TE 1382 Part1

IAEA-TECDOC-1382
Evaluation of high temperature

gas cooled reactor performance:
Benchmark analysis related to initial
testing of the HTTR and HTR-10
November 2003
The originating Section of this publication in the IAEA was:

Nuclear Power Technology Development Section
International Atomic Energy Agency
Wagramer Strasse 5
P.O. Box 100
A-1400 Vienna, Austria
EVALUATION OF HIGH TEMPERATURE GAS COOLED REACTOR PERFORMANCE:

BENCHMARK ANALYSIS RELATED TO INITIAL TESTING OF THE HTTR AND HTR-10
IAEA, VIENNA, 2003
IAEA-TECDOC-1382
ISBN 9201162030
ISSN 10114289
IAEA, 2003
Printed by the IAEA in Austria
November 2003
FOREWORD
The Co-ordinated Research Project (CRP) on Evaluation of High Temperature Gas
Cooled Reactor (HTGR) Performance was initiated by the IAEA in 1998 on the
recommendation of the Technical Working Group on Gas Cooled Reactors. This CRP was
established to foster the sharing of research and associated technical information between
participating Member States in the ongoing development of the HTGR as a future source of
nuclear energy for high temperature process heat applications and the production of
electricity.
The IAEA has facilitated an extensive programme that addresses the technical
development of advanced gas cooled reactor technology. This CRP complements other
recently completed CRPs in validating safety and performance capabilities of the HTGR. For
this CRP, computer codes and models are verified through actual test results from operating
reactor facilities. Specifically, the IAEA is grateful to the Japan Atomic Energy Research
Institute for providing the HTTR, Chinas Institute of Nuclear Energy Technology for the
HTR-10 and the Russian Federations OKBM and the Kurchatov Institute for the ASTRA
critical facility in support of this CRP.
The focus of this TECDOC is on core physics benchmarks in conjunction with initial
testing of the HTTR and HTR-10 and also selected thermal hydraulic benchmarks on the
HTTR. The following Member State national institutions participated in the performance of
this CRP:
Institute of Energy Technology (INET), Beijing, China

SACLAY (CEA), Gif-sur-Yvette, France
Research Centre Juelich (ISR), Juelich, Germany
National Nuclear Energy Agency (BATAN), Serpong, Indonesia
Japan Atomic Energy Research Institute (JAERI), Oarai, Japan
Nuclear Research and Consultancy (NRG), Petten, Netherlands
OKBM/Kurchatov Institute, Nizhny Novgorod, Russian Federation
Pebble Bed Modular Reactor (PBMR), Centurion, South Africa
Department of Nuclear Engineering, Hacettepe University, Ankara, Turkey
Oak Ridge National Laboratory (ORNL), Oak Ridge, TN, United States of America
The development of this report was co-ordinated by K. Kunitomi (JAERI), Y. Sun (INET)
and S. Ball (ORNL) with final compilation by H.L. Brey (Consultant to JAERI). The IAEA
officer responsible for this publication was M. Methnani of the Division of Nuclear Power.
EDITORIAL NOTE
The use of particular designations of countries or territories does not imply any judgement by the
publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and
institutions or of the delimitation of their boundaries.
The mention of names of specific companies or products (whether or not indicated as registered) does
not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement
or recommendation on the part of the IAEA.
CONTENTS
CHAPTER 1. INTRODUCTION .............................................................................................. 1
CHAPTER 2. HIGH TEMPERATURE TEST REACTOR (HTTR)
REACTOR PHYSICS BENCHMARKS ........................................................... 3
2.1. HTTR general information................................................................................................. 3
2.1.1. Facility description.................................................................................................. 3
2.1.2. Reactor physics benchmark problem descriptions ................................................ 25
2.2 Reactor physics benchmark analysis and results............................................................... 27
2.2.1 Japan....................................................................................................................... 27
2.2.2. France .................................................................................................................... 50
2.2.3. Germany ................................................................................................................ 69
2.2.4. Indonesia ............................................................................................................... 93
2.2.5. Netherlands............................................................................................................ 98
2.2.6. Russian Federation .............................................................................................. 117
2.2.7. Turkey ................................................................................................................. 131
2.2.8. United States of America .................................................................................... 136
2.3. References to Chapter 2 .................................................................................................. 141
CHAPTER 3. HIGH TEMPERATURE TEST REACTOR
THERMAL HYDRAULIC BENCHMARKS ............................................... 145
3.1. Thermal hydraulic benchmark information.................................................................... 145
3.1.1. Description of systems for VC and LP benchmarks ........................................... 145
3.1.2. HTTR thermal hydraulic benchmark problem descriptions................................ 156
3.2. Thermal hydraulic benchmark analysis/results .............................................................. 157
3.2.1. Japan.................................................................................................................... 157
3.2.3. South Africa ........................................................................................................ 183
3.2.5. France .................................................................................................................. 218
CHAPTER 4. THE HIGH TEMPERATURE GAS COOLED REACTOR TEST
MODULE CORE PHYSICS BENCHMARKS............................................. 227
4.1. HTR-10 general information.......................................................................................... 227
4.1.1. Facility description.............................................................................................. 227
4.1.2. Benