Z Z Z Z Z Z Z Z Z Z
5.80 Y Y Y Y Y Y Y Y Y Y Y YY
X X X X X X X X X X
X X
5.60
5.40
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Time (s)
Fig. 124 – Mass Flowrate in the other channels of the affected GDH
176
WinGraf 4.1 - 08-19-2005
1.20
1.00 X X X X X X X
X
X X
X
X
.80
XXX RBMK_FC_BLOCKAGE voidg214560000
Void Fraction
.60
.40
.20
0
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Time (s)
Fig. 125 – Void Fraction at the Blocked Channel Exit
2.90E+06
Blocked_Channel
29-30
29-31
2.70E+06
30-29
30-30
30-31
2.50E+06
31-30
Channel Power (W)
31-31
31-32
2.30E+06
2.10E+06
1.90E+06
1.70E+06
1.50E+06
0.0 5.0 10.0 15.0 20.0 25.0
Time (s)
Fig. 126 – Affected FC power and power (per unit FC) in neighbouring FC.
177
WinGraf 4.1 - 08-19-2005
2000
XXX RBMK_FC_BLOCKAGE httemp214500112

1800 YYY RBMK_FC_BLOCKAGE httemp214500212
ZZZ RBMK_FC_BLOCKAGE httemp214500312
VVV RBMK_FC_BLOCKAGE httemp214500412
1600 JJJ RBMK_FC_BLOCKAGE B
O
httemp214500512
HHH RBMK_FC_BLOCKAGE
#
B
O
A
H
C
DJ H
D
httemp214500612
E F
F V
### RBMK_FC_BLOCKAGE httemp214500712
V
Z G
1400 OOO RBMK_FC_BLOCKAGE IG
httemp214500812
Temperature (K)
#
O
B
A
H
D
CJ
AAA RBMK_FC_BLOCKAGE
E
F httemp214500912
BBB RBMK_FC_BLOCKAGE
V
Z httemp214501012
CCC IG
RBMK_FC_BLOCKAGE httemp214501112
1200
#
O
H
B
A DDD RBMK_FC_BLOCKAGE
L
MY L
Y
httemp214501212
#
B
O
A
H
CJ
D D
C
E
FJ X
B
O
A
#
HJ FE V
Z LY
D
C
E
F V
Z IGEEE RBMK_FC_BLOCKAGE
M httemp214501312
#
B
O
A
H
D
C
E
FJ V
Z IG FFF X
IG
1000 B
#
O
A
H
D
C
J
E V
Z
IG LY P N
N
B
#
O
A
H
C
J V
Z L L
Y
M
X
MXGGG RBMK_FC_BLOCKAGE httemp214501512
V
Z M
Y
X III N
RBMK_FC_BLOCKAGE
P httemp214501612
O
A
#
H
JV Y
X
Z Y
X N LLL RBMK_FC_BLOCKAGE httemp214501712
800 Y N
P P
#
H
JV X MMM RBMK_FC_BLOCKAGE httemp214501812
Z Y
X NNN RBMK_FC_BLOCKAGE httemp214501912
Y
X PPP RBMK_FC_BLOCKAGE httemp214502012
600 Z
X
Y JV
Y
Z
X
400
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5
Time (s)
Fig. 127 – Rod surface temperature at different elevations in the affected FC.
WinGraf 4.1 - 09-09-2005
800
XXX RBMK_FC_BLOCKAGE httemp214400101
YYY RBMK_FC_BLOCKAGE httemp214400201
ZZZ RBMK_FC_BLOCKAGE httemp214400301
750 VVV RBMK_FC_BLOCKAGE httemp214400401
JJJ RBMK_FC_BLOCKAGE httemp214400501
HHH RBMK_FC_BLOCKAGE httemp214400601
AB B
O
D
### OC
D
RBMK_FC_BLOCKAGE
FF
#
E G
httemp214400701
OOO G
RBMK_FC_BLOCKAGE
I H httemp214400801
700 H L
AAA RBMK_FC_BLOCKAGE
JL
M httemp214400901
Temperature (K)
BBBA
OB
D
C RBMK_FC_BLOCKAGE httemp214401001
CCC E
#
F RBMK_FC_BLOCKAGE httemp214401101
IG
H NN
DDDJLM RBMK_FC_BLOCKAGE
PV httemp214401201
650 V
EEE RBMK_FC_BLOCKAGE httemp214401301
A
OD
E
FC
B
#
A
OD
E
FC
B
# IG FFFN RBMK_FC_BLOCKAGE httemp214401401
D
F
C IG H
L V P
E
B
A
IG
O# L
H JM GGG RBMK_FC_BLOCKAGE httemp214401501
A
OD
E
FC
B
IG
# L
M
H JM N III RBMK_FC_BLOCKAGE httemp214401601
D
E
C
B
A
O J
600 C
B
A # JH N V P LLL RBMK_FC_BLOCKAGE httemp214401701
O
# H V P
A
O H J V Z
Q Q
#
H J V MMM RBMK_FC_BLOCKAGE
R httemp214401801
#
H J V V NNN RBMK_FC_BLOCKAGE httemp214401901
V Z Z Z Z S
Z J
V
Z JZ
V Z Z Z Z Q
R PPP Q
R
S
T
S
RBMK_FC_BLOCKAGE
T httemp214402001
550 QQQ RBMK_FC_BLOCKAGE httemp214402101
Y
X Y
X Y
X X
Y Y
X Y
X Y
X Y
X Y
X Y
X RRR Y
X XYY
SSS RBMK_FC_BLOCKAGE httemp214402301
TTT RBMK_FC_BLOCKAGE httemp214402401
500
0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 22.5 25.0
Time (s)
Fig. 128 – PT temperature at different elevations in the affected FC.
178
5.1.2.3. The Fuel Channel Blockage: the FC criticality calculations
In order to perform precise criticality analyses for a FC flow blockage accident,
Monte-Carlo code MCNP5 calculations were performed considering as calculation
domain a single FC and a 3x3 lattice of RBMK cells. Boundary conditions for
MCNP5 input were derived by a previous transient calculation by state-of-the-art
codes HELIOS/RELAP5-3D.
The changes of the main physical parameters (e.g. fuel and water/steam
temperature, water density, graphite temperature) at different time intervals of the
FC blockage transient were evaluated by a RELAP5-3D calculation. This
information was used to set up further MCNP5 inputs. Criticality analyses were
performed for different systems (single channel and lattice) at those transient’
states, obtaining global criticality versus transient time. Finally the weight of each
parameter’s change (fuel overheating and channel voiding) on global criticality was
assessed.
5.1.2.3.1. Introduction
Cell kinf was calculated for several physical states occurring during a Fuel Channel
Blockage (FCB) event. Fuel and moderator temperatures as well as coolant density
values at different time-steps were obtained by a previously executed 3D NK – TH
calculation by RELAP5-3D code. These values are given in Tab. 40.
Tab. 40 – Cell parameter variation during a FCB event by RELAP5-3D code
Avg.
Avg. Fuel Coolant Coolant
Transient Moderator
Temperature Density Temperature
Time (s) Temperature
(K) (Kg/m3) (K)
(K)
0.0 800 543 563 850
8.0 1100 18.63 836 850
15.0 1300 15.2 1000 850
20.0 1500 12.6 1194 850
In order to assess the effects played by the different models, several calculations
were run for both single FC models (with fuel 2.0% and 2.4%) and for lattice fuel
channels models, considering FC blockage occurring in the central FC of the
lattice. Results are reported hereafter.
5.1.2.3.2. Single Fuel Channel - 2.0% Fuel
Criticality Calculations
Excluding the reference state at 0.0 second (run ID Xi), 9 runs were executed (see
Tab. 40) changing, in each run:
• all parameters, thus obtaining system criticality (runs X1, X4, X7);
• only coolant parameters, thus assessing voiding effect (runs X2, X5, X8);
• or only fuel temperature, thus assessing Doppler effect (runs X3, X6, X9).
179
All parameter changes were done according to values reported in Tab. 40. Results
of cell kinf with standard deviation are reported in Tab. 41 and Fig. 129. Low
variance was obtained running 103 active cycles, simulating 104 neutrons histories
per cycle.
phenomena, of channel voiding, of Doppler effect
System criticality Channel voiding Doppler Effect

Transient
Time (s) RUN s RUN s RUN s
kinf kinf kinf
ID (pcm) ID (pcm) ID (pcm)
0.0 Xi 1.28033 15 N/A N/A N/A N/A N/A N/A
8.0 X1 1.27654 17 X2 1.28343 15 X3 1.27395 15
15.0 X4 1.27590 16 X5 1.28319 16 X6 1.27382 16
20.0 X7 1.27581 16 X8 1.28290 16 X9 1.27323 15
1.28600
1.28400
1.28200
System Kinf
1.28000
Voiding Effect
Kinf
Doppler Effect
1.27800
1.27600
1.27400
1.27200
0 5 10 15 20 25
Time (s)
Some considerations were obtained by a qualitative analysis of the previous

results. FC total reactivity during this kind of transient seemed reduced thanks to
overwhelming negative reactivity insertion by Doppler effect; this phenomenon
looked opposed to the positive reactivity insertion generated by void effect.
Reaction Rates and neutron flux tallying
In order to have a rigorous justification for the previous results, reaction rates tallies
for Xi to X3 runs were executed, assessing the effects of the main phenomena (fuel
overheating and channel voiding) on the neutron fluxes and reaction rates. Results
180
for runs Xi to X3 are given in Tab. 42. The effects of parameters changes on the
different tallies for runs X1, X2, X3 compared with run Xi is showed in Tab. 43. Low
variance for the tallies was obtained running a large number (108) of neutron
histories. Tallies were executed dividing the energy domain into two energy groups:
the fast group (from 14 MeV to 0.625 eV) and the thermal group (from 0.625 eV to
0 eV).
Tab. 42 – 2.0% fuel cell reaction rates and fluxes for runs Xi to X3
Xi X1 X2 X3
TALLYING
Fuel 2.0% Thermal Thermal Thermal Thermal
Total Total Total Total
Fast Fast Fast Fast
3.52E-04 3.42E-04 3.44E-04 3.50E-04 3.85E-
Sfission * F 3.87E-04 3.86E-04 3.88E-04
04
3.49E-05 4.34E-05 4.34E-05 3.48E-05
Sabsorption 1.49E-04 1.47E-04 1.47E-04 1.48E-04
(fuel) * F 2.56E-
2.54E-04 2.74E-04 2.72E-04
1.05E-04 1.28E-04 1.25E-04 1.08E-04 04
(excluding
fission)
Sabsorption * F  2.27E-05 3.56E-06 3.58E-06 2.26E-05 2.39E-
2.40E-05 3.87E-06 3.89E-06
(in the water) 1.33E-06 3.10E-07 3.11E-07 1.33E-06 05
2.90E-03 3.21E-03 3.21E-03 2.90E-03 8.38E-

F (in fuel) 8.39E-03 1.03E-02 1.03E-02
5.49E-03 7.09E-03 7.11E-03 5.48E-03 03
Sscattering * F  2.12E-03 2.28E-03 2.28E-03 2.12E-03

3.67E-
(in the 3.67E-03 4.32E-03 4.33E-03
1.55E-03 2.04E-03 2.05E-03 1.55E-03 03
graphite)
Tab. 43 – 2.0% fuel cell – Ratio of Tallies, %
X1/Xi, % X2/Xi, % X3/Xi, %

Ratios of Tallies - Fuel
2.0% Thermal Total Thermal Total Thermal Total
Fast Fast Fast
-2.8 -2.2 -0.5
Sfission * F -0.3 0.2 -0.5
24.3 24.5 -0.2
Sabsorption (fuel) * F -1.2 -0.8 -0.4
8.0 7.1 0.9
(excluding fission) 21.1 18.3 2.7
Sabsorption * F -84.3 -84.2 -0.4
-83.9 -83.8 -0.4
(water) -76.7 -76.7 -0.4
10.7 10.7 -0.1
F (in fuel) 22.8 23.0 -0.2
29.1 29.5 -0.2
Sscattering * F 7.3 7.4 -0.1
17.6 17.8 -0.1
(graphite) 31.9 32.1 -0.1
181
Considering Tab. 43, the followings can be stated.
• Channel voiding and steam overheating (run X2) is causing a large drop in
the neutron absorption by the water (-83%) and a shift of the energy
spectrum of the neutrons towards higher energies (flux hardening, +29%);
consequently, an increase of the neutron flux (F = n*v) results (+23%).
Faster flux is increasing the scattering reaction rates in the graphite block
(+32% in epithermal-fast zones); hardened flux is increasing neutron
absorption and neutron fissions in epithermal-fast zones (+18% and
+24.5% respectively). Neutron multiplication per generation (i.e., kinf) is
increased as result from the sum of all these effects.
• Fuel overheating (run X3) as a sensible impact on the neutron population,
interesting neutron absorption epithermal-fast zones; consequently a
decrease on neutron population and fissions result.
• The results from the physical simulation of both effects and its comparison
with the reference case (fuel and coolant overheating, run X1) is
reproduced in the first column of Tab. 42 and Tab. 43 respectively. The
coolant overheating is causing a flux hardening that results in an increase
of the scattering reaction rates in the graphite block, in an increase of the
neutron flux, and in an increase of the fast fissions (ε term in the kinf factor).
Fuel overheating is instead causing an increase of the neutron absorption
(p term in the kinf factor) in the and contributing to the decrease of thermal
fissions. The overall number of fissions (tallying on fast and thermal
energies groups) is therefore decreased. Thus, the system criticality
decrease (see Fig. 129).
5.1.2.3.3. Single FC – 2.4% Fuel
Criticality Calculations
The same considerations exposed in the previous paragraph can be applied for the
single FC with the 2.4% enriched fuel. The identification of the different cases is
done using the same nomenclature specified above. From Tab. 44 and Fig. 130
respectively, it results that the system kinf is always decreasing, also when
considering channel voiding effect only. The explanation of these phenomena is
given in the next paragraph.

Transient
kinf kinf kinf
0.0 Xi 1.33503 15 N/A N/A N/A N/A N/A N/A
8.0 X1 1.32694 17 X2 1.33383 16 X3 1.32873 16
15.0 X4 1.32680 16 X5 1.33369 16 X6 1.32850 16
20.0 X7 1.32637 16 X8 1.33375 16 X9 1.32788 15
182
1.33600
1.33500
1.33400
1.33300
System Kinf
1.33200 Voiding Effect
Doppler Effect
1.33100
Kinf
1.33000
1.32900
1.32800
1.32700
1.32600
1.32500
0 5 10 15 20 25
Time (s)
Reaction Rates and neutron flux tallying
As for the 2.0 % single FC case (see previous paragraphs for calculation
parameter specifications), several tallies were executed, measuring the effects of
the main phenomena (fuel overheating and channel voiding) on the neutron fluxes
and reaction rates. Results for runs Xi to X3 are given in Tab. 45. The effects of
parameters changes on the different tallies for runs X1, X2, X3 compared with run
Xi is showed in Tab. 46.
183
Tab. 45 – 2.4% fuel cell reaction rates and fluxes for runs Xi to X3
Xi X1 X2 X3
TALLYING
Fast Fast Fast Fast
3.65E-04 3.52E-04 3.55E-04 3.63E-04
Sfission * F 4.04E-04 4.01E-04 4.03E-04 4.02E-04
3.91E-05 4.87E-05 4.88E-05 3.91E-05
(fuel) *F 2.45E-04 2.64E-04 2.62E-04 2.47E-04
(excluding 1.06E-04 1.28E-04 1.25E-04 1.09E-04
fission)
Sabsorption * 2.00E-05 3.10E-06 3.12E-06 1.99E-05
F 2.13E-05 3.41E-06 3.43E-06 2.12E-05
(in the 1.32E-06 3.06E-07 3.06E-07 1.31E-06
water)
2.54E-03 2.79E-03 2.79E-03 2.54E-03
F (in fuel) 8.01E-03 9.84E-03 9.86E-03 7.99E-03
5.47E-03 7.05E-03 7.07E-03 5.46E-03
Sscattering * 1.97E-03 2.09E-03 2.10E-03 1.97E-03
F 3.52E-03 4.13E-03 4.14E-03 3.51E-03
(in the 1.55E-03 2.04E-03 2.04E-03 1.54E-03
graphite)
phenomena, of channel voiding, of Doppler effect, %
X1/Xi, % X2/Xi, % X3/Xi, %

TALLYING Fuel 2.4% Thermal Thermal Thermal
Total Total Total
Fast Fast Fast
-3.3 -2.8 -0.5
Sfission * F  -0.6 -0.1 -0.5
24.4 24.6 -0.2
Sabsorption (fuel) * F  -1.8 -1.3 -0.4
8.0 7.0 0.9
(excluding fission) 20.7 18.0 2.6
Sabsorption * F  -84.5 -84.4 -0.4
-84.0 -83.9 -0.4
(water) -76.8 -76.8 -0.4
9.8 9.9 -0.1
F (in fuel) 22.9 23.1 -0.2
29.0 29.3 -0.2
Sscattering * F  6.2 6.3 -0.1
17.4 17.6 -0.1
(graphite) 31.7 31.9 -0.1
From Tab. 45 and Tab. 46, the following considerations result:

• channel voiding and steam overheating (run X2) is causing a large drop in
neutron absorption by water (-84%) and a shift of the neutron energy
spectrum towards higher energies (+29%). Consequently, an increase of
184
the neutron population (+23%) results. The faster neutron flux is increasing
the scattering reactions in the graphite block (+32% in epithermal-fast
zones) as well as epithermal neutron absorption and fast fissions reaction
rates in the epithermal-fast zones (+18% and +24.6% respectively). The
voiding effect on kinf is negative (kinf is slightly decreasing ) because the
fissions reaction rates are decreasing in thermal zone more than for the
2.0% fuel case. Therefore, a sensible reduction of the number of neutrons
from fissions results, causing a reduction of the kinf.
• Fuel overheating (run X3) has a sensible impact on the neutron population,
interesting neutron absorption epithermal-fast zones; consequently a slight
decrease of the neutron population and of fissions reaction rates result.
The magnitude of the reductions is the same as for the 2.0% fuel case.
• The results from the physical simulation of both effects and its comparison
with the reference case (fuel and coolant overheating, run X1) is
reproduced in the first column of Tab. 45 and Tab. 46 respectively. The
coolant overheating is causing a flux hardening that results in an increase
of the scattering reaction rates in the graphite block, in an increase of the
neutron flux, and in an increase of the fast fissions. Fuel overheating is
instead causing an increase of the neutron absorption and contributing to
the decrease of thermal fissions. The overall number of fissions (tallying on
fast and thermal energies groups) is therefore decreased, in this case
more than in the 2.0% fuel case. Thus, the system criticality decreases
(see Fig. 130).
5.1.2.3.4. Lattice Cell Analysis

Similar analyses were performed for studying the effects of channel voiding and
fuel overheating phenomena when they take place in a FC of a lattice, i.e.
simulating an event more similar to what is occurring in a real RBMK core. For this
purpose, the 3x3 lattice model was used, considering blockage of the central FC.
Analyses were performed for the following configurations:
1. 2.0 % fuel lattice
2. 2.4 % fuel lattice
3. 2.4 % fuel lattice with 2.0 % fuel for blocked channel
Results of these analyses are reported in the following paragraphs.
2.0 % Fuel : Criticality Calculations
Results of criticality calculations are given in Tab. 47 and in Fig. 131. Very low
variance was obtained running 104 neutron histories for 104 active neutron cycles
(108 MH). Doppler and voiding effect in the central FC produced the same
qualitative effects on the lattice reactivity as in the case of a single FC with 2.0%
enriched fuel (i.e., positive and negative insertion, for voiding and Doppler
respectively). System criticality is instead, in this case, increased.
It should also be noted that in all these simulations with lattice of FCs, the
magnitude of kinf change is smaller than in the case of single FCs simulations (tens
of pcm compared to hundreds of pcm, respectively).
185
For the identification of the different cases, the same nomenclature specified in the
previous paragraphs was used.
considering occurrence of all phenomena, of channel voiding, of Doppler effect

Transient
kinf kinf kinf
0.0 Xi 1.28030 5 N/A N/A N/A N/A N/A N/A
8.0 X1 1.28068 5 X2 1.28139 5 X3 1.27957 5
15.0 X4 1.28066 5 X5 1.28128 5 X6 1.27966 5
20.0 X7 1.28058 5 X8 1.28132 5 X9 1.27957 5
1.2820
System Kinf
Voiding Effect
1.2815 Doppler Effect
1.2810
Kinf
1.2805
1.2800
1.2795
1.2790
0 5 10 15 20 25
Time(s)
2.0 % Fuel : Reaction Rates and neutron flux tallying
As for the single FC calculations, several reaction rates and flux tallies were
executed assessing the effects of the main phenomena. Results for runs Xi to X3
are given in Tab. 48 and Tab. 49, reporting values for the central channel (blocked
channel) and for the border channels of the lattice. The effects of parameters
changes on the different tallies for runs X1, X2, X3 in comparison with run Xi is
186
showed in Tab. 50. Low variance for the tallies was obtained running a large
number of neutron histories (108 MH).
Tab. 48 – 2.0% lattice fuel cells reaction rates and fluxes for runs Xi and X1
Xi X1
TALLYING
Central Border Central Border
Lattice
Fast Fast Fast Fast
3.91E-05 4.30E- 3.91E-05 4.30E- 3.42E-05 3.97E-05 4.36E-
Sfission * F 3.81E-05
3.88E-06 05 3.88E-06 05 3.96E-06 3.96E-06 05
(fuel) * F  2.82E- 2.82E- 2.87E-
2.60E-05
(excluding 05 05 05
1.17E-05 1.17E-05 1.13E-05 1.20E-05
fission)
Sabsorption * F 2.52E-06 2.52E-06 3.55E-07 2.55E-06
2.67E- 2.67E- 2.70E-
(in the 3.83E-07
1.48E-07 06 1.48E-07 06 2.76E-08 1.52E-07 06
water)
3.22E-04 9.32E- 3.23E-04 9.33E- 3.18E-04 3.27E-04 9.49E-
F (in fuel) 9.82E-04
6.10E-04 04 6.11E-04 04 6.64E-04 6.22E-04 04
Sscattering * F  2.36E-04 2.36E-04 2.28E-04 2.39E-04
4.08E- 4.08E- 4.15E-
(in the 4.12E-04
1.72E-04 04 1.72E-04 04 1.84E-04 1.76E-04 04
graphite)
Tab. 49 – 2.0% lattice fuel cells reaction rates and fluxes for runs X2 and X3
X2 X3
TALLYING
Lattice
Fast Fast Fast Fast
3.43E-05 3.97E-05 3.89E-05 4.28E- 3.91E-05 4.30E-
Sfission * F 3.83E-05 4.36E-05
3.97E-06 3.96E-06 3.86E-06 05 3.88E-06 05
Sabsorption
1.47E-05 1.67E-05 1.65E-05 1.65E-05
(fuel) * F  2.85E- 2.82E-
2.58E-05 2.87E-05
(excluding 05 05
1.11E-05 1.20E-05 1.20E-05 1.17E-05
fission)
Sabsorption * F 3.56E-07 2.55E-06 2.51E-06 2.52E-06
 2.66E- 2.66E-
3.84E-07 2.70E-06
(in the 2.76E-08 1.52E-07 1.47E-07 06 1.48E-07 06
water)
187
Tab. 49(cont.) – 2.0% lattice fuel cells reaction rates and fluxes for runs X2 and X3
X2 X3
TALLYING
Lattice
Fast Fast Fast Fast
3.17E-04 3.27E-04 3.22E-04 9.29E- 3.23E-04 9.33E-
F(in fuel) 9.84E-04 9.49E-04
6.67E-04 6.22E-04 6.07E-04 04 6.11E-04 04
Sscattering *
2.28E-04 2.39E-04 2.36E-04 2.36E-04
F 4.08E- 4.08E-
4.12E-04 4.15E-04
(in the 04 04
1.84E-04 1.76E-04 1.72E-04 1.72E-04
graphite)
Tab. 50 – 2.0% lattice fuel cell – Ratio of Tallies, %
Ratios of X1/Xi , % X2/Xi , % X3/Xi , %

Tallies -
Central Border Central Border Central Border
Lattice
Fuel Th. Th. Th. Th. Th. Th.
2.0% Tot. Tot. Tot. Tot. Tot. Tot.
Fast Fast Fast Fast Fast Fast
-12.7 1.4 -12.3 1.4 -0.4 0.0
Sfission* F  -11.3 1.4 -11.0 1.4 -0.4 0.0
2.1 2.1 2.4 2.1 -0.3 0.0
Sabsorption* -11.4 1.4 -11.1 1.4 -0.3 0.0
-8.0 1.8 -8.7 1.8 0.8 0.0
F (fuel) -3.3 2.4 -5.4 2.4 2.4 0.0
Sabsorption* -85.9 1.4 -85.8 1.4 0.0 0.0
F -85.6 1.4 -85.6 1.4 -0.4 0.0
-81.4 2.4 -81.3 2.4 -0.5 0.0
(water)
-1.5 1.4 -1.6 1.4 0.0 0.0
F(fuel) 5.3 1.7 5.6 1.7 -0.4 0.0
8.9 1.9 9.4 1.8 -0.5 0.0
Sscattering* -3.3 1.1 -3.3 1.1 0.0 0.0
F 1.0 1.7 1.1 1.7 -0.1 0.0
7.0 2.5 7.2 2.5 -0.2 0.0
(graphite)
From Tab. 48 to Tab. 50, the followings can be stated.

• Central FC coolant overheating (run X2) is causing a large drop in the
neutron absorption by the water (-86%) and a shift of the energy spectrum
of the neutrons towards fast energies (flux hardening, +9.4%, sensibly
less than +29% occurring in the single FC model because of neutron
leakages). This is causing a neutron leakage towards border FC of the
lattice; accordingly, there is an increase of the fast scattering (+7.2%) in
the central graphite block and a decrease of total fission (-11%) and
absorption (-8.7%) reaction rates in the fuel of the central FC. The
reaction rates of the border FCs are instead all increased of some
percent.
188
• Fuel overheating is affecting only the reaction rates of the central FC.
Neutron absorption is increased in the fuel.
• The results from the physical simulation of both effects (fuel and coolant
overheating, run X1) are reported in the first column of Tab. 48 and Tab.
50 respectively. The coolant overheating in the central FC is causing a
flux hardening that results in an increase of the scattering reaction rates in
the graphite block, in an increase of the neutron flux and fast fissions and
in an increase of neutron leakages toward peripheral FCs. Reaction rates
in the peripheral FCs increase of some percents, while the overall fission
and absorption reaction rates of the central FC are sensibly reduced (-
11% and -8% respectively). The system neutron multiplication capability
(i.e., kinf) is slightly increased (see Fig. 131).
2.4 % Fuel: Criticality Calculations
Results of criticality calculations for the 2.4% enriched fuel are given in Tab. 51 and
in Fig. 132. Also in this case, very low variance was obtained running 104 neutron
histories for 104 active neutron cycles (i.e., 108 MH). Doppler and voiding effect in
the central FC produced negative and positive reactivity insertion respectively;
system criticality did not change. For the indication of the different cases, the same
nomenclature specified in the previous paragraphs was used.

Transient
kinf kinf kinf
0.0 Xi 1.33513 5 N/A N/A N/A N/A N/A N/A
8.0 X1 1.33511 5 X2 1.33573 5 X3 1.33435 5
15.0 X4 1.33512 5 X5 1.33575 5 X6 1.33442 5
20.0 X7 1.33503 5 X8 1.33574 5 X9 1.33439 5
189
1.3360
1.3358
1.3356
System Kinf
Doppler Effect
1.3352
Kinf
1.3350
1.3348
1.3346
1.3344
1.3342
0 5 10 15 20 25
Time(s)
2.4 % Fuel: Reaction Rates and neutron flux tallying
Reaction rates tallies were executed assessing the effects of the main phenomena.
Results for runs Xi to X3 are given in Tab. 52 and Tab. 53 (see the previous
paragraphs for the nomenclature and for details about calculation procedure),
reporting tallies values for the central channel (blocked channel) and for the border
channels of the lattice. The effects of parameters changes on the different tallies
for runs X1, X2, X3 in comparison with run Xi is showed in Tab. 54.
190
Tab. 52 – 2.4% lattice fuel cells reaction rates and fluxes for runs Xi and X1
Xi X1
TALLYING
Lattice Fuel
2.4% Thermal Thermal Thermal Thermal
Fast Fast Fast Fast
4.05E-05 4.05E-05 3.53E-05 4.10E-05
Sfission* F 4.49E-05 4.49E-05 3.98E-05 4.55E-05
4.35E-06 4.35E-06 4.44E-06 4.44E-06
(fuel) *SF 2.72E-05 2.72E-05 2.50E-05 2.77E-05
(excluding 1.18E-05 1.18E-05 1.14E-05 1.21E-05
fission)
2.37E-06 2.37E-06 3.38E-07 2.40E-06
(water) * F 1.46E-07 1.47E-07 2.72E-08 1.50E-07
2.82E-04 2.83E-04 2.77E-04 2.87E-04
F(in fuel) 8.90E-04 8.91E-04 9.39E-04 9.06E-04
6.07E-04 6.08E-04 6.62E-04 6.20E-04
Sscattering 2.19E-04 2.19E-04 2.11E-04 2.21E-04
(graphite) * 3.91E-04 3.91E-04 3.94E-04 3.97E-04
F 1.72E-04 1.72E-04 1.84E-04 1.76E-04
Tab. 53 – 2.4% lattice fuel cells reaction rates and fluxes for runs X2 and X3
X2 X3
TALLYING
Lattice
Fast Fast Fast Fast
3.55E-05 4.11E-05 4.03E-05 4.05E-05
Sfission * F 3.99E-05 4.55E-05 4.47E-05 4.48E-05
4.44E-06 4.44E-06 4.33E-06 4.35E-06
(fuel) * F 2.48E-05 2.77E-05 2.74E-05 2.72E-05
(excluding 1.11E-05 1.21E-05 1.21E-05 1.18E-05
fission)
3.39E-07 2.26E-06 2.36E-06 2.37E-06
(water) * F 2.72E-08 1.50E-07 1.46E-07 1.47E-07
2.77E-04 2.87E-04 2.82E-04 2.83E-04
F (in fuel) 9.41E-04 9.06E-04 8.87E-04 8.91E-04
6.64E-04 6.19E-04 6.05E-04 6.08E-04
(graphite) * 3.95E-04 3.97E-04 3.91E-04 3.91E-04
F 1.84E-04 1.76E-04 1.72E-04 1.72E-04
191
Tab. 54 – 2.4% lattice fuel cells – Ratio of Tallies, %
Ratios of X1/Xi, % X2/Xi, % X3/Xi, %

Tallies -
Central Border Central Border Central Border
Lattice
Fuel Th. Th. Th. Th. Th. Th.
2.4% Tot. Tot. Tot. Tot. Tot. Tot.
Fast Fast Fast Fast Fast Fast
-12.7 1.3 -12.5 1.3 -0.4 0.0
Sfission * F -11.3 1.4 -11.0 1.4 -0.4 0.0
2.1 2.1 2.2 2.1 -0.3 0.0
Sabsorption* -11.5 1.3 -11.3 1.3 -0.3 0.0
-7.9 1.8 -8.7 1.8 1.0 0.0
F (fuel) -3.2 2.4 -5.4 2.4 2.7 0.0
Sabsorption * -86.0 1.3 -86.0 1.3 -0.3 0.0

-85.7 1.4 -85.7 1.4 -0.3 0.0
F (water)
-81.4 2.4 -81.4 2.4 -0.5 0.0
F (in -1.8 1.4 -2.0 1.4 0.0 0.0
5.5 1.7 5.7 1.7 -0.3 0.0
fuel) 8.9 1.9 9.3 1.8 -0.4 0.0
Sscattering * -3.8 1.0 -3.8 1.1 0.0 0.0
F 0.9 1.6 1.0 1.7 -0.1 0.0
(graphite) 6.9 2.4 7.1 2.4 -0.2 0.0
From to Tab. 52 to Tab. 54, the followings can be stated:

• Central FC coolant overheating (run X2) is causing a large drop (-86%) in
central FC of neutron absorption rates and a shift of the neutron energy
spectrum toward higher energies (+7.1% for fast scattering reaction rates
in central channel graphite and +9.3% increase of the fast flux). This
causes an increase of neutron leakage from the central FC to the border
FCs, a sensible reduction of the fissions and neutron absorption in the fuel
of central FC (-11%). The reaction rates in the border FCs are instead
increased of some percent by the neutron leakages from the central FC.
• Central FC fuel overheating (run X3) is causing an increase of 1.0% in the
neutron absorption rate. Border FCs are not perturbed by this
phenomenon.
• The results from the physical simulation of both effects (fuel and coolant
overheating, run X1) are reported in the first column of Tab. 52 and Tab.
54. The involved phenomena are the same as those reported in the 2.0%
case analysis. The kinf resulted to be not affected by these phenomena,
i.e., there is a very close balance between neutron absorption and neutron
production .
192
2.4 % fuel lattice with 2.0 % fuel in the central channel : criticality calculations
The purpose of this simulation was to simulate one of the possible zone
configuration of an RBMK core. Today, the RBMK core is mostly composed by
2.4% FCs and by few 2.0% FCs (e.g, a configuration for Smolensk-3 NPP core
resulted in about 1522 channel with 2.4% fuel and just 48 channels with 2.0% fuel).
The following model was considered:
• Central FC with 2.0% fuel
• Border elements of a 3x3 FC lattice with 2.4% fuel
The results of criticality calculations are reported hereafter (see Tab. 55 and Fig.
133).
Tab. 55 – 2.4% fuel lattice w/ central 2.0% FC: criticality during a FCB event on
central channel considering occurrence of all phenomena, of channel voiding, of
Doppler effect

Transient
Time (s) RUN s RUN RUN s
kinf kinf s (pcm) kinf
ID (pcm) ID ID (pcm)
0.0 Xi 1.32934 5 N/A N/A N/A N/A N/A N/A
8.0 X1 1.33021 5 X2 1.33068 5 X3 1.32862 5
15.0 X4 1.33012 5 X5 1.33074 5 X6 1.32866 5
20.0 X7 1.33004 5 X8 1.33068 5 X9 1.32857 5
1.3310
1.3305
1.3300
System Kinf
Kinf

Doppler Effect
1.3290
1.3285
1.3280
0 5 10 15 20 25
Time(s)
Fig. 133 – 2.4% fuel lattice w/ central 2.0% FC: criticality during a FCB event on
central channel considering occurrence of all phenomena, of channel voiding, of
Doppler effect
193
2.4 % fuel lattice with 2.0 % fuel in the central channel : Reaction Rates and
neutron flux tallying
Reaction rates tallies were executed assessing the effects of the main phenomena.
Results for runs Xi to X3 are given in Tab. 56 and Tab. 57 (see the section
5.1.2.3.2 for the nomenclature and for details about calculation procedure),
reporting tallies values for the central FC (blocked channel) and for the border FCs
of the lattice. The effects of parameters changes on the different tallies for runs X1,
X2, X3 in comparison with run Xi is showed in Tab. 58.
Tab. 56 – 2.4% lattice fuel w/ 2.0% central FC: reaction rates and fluxes for runs Xi
and X1
TALLYING Xi X1
Lattice FC
2.4% -
Central Thermal Thermal Thermal Thermal
FC 2.0% Total Total Total Total
Fast Fast Fast Fast
3.73E-05 4.07E-05 3.23E-05 4.13E-05
Sfission * F 4.10E-05 4.51E-05 3.62E-05 4.58E-05
3.71E-06 4.37E-06 3.81E-06 4.46E-06
(fuel) * F 2.71E-05 2.73E-05 2.50E-05 2.78E-05
(excluding 1.14E-05 1.18E-05 1.11E-05 1.21E-05
fission)
2.54E-06 2.38E-06 3.63E-07 2.42E-06
(water) * F 1.44E-07 1.47E-07 2.70E-08 1.50E-07
3.08E-04 2.84E-04 3.02E-04 2.88E-04
F (in fuel) 8.87E-04 8.96E-04 9.35E-04 9.11E-04
5.79E-04 6.12E-04 6.34E-04 6.23E-04
(graphite) 3.92E-04 3.93E-04 3.96E-04 3.99E-04
*F 1.68E-04 1.72E-04 1.80E-04 1.76E-04
194
Tab. 57 – 2.4% lattice fuel w/ 2.0% central FC: reaction rates and fluxes for runs
X2 and X3
TALLYING X2 X3
Lattice FC
2.4% -
Central Thermal Thermal Thermal Thermal
FC 2.0% Fast Fast Fast Fast
3.25E-05 4.13E-05 3.71E-05 4.07E-05
Sfission * F 3.63E-05 4.58E-05 4.08E-05 4.51E-05
3.82E-06 4.46E-06 3.70E-06 4.37E-06
(fuel) * F 2.48E-05 2.78E-05 2.73E-05 2.73E-05
(excluding 1.09E-05 1.21E-05 1.17E-05 1.18E-05
fission)
3.65E-07 2.42E-06 2.53E-06 2.38E-06
(water) * F 2.71E-08 1.50E-07 1.43E-07 1.47E-07
3.01E-04 2.88E-04 3.08E-04 2.84E-04
F (in fuel) 9.37E-04 9.11E-04 8.84E-04 8.96E-04
6.36E-04 6.23E-04 5.76E-04 6.12E-04
(graphite) * 3.96E-04 3.99E-04 3.92E-04 3.93E-04
F 1.80E-04 1.76E-04 1.68E-04 1.72E-04
Tab. 58 – 2.4% lattice fuel cell w/ 2.0% central FC: Ratio of Tallies, %
Ratios of X1/Xi, % X2/Xi, % X3/Xi, %

Tallies -
Lattice Central Border Central Border Central Border
FC 2.4% - Th. Th. Th. Th. Th. Th.
central FC Tot. Tot. Tot. Tot. Tot. Tot.
2.0% Fast Fast Fast Fast Fast Fast
-13.2 1.3 -12.9 1.3 -0.5 0.0
Sfission * F -11.8 1.4 -11.4 1.4 -0.4 0.0
2.7 2.1 2.9 2.1 -0.3 0.0
Sabsorption * -11.9 1.3 -11.6 1.3 -0.4 0.0
-7.9 1.7 -8.7 1.8 0.8 0.0
F (fuel) -2.4 2.3 -4.6 2.3 2.5 0.0
Sabsorption * -86.0 1.3 -85.9 1.3 -0.3 0.0

-85.7 1.4 -85.7 1.4 -0.3 0.0
F (water)
-81.2 2.3 -81.2 2.3 -0.5 0.0
-2.0 1.4 -2.1 1.4 0.0 0.0
F (in fuel) 5.5 1.6 5.7 1.6 -0.3 0.0
9.5 1.8 9.9 1.8 -0.5 0.0
Sscattering * F -3.4 1.1 -3.5 1.1 0.0 0.0

1.0 1.6 1.1 1.6 -0.1 0.0
(graphite) 6.9 2.3 7.2 2.3 -0.2 0.0
From Tab. 56 to Tab. 58 the followings can be stated:
195
• the involved phenomena are the same as for the other cases analyzed
(see 2.0% and 2.4% lattice case analyses); this configuration allows a
stronger hardening of the flux (+9.5% compared to +8.9% for the 2.0% and
2.4% fuel) that is reflected in a greater decrease of the fissions in the
central FC (-11.8% versus -11.3% for the other cases) and of the thermal
absorption (-11.9% versus -11.4/-11.5% for the other cases); on the other
hand, fast neutron fission are increased more than in the other cases
(+2.7% versus +2.1%).
• the system kinf increased more than for the 2.0% fuel lattice case (+87 pcm
versus +38 pcm). This could be explained by the higher enrichment of the
fuel lattice considered in this case.
5.1.2.3.5. Conclusions
Criticality and reaction rates analyses of a FC blockage accident were performed,
deriving boundary conditions and main parameters changes by a previously
executed RELAP5-3D calculation. It was found thatt during such type of accident
the fission reaction rates were significantly reduced (average of -11%) in the FC
where the blockage was occurring.
This was due to:
• the increased neutron leakage towards the other un-perturbed FCs; this
phenomenon was caused by the shift of the neutron energy spectrum
towards epithermal-fast regions (flux “hardening”) as a consequence of the
coolant overheating and density reduction;
• the increased neutron absorption by Doppler effect in the overheated fuel
of the blocked FC.
The other important outcome of these calculations was the estimation of the
neighboring channels behavior; because of the increased neutron leakage from the
blocked FC, all the reaction rates in these un-perturbed FCs are increased by 1-
2%, thus not producing any relevant effects of safety concern.
Finally, it should be noted that these calculations reported here were performed
assuming fresh fuel (i.e., no presence of Plutonium was considered). Therefore,
further MCNP calculations should consider, through the use of appropriate
depletion codes (e.g., ORIGEN code), the effects of actinides and fission products.
Other improvements could be obtained by the modeling of the Additional Absorber
also.
196
5.1.3. LOCA events
5.1.3.1. CPS LOCA
The CPS is cooled by a low pressure loop independent from the MCC. However
leakages from the loop may occur. This justifies the analysis of the CPS-LOCA
event. The main purpose for the analysis is the calculation of the overall core
power response.
It was supposed that the initiating event was the rupture of the CPS cooling system
collector and that the voiding of the CR channels occur in 40 seconds (see Fig.
138). This last assumption is considered conservative [81].
The time trend of the main events is:
1) 0.0sec - onset of the CPS LOCA
2) 24.5 sec – AZ scram signal for reactor power equal to 110% of Nominal
Power (Fig. 134)
3) 50.0 sec – voiding of the CPS LOCA completed (Fig. 138)
The power surge in the FC is roughly the same (Fig. 135). Because the actuation
of AZ scram signal, the fuel clad temperatures at the hot spots is not of safety
concern (Fig. 137).
x 10 9 WinGraf 4.1 - 08-19-2005
4.5
3.5
XXXXX
3 XXX RBMK_CPS_LOCA rkotpow0
2.5
Power (W)
1.5
.5
X
X X X X X X X X X X X X X
0
-.5
0 50.0 100.0 150.0 200.0 250.0 300.0 350.0
Time (s)
197
3.0E+06
FC_669
FC_265
FC_244
2.5E+06
FC_175
FC_259
FC_214
2.0E+06
Power (W)
1.5E+06
1.0E+06
5.0E+05
0.0E+00
0 50 100 150 200 250 300 350
Time (s)
Fig. 135 – Power in FC of 11th GDH of right MCC part
WinGraf 4.1 - 08-19-2005
5.40
5.20 HH
XXX RBMK_CPS_LOCA mflowj214
YYY RBMK_CPS_LOCA mflowj259
ZZZ RBMK_CPS_LOCA mflowj265
VVV RBMK_CPS_LOCA mflowj669
5.00
JJJ RBMK_CPS_LOCA mflowj175

H HHH RBMK_CPS_LOCA mflowj244
X
4.80 V Z VX Z VXZ V Z Z
H HZ V X
HZ VX
J H YJ H Y
J H X J VH
X J VH XZ VH
J
XZ VH
J
XZ VH
J
XZ VH
XZ VH
X
J J
Y J
Y Y Y Y Y Y Y J
Y J
Y
JJ J Z
Y
4.60 X
V
VVV
ZZZZ
XXXXX
4.40
YYYY Y
4.20
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 136 – Equivalent Mass Flowrate in FC of 11th GDH of right MCC part
198
WinGraf 4.1 - 08-19-2005
580.0
575.0 YYYY Y
ZZZZ
XXXXX
VVV
V
570.0
565.0 X XXX RBMK_CPS_LOCA httemp214500612

Temperature (K)
YYY RBMK_CPS_LOCA httemp259500612

ZZZ RBMK_CPS_LOCA httemp265500612
560.0 Z VVV RBMK_CPS_LOCA httemp669500612
555.0 Y
550.0 V
XZ
YV Z Y
X V XZ Y VX Z Y V X Z Y V XZ Y VX Z YV X Z Y V XZ Y VX Z YV X Z Y V XZ Y VX
545.0
540.0
0 50.0 100.0 150.0 200.0 250.0 300.0
Time (s)
Fig. 137 – Hot Spot Clad Temperature in FC 214, 259, 265, 669
WinGraf 4.1 - 08-19-2005
1.40
1.20 XXX RBMK_CPS_LOCA voidg288460000

YYY RBMK_CPS_LOCA voidg288450000
ZZZ RBMK_CPS_LOCA voidg288440000
1.00 XYJ#
VAXZY
H J#
OVH
B XC
AZY
JO#
VH
B X
AZY
DCJEVH
O#
B XC
A
DZY
F JEVH
O#
B XZY
GC
A
DF JE VH
IO#
B XZY
GC
A
DF
L JE MH
IO#
V
B X
A
DZY
GC
F
L JE MH
IO#
N V
B XZVVV
GC
A
D
PF
L Y
JN MH
IO#
E V
B X
GC
A
D
PF
ZN
L YIORBMK_CPS_LOCA
JE MH
#
V
B X
GC
A
D
PF
ZN
L Y
JE VH
IO#
B
M X
GC
A
D
PF
ZN
L Y
JE VH
IO#
B
M X
GC
A
D
PF
ZN
L Y
JE VH
IO#
B
M X
GC
A
D
PF
ZN
L Y
JEvoidg288430000
IOVH
#
B
M X
A
DF
GL
PZ
JJJ RBMK_CPS_LOCA voidg288420000
HHH RBMK_CPS_LOCA voidg288410000
### RBMK_CPS_LOCA voidg288400000
.80
Z OOO RBMK_CPS_LOCA voidg288390000
Void Fraction
AAA RBMK_CPS_LOCA voidg288380000

BBB RBMK_CPS_LOCA voidg288370000
.60 CCC RBMK_CPS_LOCA voidg288360000
DDD RBMK_CPS_LOCA voidg288350000
EEE RBMK_CPS_LOCA voidg288340000
.40 O FFF RBMK_CPS_LOCA voidg288330000
GGG RBMK_CPS_LOCA voidg288320000
III RBMK_CPS_LOCA voidg288310000
.20 Y LLL RBMK_CPS_LOCA voidg288300000
XV MMM RBMK_CPS_LOCA voidg288290000
ZH NNN RBMK_CPS_LOCA voidg288280000
X Y J
#
0 ZVZJV
XX
YY H PPP RBMK_CPS_LOCA voidg288270000
-.20
0 50.0 100.0 150.0 200.0 250.0 300.0 350.0
Time (s)
Fig. 138 – Void fraction in CPS Cooling Channel
199
5.1.3.2. GDH LOCA (rupture after the GDH check valve)
The main purpose of the analysis of the GDH LOCA simulation is the evaluation of
the overall MCC response with main consideration to the core power and the
demonstration of the capabilities of the adopted code system to deal with this
event. However, this even is not the most severe LOCA. The event is originated by
the break of one GDH, with breaking occurring after the GDH check valve.
Conditions for scram occurrence are the same as in the previous sections.
The ‘standard’ TH and 3D NK nodalization was developed in order to investigate

the effects on the all channels of the affected GDH (see Fig. 67 and Fig. 68).
The temporal development of the transient resulted to be:
1) 0 sec. – onset of the break;
2) 0.4 sec – onset of the SCV closure
3) 1.6 sec – AZ scram signal – insertion of MCR and FASS in the core
a. Insertion of all MCR with a speed of 50 cm/s
b. Insertion of all Safety Rods with a speed of 350 cm/s
Immediately after the GDH break, flow rates of FC connected to the affected GDH,
revert their flow directions (see Fig. 140 and Fig. 141). Coolant flows into reactor
compartment from the PH and from the DS connected to the FC of the affected
GDH (Fig. 142). Due to power decrease following AZ activation and flow reverse,
the void fraction at the exit of the accident FC’s decrease firstly because of
saturated water delivery from DS and then grows up to 1.0 after complete DS
depletion (Fig. 145). At the same time pressure decrease in FC’s, results in
channels’ voiding, in deterioration of heat removal from fuel rod surfaces (Fig. 146
to Fig. 150) and in void fraction growth at the entrance of the accident FC’s (Fig.
144). The clad temperature is decreased thanks to the combined effects of the
power decrease caused by the scram and thanks to the steam flow coming from
DS. In Fig. 151 it is possible to observe that the PT temperature trends are not of
safety concern
Decrease of full reactor power to 30% Nном occurs quickly after AZ signal, therefore
closure of the MCP valve plates is actuated in the NPP. Nevertheless, as Russian
calculation demonstrated, MCP flowrate increases for a while because the
reduction of the total hydraulic MCC resistance for the leak opening. At the end of
calculation the pressure in the intact GDH of the left half starts to decrease,
causing left half MCP degradation due to cavitations.
200
x 10 9 WinGraf 4.1 - 08-19-2005
3.5
3 X
2.5
X
Power (W)
2
XXX RBMK_GDH_RUPTURE rkotpow0
1.5
1 X
X
.5 X
XX
XXX
X X X X X X X X X X X X X X
0
-.5
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
WinGraf 4.1 - 08-19-2005
8.0
6.0
4.0 XXX RBMK_GDH_RUPTURE mflowj214

YYY RBMK_GDH_RUPTURE mflowj259
2.0 ZZZ RBMK_GDH_RUPTURE mflowj265
0
Y
XY
X
XZ
-2.0
YY
XZ Z Y
X Y
Z
X Y
X Y
Z
X Y
X
X Y
Z Y Y
Z
X Y
X
Z Y X Y
Z
X
-4.0 X X
Y
Z Y
Y X
ZYY
-6.0 YY
X X
-8.0 X
X
X
-10.0
X
-12.0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 140 – Mass Flow rate in affected Channels 214, 259, 265
201
WinGraf 4.1 - 08-19-2005
8.0
6.0
4.0
2.0 XXX RBMK_GDH_RUPTURE mflowj669

YYY RBMK_GDH_RUPTURE mflowj175

0 ZZZ RBMK_GDH_RUPTURE mflowj244
XX
YY
XZ
-2.0
XX
Y Y
X X
Z
Y X
Y X
Z
Y X
Y X
X X X
Z
Y X X
Z
Y X X Z
-4.0 Z Y Y
Y XX Y Y Y
X
-6.0 ZXX Z
Y
-8.0 Y YY
Z
Y
-10.0
Y
Z
-12.0
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 141 – Mass Flow rate in affected Channels 669, 175, 244
WinGraf 4.1 - 08-19-2005
1250
1000
XXX RBMK_GDH_RUPTURE mflowj90020000
750
XXXXXX
XXXX X X YYYX RBMK_GDH_RUPTURE mflowj91010000
X X X X X X X X X X
500
250
0
Y Y Y Y Y Y Y
YY Y Y Y Y Y Y
-250 Y Y
YY
Y
-500
-750
YY
-1000
-1250
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 142 – Mass Flowrate at GDH rupture
202
WinGraf 4.1 - 08-19-2005
6000
5500 X
XXX X
XXXX
5000 XXX X XXX RBMK_GDH_RUPTURE Mass_Flow
4500
Flowrate (Kg/s)
4000 X X
3500
3000
2500
2000 X X X X
1500 X X X X X
1000
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
Fig. 143 – Imposed Mass flow rate in damaged MCC part
WinGraf 4.1 - 08-19-2005
1.20
1.00 VVV
YYYY
Z Z
XXXX
J ZY V
Y
Z YV ZV
JY
H
X HV
JY
X JV
Y
Z
HX V
JY
HX HV
JY
ZX V
Y
JH
X
X J HY
XZYYY
VVV VV JH
X V
YX
JH
Z V
Y
JX
H V
Y
ZX
JH V
Y
JH
X JV
Y
ZH
X
J X
JH
HX JJXX
.80 HJXHH
JH
Void Fraction
.60
XXX RBMK_GDH_RUPTURE voidg214260000

YYY RBMK_GDH_RUPTURE voidg259260000
.40 ZZZ RBMK_GDH_RUPTURE voidg265260000
VVV RBMK_GDH_RUPTURE voidg669260000
JJJ RBMK_GDH_RUPTURE voidg175260000
.20 HHH RBMK_GDH_RUPTURE voidg244260000
-.20
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 144 – Void Fraction at Core Inlet
203
WinGraf 4.1 - 09-02-2005
1.10
1.00 Y X Y
Z
X Y
X Y
Z
X Y
X Y
Z
X Y
X
X Y Y X
Z
X X
Y Z
Y
XZ Y Z
Y
X
.90 X
X
Y
Z
.80 Y
XXX RBMK_GDH_RUPTURE voidg244560000
.70 Z
Void Fraction
X YYY RBMK_GDH_RUPTURE voidg175560000

X ZZZ RBMK_GDH_RUPTURE voidg214560000
.60 XY
Y XX
.50 X Y Y
Y
Y
.40 X
Y Z
X
.30
Z
.20
.10
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Time (s)
Fig. 145 – Void Fraction at core outlet
WinGraf 4.1 - 08-19-2005
800
XXX RBMK_GDH_RUPTURE httemp214500112

YYY RBMK_GDH_RUPTURE httemp214500212
750 Z
V ZZZ RBMK_GDH_RUPTURE httemp214500312
VVV RBMK_GDH_RUPTURE httemp214500412
JJJ RBMK_GDH_RUPTURE httemp214500512
Y
X HHH RBMK_GDH_RUPTURE httemp214500612
700 ### RBMK_GDH_RUPTURE httemp214500712
OOO RBMK_GDH_RUPTURE httemp214500812
Temperature (K)
AAA RBMK_GDH_RUPTURE httemp214500912

X BBB RBMK_GDH_RUPTURE httemp214501012
Y
V
J
650 Y CCC RBMK_GDH_RUPTURE httemp214501112
X DDD RBMK_GDH_RUPTURE httemp214501212
EEE RBMK_GDH_RUPTURE httemp214501312
Z## A
O#B
OB FFF RBMK_GDH_RUPTURE httemp214501412
HJHJHJH
X
V
J C
AD
O
Y #
H B
C GGG RBMK_GDH_RUPTURE httemp214501512
600 VVZVJV O
#
H
ZJV
X III RBMK_GDH_RUPTURE httemp214501612
YYX
X Y LLL RBMK_GDH_RUPTURE httemp214501712
X JV
X#
Z
H
O
A
B
C
D
Y
E
F JV
#
H
O MMM RBMK_GDH_RUPTURE httemp214501812
YY
XX D
E
B
C
A
#
O
H F
D
E
B
C IG XY
B
C
D
E
IF Z
JV
Y
H
X
#
O
BA JV
XY
H
JV
XZ
Y XO
#
H
JV
Y L C
D
E
IF
M
G
L #
O
B
C
D
EL NNN
IF
NM Z
V
J
Y
H
X
#
O
B
C
DA RBMK_GDH_RUPTURE httemp214501912
550 E
IF
M
P
G
L
N NP
M
L NP P P
PPP IF
E
D
B
C M
IG
L
E
F
D L
DM
N
IF LNM
IG
RBMK_GDH_RUPTURE
F P
M
N
IO
L P
N
L
M
IG P
M
N
httemp214502012
#
H
J
V
Y
XO #
H
JZ
V
Y
X
C
B
A
O JE
C
#
H
V
Y
XBO D
C
#
H
JZ
V
Y
X
E
B
A
O E
F
D
C
B
#
H
J
V
Y
X
E
D
B
C
O
#
J
H
V
Z
Y
X
F
A F
D
E
#
O
H
Y
X
500
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
Fig. 146 – Clad Temperature in FC 214
204
WinGraf 4.1 - 08-19-2005
850

800 YYY RBMK_GDH_RUPTURE httemp259500212
ZZZ RBMK_GDH_RUPTURE httemp259500312
Z
V VVV RBMK_GDH_RUPTURE httemp259500412
750 HHH RBMK_GDH_RUPTURE httemp259500612
Y ### RBMK_GDH_RUPTURE httemp259500712
X
Temperature (K)

700 AAA RBMK_GDH_RUPTURE httemp259500912
X
Y BBB RBMK_GDH_RUPTURE httemp259501012
V
YJ CCC RBMK_GDH_RUPTURE httemp259501112
X X Z DDD RBMK_GDH_RUPTURE httemp259501212
JV
Y
650 H EEE RBMK_GDH_RUPTURE httemp259501312
FFF RBMK_GDH_RUPTURE httemp259501412
# GGG RBMK_GDH_RUPTURE httemp259501512
JV
HA
O
#
J
H Z
J
V
600 ZB
V#
O
HC
J
V B
A
#
O
H
J
H
#
Y
X
O
A
B V
J
III RBMK_GDH_RUPTURE httemp259501612
Y
XY V
ZDB
#
O
J
C
H C
D X
Y
H
#
O
XY V E B
C Z
X
V
Y
J LLL RBMK_GDH_RUPTURE httemp259501712
XYXY D
E
F H
#
O
A
B
X C
D
E
F X
Y
V
J
H
#
O Z MMM RBMK_GDH_RUPTURE httemp259501812
X IG B
C
D
E
F X
Y
V
J
OH
#
A
B X
Y
J
VH Z RBMK_GDH_RUPTURE
IL IC
DE
F
G
M
L O
C
D#
B X
Y
J
V
NNN D
H
#
OA httemp259501912
IE
F
M
L
N B
C
E
IF
P
G
M
550 PPP LN F
D
E
C
B
IO
H
Y P
N
M
L
JV
#
X DE
C
B
#
O
H
P
N
M
G
A N
L
IE
D
CB
#
P
M
F
O N
M
L
G
E
D
C
B
A
#IP
RBMK_GDH_RUPTURE
IFL F
O N
M
L
E
D
C
B
#
POhttemp259502012
IF N
L
F
D
C
O P
M
IG
AE
B P
N
M
F
E
D
O
X
YJV
Z H
JV
Y
X H
JV
Z
Y
X H
JV
Y
X H
JV
Y#
Z
X #
H
Y
X
500
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
WinGraf 4.1 - 08-19-2005
800

Z YYY RBMK_GDH_RUPTURE httemp265500212
750 V ZZZ RBMK_GDH_RUPTURE httemp265500312
Y
X JJJ RBMK_GDH_RUPTURE httemp265500512
HHH RBMK_GDH_RUPTURE httemp265500612
Temperature (K)

Y BBB RBMK_GDH_RUPTURE httemp265501012
650 XVJ
X CCC RBMK_GDH_RUPTURE httemp265501112
Y
Z
V DDD RBMK_GDH_RUPTURE httemp265501212
J
H
Y
X V
JH EEE RBMK_GDH_RUPTURE httemp265501312
Y
X
# Z
V
X
J
Y
O H
# FFF RBMK_GDH_RUPTURE httemp265501412
#
J
H#
JA
VHOV B
C O
A
600 V
Z Z
JJ
H
O
#B
BH D B
C X
VY
J# GGG RBMK_GDH_RUPTURE httemp265501512
V#
O
A
Y
X
C D HO
YY
X E B Z
X
V
JOY III RBMK_GDH_RUPTURE httemp265501612
XXY C
DE H
#
F B
C
DA X
V
JO
H Y LLL RBMK_GDH_RUPTURE httemp265501712
X E
F
IG #
B
C X
Z
V
J#
H Y
DE
IF
L B
C
DO
A V
X
JO
H
# Y MMM
XZ
JV
Y RBMK_GDH_RUPTURE httemp265501812
E
F
IG
M
L B
C
D H
#
O
550
IF
EM
L
N NNNIA
B
C
D
E
M
F
L
NG
P M
N
LRBMK_GDH_RUPTURE
P MP httemp265501912
IF
D
C
B
X
#
HE
Y
O IN
L
D
C
B
#F
G
E
A
O P
N
M
L
IO
F
E
D
C
B
# M
N
L
D
C
B
#
P
IE
G
F
O
A P
N
M
L
IF
E
D
C
B
O M
N
L
F
D
C
BP
IE
G
A P
N
M
F
E
D
PPP J
V H
J
V
XZ
Y H
J
V
Y
X H
J
V
Y
XZ #
H
J
V
RBMK_GDH_RUPTURE
Y
X O
#
H
J
V
Y
XZ O
#
H
httemp265502012
Y
X
500
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
205
WinGraf 4.1 - 08-19-2005
800

YYY RBMK_GDH_RUPTURE httemp669500212
750 ZZZ RBMK_GDH_RUPTURE httemp669500312
Z
V VVV RBMK_GDH_RUPTURE httemp669500412
Y
X HHH RBMK_GDH_RUPTURE httemp669500612
Temperature (K)

BBB RBMK_GDH_RUPTURE httemp669501012
Y
650 CCC RBMK_GDH_RUPTURE httemp669501112
XJV
X
Y DDD RBMK_GDH_RUPTURE httemp669501212
ZH
J
V
H #
O
A EEE RBMK_GDH_RUPTURE httemp669501312
Y
X FFF RBMK_GDH_RUPTURE httemp669501412
#H
H
J #
O
B
600 VJ
ZJVH
V O
#
A
B Z
JV
H
# GGG RBMK_GDH_RUPTURE httemp669501512
JV
CH Y
O
X
A
B JV
Y
XY Z #
JO
B
C C
D X
Y
H
#
O III RBMK_GDH_RUPTURE httemp669501612
XXYYV D E B
C
D
E
F
Z
JV
X
Y
H
A#
O
B LLL RBMK_GDH_RUPTURE httemp669501712
X C
D
E
F JV
X
YH#
O
X IG B
C
D
E Z
Y
JV
XH# MMM RBMK_GDH_RUPTURE httemp669501812
Y
X IF
L AO
B
C
D
E
F
IG
M
L Y
JV
X
H
#
O
BCNNNJV
D Z
Y
X
H#
O
550
E
IF
M
N
L IA
B
C RBMK_GDH_RUPTURE
D
E
P
M
F
L
G
N E
F
P
D
IH
N
C
M
B httemp669501912
L
O
#
J
V
Y IP
N
M
GL
E
F
D P
M
N
L
IH
F
E P
N
M
L
IO
G
EF N
MP
IO
LF MP
N
L
IG P
N
M
X C
B
A
O
#
H
J
Z
Y
X
CD
B
O
#
J
Y
X
D
C
B
A
#
J
H
PPP RBMK_GDH_RUPTURE
V V V
Z
Y
X JE
D
C
#
HB
V
Y
X C
#
HF
E
D
B
A
O
V
Z
Y
X
E
F
D
O
#
httemp669502012
J H
Y
X
500
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
WinGraf 4.1 - 08-19-2005
660

640 YYY RBMK_GDH_RUPTURE httemp244500212
ZZZ RBMK_GDH_RUPTURE httemp244500312
Z JJJ RBMK_GDH_RUPTURE httemp244500512
620 V HHH RBMK_GDH_RUPTURE httemp244500612
X
Y ### RBMK_GDH_RUPTURE httemp244500712
Temperature (K)

600 AAA RBMK_GDH_RUPTURE httemp244500912
BBB RBMK_GDH_RUPTURE httemp244501012
Y CCC RBMK_GDH_RUPTURE httemp244501112
DDD RBMK_GDH_RUPTURE httemp244501212
580 X EEE RBMK_GDH_RUPTURE httemp244501312
FFF RBMK_GDH_RUPTURE httemp244501412
J #
X V OC
AD
BBB
C
OH
O
### #
B
DO
A
JV
H
C
E
F
IG
Z GGG RBMK_GDH_RUPTURE httemp244501512
Y
XH HA
JVH
OYYY
H
JJJ
VVV
Z D
C
BE
A
O D
C
BF
E YX JV
DO
#
H
B
C
E
560 JV#YZXXX #
H
JV
Z #
HO
JV YL
IF
X M III RBMK_GDH_RUPTURE httemp244501612
YZH
J
V X YX YX JV
D
Y#
H
L
O
IF
B
C
XZ
A
G
E JV
M
H
O
#N
XY
X D
YL
IE
B
LLL
F
C
X M
J
LRBMK_GDH_RUPTURE
P
IF
V
HN
Z httemp244501712
#
Y
X
B
D
CO
G
E
A P
M
LN
IE M
LP
N
IE LP
N
MMM RBMK_GDH_RUPTURE
IM P
N
M
L
IF httemp244501812
P
N
M
L MP
N P
N
F
D
C
B D
C
BG
F E
D
C
BF E
D
C
BG IC
DF
E L
IF
DG
E M
F
E
O
#
H O
#
HA #
HO
NNN RBMK_GDH_RUPTURE#
HA
O B
#O CA
O D
httemp244501912
B O
540 J
V
Y J
V
YZ J
V
YX J
V
Y
XZ H
J
V
Y #
H
J
VZ #
H
X X
PPP RBMK_GDH_RUPTURE X YX Y
X
httemp244502012
520
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
206
WinGraf 4.1 - 08-19-2005
640
620 Y
Z
XY
VY
Z XXX RBMK_GDH_RUPTURE httemp214400501
X YYY RBMK_GDH_RUPTURE httemp259400501
600 ZZZ RBMK_GDH_RUPTURE httemp265400501
Y
Z
Temperature (K)
Y VVV RBMK_GDH_RUPTURE httemp669400501

Y V Z
Y JJJ RBMK_GDH_RUPTURE httemp175400501
X VV V V HHH RBMK_GDH_RUPTURE httemp244400901
X Y
580 Y V Y
Z Z
J Y X V Y
X X V Y
Z
V
JXZY J
H J
H J X
J
V
X
J
H
560 X H HXH
XJVYH
XX
JJ
VY
V H H
HHJV JX H
JX H Y
V
J
H
JX H H H H H H
540 ZY
JV
X JV
X
Y JV
ZX
Y JX
Y
V Z
JV
Y
X Y
X
520
0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0
Time (s)
Fig. 151 – PT Hot Spots temperatures
207
5.2. Conclusion
A wide spectrum of combined 3D NK TH analyses by the RELAP5-3D code were
executed. The classification of the accidents was the following:
• Reactivity Initiated Accident (RIA): CR withdrawal and CR group
withdrawal
• Decrease of coolant event: Single FC blockage, GDH blockage
• LOCA: GDH LOCA, CPS cooling circuit LOCA
All of the considered accidents were DBA, with the exception of CPS cooling circuit
LOCA. The results showed that in the case of CR withdrawal and CR group
withdrawal any safety threshold was overcome as well as during the GDH LOCA.
This because of some intrinsic good qualities of the RBMK design, e.g.:
• In the case of GDH LOCA, a large water inventory is available in the DS,
allowing a cooling by a backflow of the fuel rods
• CR are located in a circuit that is physically separated from the Primary
Side. Therefore, they are operating in a low pressure environment (i.e,
namely atmospheric pressure) that excludes their rapid ejection, as in the
case, e.g., of a PWR
On the other hand, two safety barriers are lost during the FC blockage event and
during the GDH blockage. Fuel rods and PT are overheating during the FC
blockage, leading to the release of radioactivity into the ALS. The Monte Carlo
analyses allowed to analyze the physic of the neutronics phenomena involved
during a FC voiding (increased leakages, Doppler effect). The
GDH blockage is an accident leading to a sharp coolant flow reduction in all the
channels (generally 42 FC) of the affected GDH. Flow oscillations are induced and
in some FC this could lead to PT and fuel rods overheating, with their consequent
disruptions and release of radioactivity in the MCC and the ALS if any scram signal
is actuated in time. The modeling of all the single FC connected to the affected
GDH is mandatory in order to derive acceptable conclusions for a licensing safety
analyses [82].
Last, the CPS cooling circuit LOCA is a BDBA accident leading to a rapid insertion
of positive reactivity. The installation of dedicated scram rods (the safety rods) not
affected by the LOCA is avoiding dangerous consequences for the reactor.
208
6. THE EXTREME CASE
6.1. Introduction
A comprehensive safety analysis of the RBMK system should not preclude the
calculation of events occurring at low power. This because of the particular
features of this reactor that can lead to significant changes of reactivity coefficients
and in some particular situations to the neutronic decoupling of some zones of the
core (see Chapter 2 of this document for a description of the neutronics
characteristics of RBMK). The notorious event of Chernobyl sadly demonstrated
the vulnerability and the instability of these systems when brought to low power.
However, it should be stressed that after that event, extensive hardware and
procedure modifications were implemented in order to minimize or avoid some
dangerous situations. The reference RBMK unit studied during this PhD thesis
(Smolensk-3 NPP) already included some important changes that enhanced its
stability at low power (e.g., U enrichment at 2.4%, higher number of AA, the new
design of CR). Nowadays, in order to further improve the reactor safety levels (e.g.,
reducing the positive void effect), all RBMK in Russia operates with fuel having an
enrichment of 2.8%, in conjunction with the use of Erbium as burnable poison.
Thus, the analyses that are showed here applies to an intermediate step of the
evolution of the RBMK configuration.
6.2. The Chernobyl event

We reported here a brief overview of the Chernobyl accident, in order to get the
reader familiar with the quite complex chains of events that caused the reactor
destruction. The main source document for the reconstruction of this event was the
IAEA INSAG-7 report [83]. Other sources were also [84], [85].
The main aspects that differed the Chernobyl-4 from the Smolensk-3 here
analyzed are the following:
• fuel enrichment at 2.0% instead of 2.4%;
• small number of Additional Absorber,
• As a consequence of the previous bullets, a greater void coefficient of
reactivity (total vaporization of the coolant was estimated in +4-5 b)
• Different design of CPS. In particular:
o MCR with a shorter graphite follower that allowed to some water to
be present on the bottom of the CR channels when the CR itself
was completely extracted (see Fig. 152)
o Greater time for the insertion of CR (18 seconds)
The accident occurred during the execution of a turbine trip test at low power, in
the first hours of the 26 April 1986. The test was conducted in order demonstrate
the capability of the turbo-generator to produce enough power, during the run-
down, for the unit’s internal requirements. In particular, the test should have had
demonstrate the capability to power the MCP and the FW pumps.
209
The run-down concept was accepted and included in the design of NPP equipped
with RBMK reactors. For example, it was reported in the Technical Safety Report of
Smolensk that “.[..] during a DBA involving the total loss of power for the unit’s
internal requirements, cooling water should be fed to the damaged part by FW
pumps powered by the turbo-generator rundown” [86].
This kind of test was already performed at other RBMK units, considering it as a
merely electrical test. Instead, what resulted was that these tests should have had
to be considered as complex unit tests, and should have had involve the General
Designer, the Chief Design Engineer, the Scientific Manager and the Regulatory
Body.
The Chernobyl 4 test differed from the other tests and that contributed to the
accident were the following conditions:
• Reduced power (200 MWth instead of the prescribed 700 MWth)
• Extreme Xenon poisoning of the core: the test was delayed by the electric
grid manager and a wrong operator action caused a sudden reactor power
reduction (to 30 MWth) for several minutes (see Tab. 59)
• Complete extraction of several MCR, with consequent reduction of ORM
• 4 MCP in operations per side, with a decreased level of subcooling at the
inlet of the core
Fig. 152 – MCR configuration of Chernobyl-4 reactor
The sequence of the events for the Chernobyl-4 accident, as it was recorded by the
plant instrumentation, is given in report Tab. 59. This was used as reference for the
transient reconstruction (see the following paragraph).
210
Tab. 59 – Chernobyl-4 accident, sequence of the events
Time Event
25 April 1986
01:06 Start of reactor power reduction; ORM equals 31 MCRs
Start of replacement of the nitrogen—helium gas mixture with
03:45
nitrogen in the gas cooling system for the reactor graphite stack
03:47 Reactor thermal power is 1600 MW
Sequential measurement of the control system parameters and
from 04:13
vibration characteristics of turbo-generator No. 7 and turbo-
until 12:36
generator No. 8 at constant thermal power of 1500 MW
07:10 ORM equals 13.2 manual control rods
13:05 Disconnection of turbo-generator No. 7 from the system
Disconnection of the ECCS from the multipass forced circulation
14:00
circuit (MFCC)
Postponement of testing programme requested by Kiev power
14:00
grid controller
15:20 ORM equals 16.8 manual control rods
Power supply to auxiliary equipment not involved in the tests
18:50
switched to working transformer No. T6
23:10 Power reduction continued, ORM equals 26 manual control rods
26 April 1986
00:05 Reactor thermal power was 720 MW
At reactor thermal power of about 500 MW transfer made from
the local to global main range automatic power control
(automatic power controllers Nos. 1 and 2). During the transfer
00:28
there was a reduction in thermal power to 30 MW (neutron
power to zero), which was not envisaged in the testing
programme. Measures to increase the power were taken
00:34:03 Emergency fluctuations of water level in steam separator drums
The EPS trip point in response to a pressure drop in the steam
00:36:24
separator drums was changed from 55 to 50 kg/cm2
Diagnostic Parameter Recording Program (DREG) did not work
from 00:39:32
Personnel blocked the EPS signal which would have shut down
until 00:43:55
the two turbo-generators
from 00:41 Disconnection of turbo-generator No. 8 from the system to
until 01:16 determine the vibration characteristics during rundown
01:03 Reactor thermal power increased to 200 MW and stabilized
Seventh main circulating pump was put into operation (MCP No.
01:03
12)
01:07 Eighth MCP was put into operation (MCP No. 22)
from 01:12:10 DREG program did not work
until 01:18:49
211
Tab. 59(cont.) – Chernobyl-4 accident, sequence of the events
Time Event
26 April 1986
from 01:19:44 'One overcompensation upwards' signal on
until 01:19:57
The parameters were recorded on magnetic tape (calculations
were performed at the Smolensk plant after the accident using
01:22:30
the PRIZMA program; ORM proved to be equal to 8 manual
control rods)
'Oscilloscope is on' signal was given; emergency stop valves of
turbo-generator No. 8 were closed. The rundown was started of
01:23:04
four MCPs: MCPs Nos. 13 and 23 (section 8RA) and MCPs
Nos. 14 and 24 (section 8RB)
01:23:10 DBA button was pressed
'One overcompensation upwards' signal went off (it lasted 3 min
01:23:30
33 s)
EPS-S button was pressed; the EPS rods and manual control
01:23:40
rods started to move down into the core
Power excursion rate emergency protection system signals on;
01:23:43 excursion period: less than 20 s; emergency power protection
system signals actuated; power exceeded 530 MW(th)
01:23:46 Disconnection of the first pair of MCPs being 'run down'
01:23:46.5 Disconnection of the second pair of MCPs being 'run down'
Sharp reduction in the flow rates (by 40%) of MCPs not involved
in the rundown test (MCPs Nos. 11, 12, 21 and 22) and
unreliable flow rate readings of the MCPs taking part in the
01:23:47 rundown (MCPs Nos. 13, 14, 23 and 24); sharp increase of
pressure and in the water level in the SDs; signals 'failures of
measuring systems' from both main range automatic controllers
(automatic power controllers Nos. 1 and 2)
Restoration of flow rates of MCPs not involved in the rundown
test to values close to the initial ones; restoration of flow rates to
15% below the initial rate for the MCPs on the left side which
were being 'run down'; restoration of flow rates to 10% below the
01:23:48 initial rate for MCP No. 24; unreliable readings for MCP No. 23;
further increase of pressure in die steam separator drums (left
side 75.2 kg/cm2, right side 88.2 kg/cm2) and of water level in
the steam separator drums; triggering of fast acting systems for
dumping of steam to condensers Nos. 1 and 2
Emergency protection signal 'Pressure increase in reactor space
(rupture of a fuel channel)'; 'No voltage — 48 V signal (no power
supply to the servo-drive mechanisms of the EPS); 'Failure of
the actuators of automatic power controllers Nos. 1 and 2'
01:23:49
signals. From a note in the chief reactor control engineer's
operating log: "01:24: Severe shocks; the RCPS rods stopped
moving before they reached the lower limit stop switches; power
switch of clutch mechanisms is off”
212
From the review of data and from the successive simulations it was deduced that
the main causes of the accidents were the followings. The reactor, through the
series of events that preceded that test, was brought to an instable point,
characterized by:
• Complete extraction of most of the MCRs, with consequent reduction of the
ORM
• Xenon poisoning of the core with axial flux characterized by a concave
shape, with two humps, one in the bottom and one in the top of the core
(see Fig. 153)
• Low level in SD
• 4 MCP per side in operation instead of two per side
• Coolant temperature close to saturation at the core inlet
Fig. 153 – Chernobyl axial distribution of the thermal neutron flux density at h.22:00
(curve 3) and h. 00:30 (curve 4) before the accident [83]
The actuation of the scram caused a positive reactivity insertion because of the
particular design of the MCR (see Fig. 152). A further positive reactivity insertion
was introduced by the coolant vaporization bringing the reactor to a power
excursion that was anymore possible to compensate by the CPS. The rupture of
several fuel channels with a consequent cavity pressurization, top reactor plate
movement and air ingress into the reactor cavity caused several violent chemical
reactions (e.g., C+O2, H2+O, C+H2O, see [85]) that led to an explosion and a
destruction of the reactor.
213
6.3. Literature review
The problem of simulation of the Chernobyl-4 accident was immediately addressed

in the aftermath of the event. Unfortunately, at that time, coupled codes
calculations were not available because of the lack of sufficiently robust and
validated codes and because of the lack of computational power for dealing with
simulations of such a large system.
Therefore the analyses were focused on calculating the thermal-hydraulics
phenomena with simplified 0-D neutron kinetics model. An example of these works
can be found in [87]. Neutronics calculations were instead effectuated using nodal
codes with simplified TH models [88], [93] or performing static analyses using
Monte Carlo simulations [91].
The fields of investigation were the followings:

• Assessment of the positive scram reactivity
• Assessment of the reactivity introduced by the core voiding
• Calculations of the plant dynamics, including MCP cavitations
The outcomes from all these investigations demonstrated the positive scram
effects introduced by the old design of MCR (see Fig. 152). Different axial power
shapes and different burnup distribution were investigated. The burnup distribution
did not affect so much the positive scram effect. What was found to be an
important parameter was the flux axial shape: concave axial shape (neutronic
decoupling) was found to be the main cause of the positive reactivity insertion by
scram. On the other hand, convex axial shapes resulted in a negative reactivity
insertion by scram [89], [90]. Reactivity insertion values were estimated to be super
prompt critical (r> 1 b, values between 1-1.25) [91].
The core voiding, instead, was estimated to cause a greater reactivity insertion,
from 2 b up to 5.6 b [95]. The causes of the core voiding, including the pump run-
down effects and the eventual pumps cavitations were also investigated by several
TH codes [92]. The conclusions were that the main contributions to the core
voiding were given by the pump rundown and not by pumps cavitation.
6.4. Low Power analyses – the model upgrade
A transient like the Chernobyl 4 was simulated using the developed Smolensk-3
model. In order to minimize the differences between the simulation and the
accident, it was decided to start the calculations from the power reduction
sequence (see Tab. 59, h.23.10 of 25 April 1986). These required the upgrade of
the RELAP5-3D model with the inclusion of Xenon microscopic cross sections and
of a power control system for operate the reactor power descent according to the
available data.
214
6.4.1. Xenon modeling
The capability to calculate a low power scenario could imply the code capability in
handling a Xenon transient. A recent version of the RELAP5-3D code (e.g.,
RELAP5-3D v.2.4.2) allows modeling the microscopic cross section of Xenon and
its effects on different elementary fuel cells. In order to speed up calculations it was
assumed an average value of the microscopic cross section for all the fuel cell
type, independently of the fuel cell burnup value.
Thus, RELAP5-3D code calculated the steady-state Xenon number densities and
upgraded it during the transient calculations (e.g., see Fig. 157).
6.4.2. Power Control modeling

A simplified RBMK power control system was developed and implemented. The
objective was to create a control logic able to move a CR bank and to obtain the
desired power history. The system should also be able to operate a scram signal
and to compensate automatically the Xenon build-up or reduction. The scheme of
the control logic is given in Fig. 154.
PROGRAMMED
POWER (cntrlvar)
-
S ε VCR = k*ε Lag

+
REACTOR
POWER (rkotpow)
CR pos. (i) = CR pos(i-1) + VCR*∆t
Fig. 154 – CR bank 2 actuation logic
The control logic is generating a signal operating on the CR bank velocity, with
imposed limits for the minimum and maximum CR extraction as well as limits for
the minimum and maximum velocity. An external general table gives as input the
desired power history (power versus time). The control logic is calculating the error
between the desired power and the actual reactor power and using this error for
generating a proportional signal for the calculation of the velocity of the CR. A lag is
used for assuring the control system stability.
Therefore the CR positions are updated at each time step calculating the obtained
CR velocity (see Fig. 154).
The control logic and its interaction with Xenon build-up was thoroughly tested in
several conditions and on different reactor core design (VVER-440) [96].For the
Smolensk-3 reactor, the developed CR logic was applied to a group of 22 MCR, all
inserted at the same depth. This assured enough reactivity for the Xenon transient.
215
6.5. The Power-reduction transient
In order to reproduce as well as possible the status of the Chernobyl-4 reactor
before the accident, a preliminary power-reduction transient was executed, using
as imposed events those reported in Tab. 59. The transient was aimed to obtain
the following reactor core conditions:
• Power reduction from 1600 MWth to 200 MWth
• Xenon build-up
• 3 MCP per side running in stable conditions
• The maximum number of MCRs extracted (reduced ORM)
• A stable level in the SDs
Reactor power was reduced from 50% of Nominal Power (i.e., 1600 MWth) to
roughly 500 MW in 90 minutes (see Tab. 59 and Fig. 155). Then, after a sudden
power reduction to low power values (50 MWth) for 5 minutes, the reactor power
was increased and stabilized around 200 MWth in half an hour.
x 10 9 WinGraf 4.1 - 11-19-2008
1.75
XX
1.5
X XXX CH1d rkotpow0
X
1.25
Reactor Power (W)
1 X
X
X
.75
X
X
.5 X
.25
XX X
X
X X X
0
-.25
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0 1000.
Time (s)
Fig. 155 – Simulation of Chernobyl event in the Smolensk NPP: Reconstructed-

imposed reactor power before the event start (t = 1240 s in the scale above)
The MCRs operated by the power control system were at the beginning inserted
and then slowly extracted for compensating Xenon buildup. At around time
t=+4680 sec. (corresponding to the 00:30 of the 26 April 1986), they were
completely inserted, together with the Safety Rods, in order to simulate the abrupt
power reduction that occurred during the Chernobyl-4 accident and the further
Xenon concentration increase (see Fig. 156 and Fig. 157).
216
WinGraf 4.1 - 11-19-2008
9.00
8.00
7.00 XXX CH4 CR XX
6.00
CR Insertion (m)
5.00
X
4.00 X
X
X
3.00
2.00 XX X X X X X X X X X
X X X
1.00 X
X X
0
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 156 – Simulation of Chernobyl event in the Smolensk NPP: MCR operated by
power control logic - insertion
x 10 13 WinGraf 4.1 - 11-19-2008
9
XXX CH4 conxe30259
YYY CH4 conxe31374
ZZZ CH4 conxe31499
8 VVV CH4 conxe31587
JJJ CH4 conxe31911
HHH CH4 conxe100259 Z
Z
### CH4 conxe101374 Z
7 OOO CH4 conxe101499 Z
Xe Concentration (a/cm3)
AAA CH4 conxe101587 Z Z

B
B
H
Z Z
B H Y
BBB CH4 conxe101911 Z Y H Y
Z Z Y B
H X
6 Z YBYBY B
H X X
Z H H
Z Z
Z Y B YB X J J
Z Z Y B YBY B H HX X X J V J V V V
Z Z Z Y Y H H X X J J
Y Y Y Y Y H X H X X J JV V V O
5
Y Y Y
H H H XHX J J J V J
V V O O
A
O
A A
X X X
J J J J V O OO A
XX X X X X
J J V V OVO V O V O O
A A AA
V V V V A A A O
O A A
# # # #
# # # # # # # # # # # #
4
2
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 157 – Simulation of Chernobyl event in the Smolensk NPP: Xenon buildup
The power was then stabilized at 200 MWth in half an hour by extracting again
Safety Rods, the MCR operated by the power control system and other 10 MCRs
already completely inserted (see Fig. 155 and Fig. 156). 3 MCP per side were in
operation, in stable conditions. This was achieved by increasing the pressure
losses at the MCP exit, simulating the throttling of the regulation valve (see Fig.
158).
217
WinGraf 4.1 - 11-19-2008
2250
2000
1750
1500
X
Z Y X
Y
ZH X
Y
Z H
J Z
XX
Y Y
X ZYV X
XZ Y V
Z JX ZH
Y V XZ
JY X
Y
J Z JV
JH V H J V V X
Y X
Y
ZH X
Y
Z H X
Y
Z H X
Y
ZH
JZ XX
VYYZJ V X
Y
Z H X
Y
Z H X
Z Y X
Y
Z H
Mass Flow (Kg/s)
V H X
Y
V JV
H J V J V JV JV
HH J V J V JH V J
1250
1000
XXX CH1d mflowj33010000
750 YYY CH1d mflowj43010000
ZZZ CH1d mflowj53010000
VVV CH1d mflowj533010000
500 JJJ CH1d mflowj543010000
HHH CH1d mflowj553010000
250
-250
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 158 – Simulation of Chernobyl event in the Smolensk NPP: MCPs mass flow,
RHS and LHS
A sufficient water level in the SDs was also achieved (Fig. 159) by some manual
FW injections in order to avoid the risk of MCPs cavitation (Fig. 160).
WinGraf 4.1 - 11-19-2008
180
160
XXX CH4 cntrlvar401
YYY CH4 cntrlvar403
140 ZZZ CH4 cntrlvar405
VVV CH4 cntrlvar407
SD Liquid Volume (m3)
120 V
Z V
Z
V
Z X
X Y X X X
Y
100 V
ZX X X X V X V Y
Z Y X
V V V V XX Y
Y Y Y Y
Z Z Z
V V
Z VZ Y Y Z Y Z Z
Z Z Z X Y V V Z
80 XX Z V Z Z V V V V
Y X X X X X X Z
X Y
Y Y Y Y Y Y Y
60
40
20
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 159 – Simulation of Chernobyl event in the Smolensk NPP: SD liquid level
218
WinGraf 4.1 - 11-19-2008
250
200 XXX CH4 mflowj382000000

XY
X
X
Y YYY CH4 mflowj432000000
ZY
XV ZZZ CH4 mflowj882000000
Z
X
Y VVV CH4 mflowj932000000
V
Z
150 X
Y X V
Y XV
Z YZ
V
ZY
XV
Mass Flow (Kg/s)
Z
X
Y
V
Z
100 X
Y
V
Z
X
Y
V
Z Y
XV
ZY
X Z
V Y
X Z
V Y
XV
ZY
X
50
V
Z
X V
Z Y
X
Z Y
X V X V
Z Y V
Z
XX
YY
V Z Y
Z V
0
-50
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 160 – Simulation of Chernobyl event in the Smolensk NPP: FW mass flow per
SD, LHS and RHS
WinGraf 4.1 - 11-19-2008
.70
.60 XXX CH1d voidg307200000

YYY CH1d voidg308200000
XY
X Y ZZZ CH1d voidg807200000
X
.50 Y Y
ZXZ X Z Y VVV CH1d voidg808200000
X Y HZ Y
X
V V J Z X X
H Y
H
JZ
VJ V J # JJJ CH1d voidg809200000
V HHH CH1d voidg810200000
X
Z Y
H X X Y
Y
Y Z H X ### CH1d voidg320200000
.40 O
#
V Z
H
O
# ZH X
Y
HOOO CH1d voidg322200000
Z
Void Fraction
V AO
J X
JO
# Z X
H
Y # A
VJ O# Z Y
AAA H CH1d
X
Y
H voidg323200000
VA V JA O Z
#
.30 J O V JAO
V
# A #
VJA
J
.20
H
X
Y
H Z A
X
Y
H Z
X
Y A V J
O
#
.10 V
O
# J
X Z
H
Y V JA
JA O
#
V H
XO
ZYY H
X #
ZJ Z
0 O
V
# JAA
# O
V
-.10
-7000. -6000. -5000. -4000. -3000. -2000. -1000. 0
Time (s)
Fig. 161 – Simulation of Chernobyl event in the Smolensk NPP: Void fraction at the
core outlet, right and left side
The obtained reactor state was then used for the transient Chernobyl-like accident
simulation.
219
6.6. The “Chernobyl-like” event
The sequence of events that were implemented in the RELAP5-3D code for
reproducing the Chernobyl-4 event are given in Tab. 60. The correspondence
between the RELAP5-3D transient time and the accident sequence of the events is
given.
Tab. 60 – Sequence of the events (actual versus reconstructed)
Time transient Time (s)

ACCIDENT EVENTS RELAP EVENTS
(26 April 1986) RELAP5-3D
Thermal Power Thermal Power
1.02.00 0.
Stabilized at 200 MW Stabilized at 200 MW
Opening Isolation
1.03.00 60. MCP-7 ON
Valves of MCP-63
1.03.05 65. MCP-63 ON
1.06.00 240. FW Adjustment N/A
Opening Isolation
1.06.55 295
Valves of MCP-563
1.07.00 300 MCP-8 ON MCP-563 ON
FW at 90 Ton/h for
FW transient values
1.09.00 420 RHS; FW at 180
(Reduction)
Ton/h for LHS
1.18.52 1012 DBA Signal N/A
1.22.30 1230
Test Beginning
Closure of SV on SV 397 Closed;
1.23.04 1264 Right side of MCC, MCPs 33,43,533,543
rundown of 4 MCP, 2 rundown
per side
EPS5 Button
Scram by MCR#1
1.23.40 1300 pressed;
and Safety CRs
CRs Inserted
1.23.43 1303
1.23.47 1307
1.23.48 1308
1.23.49 1309
1.24.00 1320 Reactor Explosion
1.28.04 1564 End Of Calculation
In the following pages, the main results are reported. A couple of sensitivities were
also run. In the first one a positive reactivity perturbation was introduced at the
bottom of the reactor core by the extraction of the shortened CR, thus reproducing
a key phenomenon of the Chernobyl transient. In the second sensitivity, no scram
signal was activated in order to assess the dynamic of the plant.
220
6.6.1. Reference Transient
In the following figures the trends of the main parameters are reported. The reactor
power before the test is kept stable at 200 MWth by the power control system. CR
are compensating Xenon buildup (see Fig. 162 and Fig. 163).
x 10 8 WinGraf 4.1 - 11-19-2008
2.2
2 X X X X X X
1.8
1.6
XXX CH1d rkotpow0
Reactor Power (W)
1.4
1.2
.8
.6
.4
.2
0 200. 400. 600. 800. 1000. 1200.
Time (s)
Fig. 162 – Simulation of Chernobyl event in the Smolensk NPP: Reactor Power
before the Test
WinGraf 4.1 - 11-19-2008
8.00
7.00
6.00
5.00
CR insertion (m)
4.00
3.00 XXX CH1d CR
2.00
X X X X X X
1.00
-1.00
0 200. 400. 600. 800. 1000. 1200.
Time (s)
Fig. 163 – Simulation of Chernobyl event in the Smolensk NPP: CR before the test.
221
At 1.03.05 and 1.07.00 two MCP, one per side, are turned on (Fig. 164). This is
increasing the mass flow rate per each circuit, redistributing the mass flow rate in
each MCP (Fig. 164 and Fig. 165).
WinGraf 4.1 - 11-19-2008
2000
1800
1600
1400
XX
Mass Flow (Kg/s)
1200
X
Y X
Y X
Y X
Y
Z X
Y
Z
1000
800
600 YYY CH1d mflowj43010000
400 VVV CH1d mflowj63010000
JJJ CH1d mflowj533010000
200 ### CH1d mflowj553010000
OOO CH1d mflowj563010000
0
-200. 0 200. 400. 600. 800. 1000. 1200.
Time (s)
Fig. 164 - Simulation of Chernobyl event in the Smolensk NPP: MCP activation
WinGraf 4.1 - 11-19-2008
5500
5000
X X X X X
4500 Y Y Y Y
XXX CH1d Left

Mass Flow (Kg/s)
4000 X
YYY CH1d Right
3500
3000
2500
2000
1500
0 200. 400. 600. 800. 1000. 1200.
Time (s)
Fig. 165 – Simulation of Chernobyl event in the Smolensk NPP: Left and Right side
Mass Flow (Kg/s)
At 1.09.00 a FW perturbation is operated, reducing the mass flow on the right side
of the MCC (Fig. 166).
222
WinGraf 4.1 - 11-19-2008
60.0
50.0
40.0

Mass Flow (Kg/s)
YYY CH1d mflowj432000000

30.0 ZZZ CH1d mflowj882000000
X X VVV X
Y
X YCH1d mflowj932000000
Y
X Y
X
20.0
Z
10.0
-10.0
0 200. 400. 600. 800. 1000. 1200.
Time (s)
Fig. 166 – Simulation of Chernobyl event in the Smolensk NPP: FW perturbation
At 1.23.04 the test began, with the closure of stop valve for the RHS SD. Power
control system is deactivated and at the same time, the rundown of two MCPs per
side is operated (Fig. 167).
WinGraf 4.1 - 11-19-2008
140
120
Z Z V
Z Z Z Z Z Z Z ZZ
#
100
V
Pump velocity (rad/s)
80
V
60 X
XXX CH1d pmpvel33 V
YYY CH1d pmpvel43
40 ZZZ CH1d pmpvel53 X
V
VVV CH1d pmpvel63 X
Y
JJJ CH1d pmpvel533 V
X
Y
20 HHH CH1d pmpvel543 V
### CH1d pmpvel553 X
Y
X
Y V
OOO CH1d pmpvel563 X
Y
0 X Y
Y X Y
X JY
X JY
X HX H
JY X H
JY XX
JJ
Y
-20
1200.0 1220.0 1240.0 1260.0 1280.0 1300.0 1320.0 1340.0 1360.0
Time (s)
Fig. 167 – Simulation of Chernobyl event in the Smolensk NPP: MCPs speed
223
WinGraf 4.1 - 11-19-2008
2000
1750 VV
ZZZ
##
ZZZ
ZZ
ZZZZV
1500
1250
V
Mass Flow (Kg/s)
1000
750 YYY CH1d mflowj43010000 X V
500 VVV CH1d mflowj63010000
JJJ CH1d mflowj533010000 V
250 ### CH1d mflowj553010000 X V
OOO CH1d mflowj563010000
0 VV
XX
XYYJX
XY
YXY JXJYXHHH
JJJJ
XH
YY
XXXXXY
XYYYYY
-250
-500
1000.0 1050.0 1100.0 1150.0 1200.0 1250.0 1300.0 1350.0 1400.0
Time (s)
Fig. 168 – Simulation of Chernobyl event in the Smolensk NPP: MCP flow rates
WinGraf 4.1 - 11-19-2008
5000
XX
4500
X
4000 Y
XXX CH1d left
YYY CH1d right
X
Mass Flow (Kg/s)
3500
Y
3000 XXXX
XXX
YYYY X
YY
Y
2500 X
Y
X
2000 Y
X
YX
Y
1500 XY
YYX
1000
600.0 700.0 800.0 900.0 1000.0 1100.0 1200.0 1300.0 1400.0 1500.0 1600.0
Time (s)
Fig. 169 – Simulation of Chernobyl event in the Smolensk NPP: Circuit Flow Rate
The reduction in the circuit mass flow rates (Fig. 168 and Fig. 169) is causing a
void increase that is leading to a power reduction. This is consistent with the new
neutronics characteristics of the RBMK system (see Chapter 2), because of the
negative void coefficients.
224
WinGraf 4.1 - 11-19-2008
.25
.23
XXX CH1d voidg307200000
YYY CH1d voidg308200000 X
XYXX
.20 ZZZ CH1d voidg320200000 Y YYX
YX
Y
VVV CH1d voidg322200000
JJJ CH1d voidg323200000 X
HHH CH1d voidg807200000 Y X
Y
.18
### CH1d voidg808200000 X Z V
XXY ZVZV
Z Z
Void Fraction
OOO CH1d voidg809200000

.15 AAA CH1d voidg810200000 JV
ZX
Y
.13 JVH
X
YH
.10 Z X H
JVY H
X##
ZJVY
.08 X
YX
ZJV
ZJVJ
.05 ZZ
.03
1000.0 1050.0 1100.0 1150.0 1200.0 1250.0 1300.0 1350.0 1400.0
Time (s)
Fig. 170 – Simulation of Chernobyl event in the Smolensk NPP: Void Fraction at
core exit during the transient
x 10 8 WinGraf 4.1 - 11-19-2008
2.5
2.25
2 X
X
1.75 X
XX
XX
1.5
Power (W)
XXX CH1d rkotpow0

1.25
.75
X
XX
.5 X X XX
XX
X
.25
0
1050.0 1100.0 1150.0 1200.0 1250.0 1300.0 1350.0
Time (s)
Fig. 171 – Simulation of Chernobyl event in the Smolensk NPP: Power during the
transient
The transient is then terminated by the actuation of the scram signal. The new
configuration of the MCR is anymore introducing any positive reactivity, thus
leading the reactor in safe shutdown conditions.
225
6.6.1.1. Positive Reactivity Perturbation
This sensitivity was run in order to assess the effects of a positive reactivity
insertion during the first phase of the test. This was caused in Chernobyl by the
particular design of the MCR (see Fig. 152). In the current configuration of the
RBMK, positive reactivity could be introduced in the bottom of the reactor by the
Shortened CR only. These CR are not moved during any transient. Therefore,
these results should be looked as a pure sensitivity test that should not happen
under any circumstance during the real plant operation. Results of calculations with
extraction speed of 0.5, 0.25 and 0.1 m/s are given in Fig. 172 and Fig. 173.
x 10 9 WinGraf 4.1 - 12-04-2008
X XXX TR_0.1 rkotpow0

4 X XX YYY TR_0.25 rkotpow0
X XXXX XX
XXX X ZZZ TR_0.25 rkotpow0
Power (W)
X
2
1
X
-1
1200.0 1250.0 1300.0 1350.0 1400.0 1450.0 1500.0 1550.0 1600.0 1650.0
Time (s)
Fig. 172 – Sensitivity with introduction of additional positive reactivity: Reactor

Power
WinGraf 4.1 - 12-04-2008
1.00
.80
XXX TR_0.25 cntrlvar516
YYY TR_0.5 cntrlvar516
.60 ZZZ TR_0.1 cntrlvar516
.40 X
Reactivity ($)
.20 ZZ
ZZZ
ZZZZZ
Z
0
-.20
-.40 Z
-.60
1200.0 1250.0 1300.0 1350.0 1400.0 1450.0 1500.0 1550.0 1600.0
Time (s)
Fig. 173 – Sensitivity with introduction of additional positive reactivity: Core

Reactivity
226
6.6.1.2. No Scram actuation
This sensitivity considered the execution of the test without the intervention of any
scram signal. The power, after the reduction caused by the core voiding, was easily
controlled by the power control system.
x 10 8 WinGraf 4.1 - 11-19-2008
2.5
XXX CH3 rkotpow0
2 X
X
X
XXXXXXXXXXXXX
Reactor Power (W)
1.5
X X
.5
-.5
-1
1000.0 1100.0 1200.0 1300.0 1400.0 1500.0 1600.0
Time (s)
Fig. 174 – Reactor Power without scram actuation
6.6.2. Conclusions
The calculations showed that the neutronics and hardware modifications
introduced after Chernobyl accident were effective in avoiding such kind of events.
In particular, the two main phenomena that triggered the reactor explosion at that
time were eliminated. In fact :
• the new design of MCR avoid the possibility to obtain concave flux shapes
and a positive reactivity insertion during the first phase of the scram;
• the elimination of the positive void coefficients avoids the possibility to get
a power excursions when operating the reactor at low power, i.e., with a
reduced margin of coolant subcooling
The upgrades that were introduced during the recent years on the RBMK (e.g., the
increase of U enrichment at 2.8%, the cluster MCR) and that were not present in
the Smolensk-3 configuration analyzed here, are further increasing the safety
levels of these type of reactors for transients like the Chernobyl-4 one.
227
7. CONCLUSIONS
The application of a “chain of codes” to the analysis of transients in RBMK

Smolensk 3 NPP aimed at investigating coupled 3D NK TH phenomena in the core
has been presented. The capabilities of the adopted codes to handle the resulting
complex scenarios have been demonstrated, though it was not the purpose of the
PhD activity to perform licensing calculations.
The key conclusions can be summarized as follows:
a) a pioneering application of the chain of codes RELAP5-3D-

HELIOS/DRAGON-MCNP5 to the RBMK safety technology has been
successfully effectuated;
b) detailed input decks for RBMK system have been developed for the
coupled codes and are available to the GRNSPG/UNIPI. In the case of
RELAP5-3D, the nodalization includes up to 4000 hydraulic nodes, 50000
meshes for conduction heat transfer and about 40000 meshes for the
neutron kinetics model;
c) problems were founds in calculating the correct values of void coefficients
using both deterministic neutron transport codes DRAGON and HELIOS.
The results showed a sensible under-estimation of the void reactivity.
Calculations by the Monte Carlo code MCNP5 was instead not affected by
this problem. The weight of this problem on the system code calculation
uncertainties should be evaluated in future works;
d) a key identified problem (confirming what found for TH analysis of Ignalina,
see [97]) in case of GDH-Blockage event, was the (calculated) occurrence
of inlet flow oscillations in affected FC. This caused temperature
excursions in some FC bundles and can be the precursor of severe core
damage;
e) mild and controllable power excursions were calculated in the case of
spurious withdrawal of single and of bank-of CR (CR-withdrawal and CR-
group withdrawal, respectively);
f) the CPS-LOCA event brings to scram caused by high average core power
and safety functions of the system are not challenged with main concern to
SD pressure and rod surface temperatures;
g) The analysis of the single FC-Blockage event confirmed that the affected
rod bundle power tends to decrease following the lack of cooling event.
This is due to the Doppler effect and to the increased neutron leakages
that overpasses the effect of loosing coolant. Several calculations for this
kind of transient were also run by the use of MCNP5 code. The influence of
3D NK in predicting the overall system performance following individual
channel blockage is negligible. Namely the power generated in the affected
channel does not increase, rather a decrease has been calculated;
h) A detailed simulation of the Chernobyl-type accident was performed. This
kind of analyses required the integration of a suitable Xenon cross section
model in RELAP5-3D and the implementation of a simplified power control
228
system. The results showed that the Smolensk-3 configuration of RBMK
was safe also during this low power test, avoiding the positive reactivity
excursion that led the Chernobyl reactor to a destruction.
229
230
REFERENCES
[1] E. Adamov et alii, “Status and prospects for pressure-tube water-cooled

graphite-moderated reactors”, Nuclear Engineering and Design, Pages 59-
66, Vol. 173, 1997.
[2] N.M. Sorokin, B.A. Gabaraev, Yu.M. Cherkashov, “Safe operation and life
extension of RBMK plants”, Nuclear Engineering and Design, Pages 1648-
1656, Vol. 236, 2006.
[3] A. Kaliatka, E. Knoglinger, M. W. Jankowski, D. Mazzini, A. M. Moskalev,
S. Soloviev, E. Ušpuras, “Review of the Safety Needs that are Related to
the Project, and of the Relevant Physical Phenomena”, TD.B.1.2-DER/1
rev 1, 27 May 2004.
[4] M. W. Jankowski, J. Misak, A. Kaliatka, E. Ušpuras, A. M. Moskalev, Ju. A.
Migrov. “Thermal Hydraulics Code for the Primary Loop and the
Containment System”, TD.B.2.1-DER/1, July 2004.
[5] C. Parisi, V. Malofeev, “3D Space-Time Neutron Kinetics Code”, TD.B.2.2-
DER/1, 06 April 2005.
[6] Power Reactor Information System (PRIS), IAEA Website,
http://www.iaea.org/programmes/a2/
[7] S. F. Hall, B. A. Gabaraev, “Severe Transient Analysis for RBMK Reactor
TACIS Project R 2.30/94” – Ultimate Design Basis Accident, October 1997
referring to “RBMK Safety Review Final Report” RBMK/CMC/FR Issue 1,
March 1994.
[8] “Safety assessment of design solutions and proposed improvements to
Smolensk Unit 3 RBMK nuclear power plant”, IAEA, TECDOC-722,
Vienna, 1993.
[9] K. Almenas, A. Kaliatka, E. Uspuras, “Ignalina RBMK-1500 a source book”,
Extended and updated version, LEI – April 1998.
[10]E. Urbonavicius, A. Kaliatka, “Investigation of hydrogen control system
effectiveness in RBMK-1500 containment” Proceedings of ICAPP ’05
Seoul, KOREA, May 15-19, 2005, Paper 5422
231
[11] Various, “Control Rod Withdrawal” TACIS Project R 2.30/94 Severe
Transient Analysis for RBMK Reactors, October 1997
[12] “Safety margins of operating reactors: Analysis of uncertainties and
implications for decision making”, IAEA TECDOC 1332, Vienna, 2003.
[13] “Accident analysis for Nuclear Power Plants with Graphite Moderated
Boiling Water RBMK Reactors” Safety Reports Series No. 43, Vienna (A)
2005.
[14] J. P. Weber, G. Bava, D. Bestion, A. Maccabee, D. Reichenbach, “Final
Report of Topic Group 1: System Engineering and Accident Progression
(Appendix 11 of RBMK/CMC/FR)” Western topic Group 1 members, March
1994.
[15] Gosatomnadzor, “General Regulations for Nuclear Power Plant Safety
(ОПБ-88/97)”, GAN Report НП-001-97, 1998г
[16] “Basic Safety Principles for Nuclear Power Plants” IAEA, INSAG Series
No. 12, 75-INSAG-3 Rev. 1, Vienna, 1999..
[17] “Comparison of the Russian nuclear power safety concept contained in
OPB-88 and the next lower level norms/rules with the NUSS requirements”
Extra-budgetary Programme on the Safety of WWER NPP, WWER-RD-69,
Vienna, 1994.
[18] Gosatomnadzor, “Layout of nuclear stations. The basic criteria and safety
requirements”, GAN Report НП-032-01, 2002 г.
[19] “Probabilistic Safety Assessment”, Safety Series No. 75, INSAG-6,
Vienna, 1992.
[20] A. Kaliatka, E. Knoglinger, M. W. Jankowski, D. Mazzini, A. M. Moskalev,
S. Soloviev, E. Ušpuras, “Review of the Safety Needs that are Related to
the Project, and of the Relevant Physical Phenomena”, TD.B.1.2-DER/1
rev 1, 27 May 2004.
[21] N. Aksan, D. Bessette, I. Brittain, F. D’Auria, P. Gruber, H.L.O.
Holmstrom, R. Landry, S. Naff, R. Pochard, G. Preusche, M. Reocreux, O.
Sandervag, H. Staedtke, K. Wolfert, N. Zuber, “Code Validation Matrix of
Thermal-Hydraulic Codes for LWR LOCA and Transient” OECD/CSNI 132,
March 1987.
232
[22] N. Aksan, F. D’Auria, H. Glaeser, R. Pochard, C. Richards, A. Sjoberg, “A
Separate Effect Test Matrix for Thermal-Hydraulic Code Validation:
Phenomena Characterization and Selection of Facilities and Tests”
OECD/GN (94) 82, Vols. I and II, 1993.
[23] A. Annunziato, H. Glaeser, J.N. Lillington, P. Marsili, C. Renault, A.
Sjoberg “CSNI Code Validation Matrix of Thermal-Hydraulic Codes for
LWR LOCA and Transient”, OECD/CSNI 132, Rev 1, July 1996.
[24] X-5 Monte Carlo Team, “MCNP — A General Monte Carlo N-Particle
Transport Code, Version 5”, Los Alamos National Laboratory, LA-UR-03-
1987, April 24, 2003 (Revised 10/3/05)
[25] “Accident Analysis and its associated training programme for the RBMK
1000 Kursk 1 NPP”, IAEA Technical Report, Volume II, Rev. 3, July 2001,
Vienna, Austria.
[26] A. Jasulevicius et al., “Development and Validation of a Methodology for
Neutron Dynamics Analysis of RBMK Reactors”, Proceedings of the ANS
M&C 2001 Conference, Salt Lake City, USA, September 9-12 (2001).
[27] L. A. Hageman and D. M. Young, “Applied Iterative Methods,” Computer
Science and Applied Mathematics, Orlando: Academic Press, 1981.
[28] K. I. Derstine, DIF3D: A Code to Solve One-, Two-, and Three-
Dimensional Finite Difference Diffusion Theory Problems, ANL-82-84,
Argonne National Laboratory, 1984.
[29] R. S. Varga, Matirx Iterative Analysis, Englewood Cliffs, NJ: Prentice-Hall,
1962.
[30] S. K. Zee, Numerical Algorithms for Parallel Processors Computer
Architectures with Applications to a Few-Group Neutron Diffusion
Equations, Ph. D. Dissertation, North Carolina State University, 1987.
[31] K. S. Smith, “Nodal Method Storage Reduction By Non-Linear Iteration”,
Transactions of the American Nuclear Society, Detroit, MI, June 1983, 44,
pp. 265.
[32] K. S. Smith, “QPANDA: An Advanced Nodal Method for LWR Analyses”,
Transactions of the American Nuclear Society, June 1985, pp. 265.
233
[33] K. S. Smith and K. R. Rempe, “Testing and Applications of the QPANDA
Nodal Method,”Proceedings of the International Topical Meeting on
Advances in Reactor Physics, Pittsburgh, PA, 1987, 2, pp. 861.
[34] W. L. Weaver, “Software Design and Implementation Document; Three
Dimensional Neutron Kinetics for RELAP5/MOD3”, EGG-NRE-11021,
Idaho National Engineering Laboratory, November 1993.
[35] Edenius, M., K. Ekberg, B.H. Forssen and D.G. Knott, A Fuel Assembly
Burn-up Program User’s Manual, STUDSVIK/SOA-95/01 (1995).
[36] Edenius, M. and B. Forssen, CASMO-3, A Fuel Assembly Burn-up
Program, Users Manual, STUDSVIK/NFA-89/3.
[37] Watson, J., K. Ivanov, et al., “Cross-section Generation Methodology for
Three-dimensional Transient Reactor Simulations”, ANS TANSAO 77, p.
175 (1998).
[38] Watson, J. and K. Ivanov, “Improved Cross-section Modeling Methodology
for Coupled Three-dimensional Transient Calculations”, Annals of Nuclear
Energy, 29, 937-966 (2002).
[39] Langenbuch, S., et al., “Investigation of Modeling Aspects for Coupled
Code Application in Safety Analysis”, Proc. of the Int. Conf. on
Mathematics, Computation, Reactor Physics and Environmental Analysis
(M&C 2001), Salt Lake City, Utah, USA, September 2001.
[40] Ivanov, K. and A. Baratta, “Coupling Methodologies for Best Estimate
Safety Analysis”, Proc. of the Int. Conf. on Mathematics, Computation,
Reactor Physics and Environmental Analysis (M&C’99), Madrid, Spain, 27-
30 September 1999, Vol. 1, pp. 493-502.
[41] Langenbuch, S., et al., “Interface Requirements to Couple Thermal-
hydraulic Codes to 3-D Neutronic Codes”, OECD/CSNI Workshop on
Transient Thermal-hydraulic and Neutronics Codes Requirements,
Annapolis, MD, USA, Nov. 1996.
[42] Grundmann, U., et al., “Coupling of the Thermo-hydraulic Code ATHLET
with the Neutron Kinetic Core Model Dyn3d”, Proc. of the Int. Conf. on
Mathematics and Computations, Reactor Physics and Environmental
Analysis (M&C’95), Portland, OR, USA 30 April-5 May 1995, Vol. 1, pp.
257-263.
234
[43] “CRISSUE-S-WP2: Neutronics/Thermal-hydraulics Coupling in LWR
Technolog: State-of-the-art Report (REAC-SOAR)”, OECD/NEA, 2004.
ISBN 92-64-02084-5
[44] C. Parisi, M. Cherubini, W. Giannotti, B. Neykov “Smolensk-3 NPP
RELAP5 Thermal-Hydraulic Nodalization” TD.B.2.2-DER/1, Appendix A, 06
April 2005.
[45] IAEA, “Accident Analysis for Nuclear Power Plants”, Safety Report Series
No.23, Vienna, November 2002.
[46] NRC, “Standard Review Plan for the Review of Safety Analysis Reports
for Nuclear Power Plants”, NUREG-0800, June 1996
[47] S. N. Aksan, F. D'Auria, H. Städtke, “User effects on the thermal-hydraulic
transient system code calculations”, Nuclear Engineering and Design,
Volume 145, Issues 1-2, 2 November 1993, Pages 159-174
[48] D'Auria F., "Proposal for training of thermalhydraulic system codes users",
IAEA Spec. Meet. on User Qualification and User Effects on Accident
Analysis for Nuclear Power Plants - Vienna (A), Aug. 31-Sept. 4, 1998
[49] D'Auria F., Frogheri M., Marsili P., Giannotti W., “Standardised Procedure
for thermohydraulic system code assessment", Spring 1998 CAMP
Meeting, Ankara (TK); June 24-26, 1998, IAEA Spec. Meet. on User
Qualification and User Effects on Accident Analysis for Nuclear Power
Plants - Vienna (A); Aug. 31-Sept. 4, 1998, 2nd Int. Yugoslav Nuclear
Society Conf. (YUNSC ’98) – Belgrad (YU) Sept. 28, Oct. 1, 1998.
[50] “Enhancement of Lithuanian TSO`s Licensing Assistance Capability within
Core Integrity, Control Systems Upgrading and Management of Equipment
Ageing”, Task 1 Report “RBMK – Core Safety Surveillance”
SIP/RISKAUDIT-report No. 95684-R6, May 2004
[51] N. Alexeev et alii, “The Monte Carlo codes MCNP and MCU for RBMK
criticality calculations”, Nuclear Engineering and Design, Vol.183 (1998),
pages 287-302.
[52] F. D’Auria et alii, “Development of a Code System for Severe Accident
Analysis in RBMK” Final Technical Report for TACIS Project R2.03/97, part
B, , December 2005
235
[53] C. Parisi, V. Malofeev, K. N. Ivanov ,”Neutron Cross Section Model”, Task
2.3 Technical Report for TACIS Project R2.03/97, April 2005
[54] Studsvik™ Scandpower, “HELIOS Methods”, 2000
[55] S. C. Frankle, “Summary Documentation for the ENDL92 Continuous-
Energy Neutron Data Library (Release 1)” Los Alamos National Laboratory
internal memorandum, XTM:96-05, and report LA-UR-96-327 (1996).
[56]G. Marleau, A. Hébert, R. Roy. “A User Guide for DRAGON 3.05D”,
Technical Report IGE-174 Rev. 6D. Institut de genie nucléaire, Ecole
Polytechnique de Montréal, June 2006.
[57] International Atomic Energy Agency (IAEA). “Accident Analysis and its
associated training programme for the RBMK 1000 Kursk 1 NPP”.
Technical Report. Volume II. July 2001, Vienna (Austria).
[58] International Atomic Energy Agency (IAEA). “WIMS-D Library Update.
Final report of a coordinated research project”. Vienna, 2006
[59] C. Parisi, F. D’Auria “RBMK Fuel Channel Blockage Reactivity Analysis by
MCNP5, DRAGON and RELAP5-3D© codes”, International Conference
“Nuclear Energy for the New Europe”, Portoroz, Slovenia, September 10-
13, 2007
[60] Personal communication of V. Malofeev, June 2005.
[61] P. F. Rose, Compiler and Editor, “ENDF-201, ENDF/B-VI Summary
Documentation,” BNL-NCS-17541, Brookhaven National Laboratory
(October 1991).
[62] R. J. Howerton, D. E. Cullen, R. C. Haight, M. H. MacGregor, S. T.
Perkins, and E. F. Plechaty, “The LLL Evaluated Nuclear Data Library
(ENDL): Evaluation Techniques, Reaction Index, and Descriptions of
Individual Reactions,” Lawrence Livermore National Laboratory report
UCRL-50400, Vol. 15, Part A (September 1975).
[63] D. E. Cullen, M. H. Chen, J. H. Hubbell, S. T. Perkins, E. F. Plechaty, J. A.
Rathkopf, and J. H. Scofield, "Tables and Graphs of Photon Interaction
Cross Sections from 10 eV to 100 GeV Derived from the LLNL Evaluated
Photon Data Library (EPDL)," Lawrence Livermore National Laboratory
report UCRL-50400, Volume 6, Rev. 4, Part A: Z = 1 to 50 and Part B: Z =
51 to 100 (1989).
236
[64] R. E. MacFarlane and D. W. Muir, "The NJOY Nuclear Data Processing
System Version 91," Los Alamos National Laboratory report LA-12740-M,
(October 1994).
[65] R. E. MacFarlane, D. W. Muir, and R. M. Boicourt, “The NJOY Nuclear
Data Processing System, Volume I: User’s Manual,” Los Alamos National
Laboratory report LA-9303-M, Vol. I (ENDF-324) (May 1982).
[66] J. R. Askew, F. J. Fayers, and P. B. Kemshell, “A General Description of
the Lattice Code WIMS,” J. Brit. Nucl. Energy Soc., 5, 564 (1966).
[67] C. J. Taubman, The WIMS 69–Group Library Tape 166259 , Technical
Report AEEW-M1324, United Kingdom Atomic Energy Establishment
(1975).
[68] J. J. Kim, J. T. Lee, and H. R. Kim, “Generation and Benchmarking of a 69
Group Cross Section Library for Thermal Reactor Applications,” J. Korean
Nucl. Soc., 21, 245 (1989).
[69] WLUP, Final Stage of the WIMS-D Library Update Project, www-
nds.iaea.org/wimsd (2005).
[70] J. V. Donnelly, WIMS-CRNL, A User’s Manual for the Chalk River Version
of WIMS, Technical Report AECL-8955, Atomic Energy of Canada Limited
(1986).
[71] W. M. Stacey, “Nuclear Reactor Physics”, Wiley-VCH Verlag GmbH & Co.
KGaA, 2004. ISBN-13:978-0-471-39127-2
[72] Turinsky, P., et. al., 1994. NESTLE, Few-group neutron diffusion equation
solver utilizing the nodal expansion method for eigenvalue, adjoint, fixed-
source steady-state and transient problems. Electric Power Research
Center, North Carolina State University, Raleigh, NC, USA.
[73] V. Malofeev, Personal email to C. Parisi. 31 January 2006.
[74] V. Malofeev, Personal email to C. Parisi. 31 May 2007.
[75] H. Finnemann and A. Galati, "NEACRP 3-D LWR Core Transient
Benchmark – Final Specifications," NEACRP-L-335 (Revision 1), January,
1992
[76]J. L. Judd, W. L. Weaver, T. Downar, and J. G. Joo, "A Three Dimensional
Nodal Neutron Kinetics Capability for RELAP5," Proceedings of the 1994
237
Topical Meeting on Advances in Reactor Physics, Knoxville, TN, April 11-
15, 1994, Vol. II, pp 269-280
[77] E. Bubelis, A. Kaliatka, and E. Uspuras, “RELAP-3D Code Application for
RBMK-1500 Reactor Core Analysis”, Proceedings of ICONE-10
Conference, Arlington, VA, April 14-18 (2002).
[78] E. Uspuras, A. Kaliatka, E. Bubelis, ‘Modeling of void, fast power and
graphite temperature reactivity coefficients measurements for the
©
validation of RELAP5-3D RBMK-1500 reactor model’, Nuclear
Engineering and Design, Vol. 224, pgs. 293-300
[79] E. Uspuras, E. Bubelis, ‘RELAP5-3D code validation in the neutron-
dynamic analysis of transient proceses taking place in RBMK-1500
reactors’, Nuclear Engineering and Design, Vol. 232, pgs. 29-45
[80] R. C. Little, “Neutron and Photon Multigroup Data Tables for MCNP3B,”
Los Alamos National Laboratory internal memorandum, X-6:RCL-87-225
(1987) ( available URL: http://laws.lanl.gov/projects/data/index.html ).
[81] E. Knoglinger et al., “Evaluation of the Smolensk-3 Shutdown System -
Final Report – Phase II and III “, PSI Bericht Nr. 96-10, April 1996 – ISSN
1019-0643
[82] Lombardi Costa A., Cherubini M., D’Auria F., Giannotti W., Moskalev A.
“Thermal-Hydraulic Analysis of Coolant Flow Decrease in Fuel Channels of
Smolensk-3 RBMK during GDH Blockage Event”, J. Science and
Technology of Nuclear Installations, Vol. (2007), Art. ID 87834, pp. 1-7
[83] INSAG-7, “The Chernobyl Accident: Updating of INSAG-1”, Safety Series
Report No. 75, IAEA, Vienna, 1992
[84] S. Lavrencic, “Chernobyl. Anatomia di un’esplosione”, Report AT-WAA-
00003, ENEA, Dipartimento Reattori Innovativi, 1991. Translation o fan
article of G. L’vov, correspondent of “Nauka i Zhizn”, No. 12, 1989.
[85] “Dossier Chernobyl”, an ANSALDO-NIRA report
[86]GIDROPROEKT, “1st stage Smolensk nuclear power plant; 2nd stage
Kursk nuclear power plant; 2nd stage Chernobyl nuclear power plant”,
Technical Safety Report, Rep. 176, Moscow, 1976.
238
[87] G. Fruttuoso, G.M. Galassi, M. Mazzini, “Analisi dell’incidente di
Chernobyl con l’impiego del codice di calcolo RELAP 4 MOD. 6”, ENEA-
DISP Seminar on Chernobyl Accident, 8 January 1987, Rome.
[88] E. O. Adamov, “Soviet Chernobyl model highlights role of pump trip”,
Nuclear Engineering International, 33 (406), 1998.
[89]P. S. W. Chan, A. R. Dastur, “The Sensitivity of Positive Scram Reactivity
to Neutronic Decoupling in the RBMK-1000”, Nuclear Science and
Engineering, Vol. 103, 1989. Pages 289-293
[90] V. A. Khotylev, “On the interpretation of a positive scram reactivity”,
Annals of Nuclear Energy, Vol. 23, No. 9, 1996. Pages 779-784.
[91] P. A. Landeyro, A. Buccafruni, “Time-Independent Neutronic Analysis of
the Chernobyl Accident”, Nuclear Science and Engineering, Vol. 108,
1991. Pages 126-149.
[92] H. Mochizuki, “Analysis of the Chernobyl accident from 1:19:00 to the first
power excursion”, Nuclear Engineering and Design, Vol. 237, 2007. Pages
300-307.
[93] A. D’Angelo, “Sensitivity study of the positive scram effect relevant to the
Chernobyl accident simulation”, Nuclear Science and Engineering, Vol.
125, 1997. Pages 93-100.
[94] S. Lavrencic, translation of: A. V. Krayushkin et alii, “Neutron-Physical
Investigation of the DBA of RBMK-1000”, on Atomnaya Energiya, Vol. 71,
No. 6, December 1991. ENEA Report, AT-WAA-00009, 1998.
[95] J. Chao et alii, “An analysis of the Chernobyl accident using RETRAN-
02/mod3”, Nuclear Technology, Vol. 83, 1988. Pages 289-301.
[96] C. Parisi, F. D’Auria, “Calculation of Xenon transients and implementation
of an automatic power control system on a RELAP5-3D VVER-440 model”,
Technical Note of DIMNP, 2008.
[97] E. Uspuras, A. Kaliatka, V. Vileiniskis, “Implementation of weak heat
conduction mechanism model for analysis of long-term transients using
RELAP5-3D code”, The 7th International Information Exchange Forum on
“Safety Analysis for Nuclear Power Plants of VVER and RBMK types”
(FORUM-7), 28-30 October 2003, Piestany, Slovakia.
239
[98] P. Bayless, J. Fisher, “INSP IRUG-related Activities in FY-2001”, RELAP5
Users Seminar, Elkhorn Resort, Sun Valley, Idaho, 5-7 September 2001
[99] M. Clemente, S. Langenbuch, K. Velkov, P. Kusnetzov, I. A. Steinbock,
“Application of the coupled code system ATHLET and QUABOX/CUBBOX
to an ATWS Transient ‘Total Loss-of-Feedwater’ in RBMK”, International
Conference on the Physics of Nuclear Science and Technology, Islandia
Marriott, Long Island, USA, October 5-8, 1998.
[100] A. Petry, H. Wolff, S. Langenbuch, M. Clemente “Analysen von
RBMK-Reaktoren,Gesellschaft für Anlagen- und Reaktorsicherheit (GRS)
mbH“ Report GRS – A – 2483, June 1997 (in German)
240

Nuclear Safety OF RBMK Reactors: Università Di Pisa Scuola Di Dottorato in Ingegneria "Leonardo Da Vinci"

Uploaded by

Copyright:

Available Formats

Nuclear Safety OF RBMK Reactors: Università Di Pisa Scuola Di Dottorato in Ingegneria "Leonardo Da Vinci"

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nuclear Safety OF RBMK Reactors: Università Di Pisa Scuola Di Dottorato in Ingegneria "Leonardo Da Vinci"

Uploaded by

Copyright:

Available Formats

UNIVERSITÀ DI PISA

Scuola di Dottorato in Ingegneria “Leonardo da Vinci”

Corso di Dottorato di Ricerca in

Tesi di Dottorato di Ricerca

Scuola di Dottorato in Ingegneria “Leonardo da Vinci”

Corso di Dottorato di Ricerca in

Tesi di Dottorato di Ricerca

Prof. Francesco D’Auria (UNIPI) Firma_____________

Dr. Gianni Petrangeli Firma_____________

Prof. Eugenius Uspuras (LEI) Firma_____________

Dr. Sergei L. Soloviev (NIKIET) Firma_____________

Le attività di ricerca hanno contemplato lo sviluppo di una complessa

Dopo la validazione dei modelli sviluppatii, si sono eseguite analisi di transitori a

L’ultima parte delle attività di dottorato si sono concentrate sull’analisi di un

In conclusione, le analisi avanzate effettuate nell’ambito di questa Tesi di

2. THE DESCRIPTION OF THE RBMK ...................................................... 5

3. THE BACKGROUND FOR RBMK ACCIDENT ANALYSIS ............... ..47

4. THE TOOLS, THE METHODOLOGIES AND THE QUALIFICATION OF

5. THE APPLICATION TO REALISTIC TRANSIENTS

6. THE EXTREME CASE ........................................................................ 209

7. CONCLUSIONS .................................................................................. 228

1.1. State of the Art on the RBMK Safety Analyses

In particular, immediate actions focused on plant modifications, and concerned

Objectives of these Projects were NPP hardware upgrading, personnel training,

According to the requirements of the Russian nuclear regulatory authority

• The estimated probability of an event with large release should be

Thus, according to requirements OPB-88/97 of the Russian nuclear regulatory

Fig. 1 – Overall view of the RBMK NPP system

1-separator drum, 2-downcomers, 3-suction header, 4-suction piping of the MCP,

Fig. 3 – RBMK Main Circulation Circuit (MCC) flow diagram

Fig. 4 – Key elevations in the MCC of the RBMK system (Smolensk 3)

Fig. 5 – The MCP for RBMK NPP.

The coolant is distributed to individual GDH by means of twenty-two pipes 325x15

1 – isolation and control valve, 2 – ball type flow-rate

Fig. 7 – Geometrical configuration of RBMK GDH

The separator drum has the following functions:

Fig. 10 – Layout of the two DS and related connection

2.1.2. The core

Fig. 11 – Geometrical configuration of RBMK core: a) detail of active region; b)

The interaction of fast neutrons can lead to dimensional changes in various

Fig. 12 – Geometrical configuration of a) RBMK fuel bundle, b) fuel rod

1 – steel biological shield

Fig. 13 – Geometrical and material configuration of RBMK FC

1 – support tube, 2 – absorbing

1 – top storage tank

Fig. 17 – Flow diagram of the CPS cooling system

2.1.2.3. The scram signals

Tab. 2 – CPS Modes of Operation

The following should be noted concerning the CPS CR grouping:

Reactor L (cm) tth1/2 (cm) M (cm) Diameter (L) Diameter (M)

Tab. 5 – Neutronic characteristics of a RBMK fuel cell

Neutronic characteristics of a RBMK fuel cell

Percentages of fissions thermal range (E<0.625 eV): 91.67%

Fig. 19 – Thermal neutron flux distribution

It is clear in Fig. 19 the neutron thermalization and accumulation in the graphite

Fig. 20 – Fast neutron flux

BWR-6 2894 56 19.8 44 ~54

2.1.2.5. Reactivity Control

2.1.2.5.1. Fuel Cell Overmoderation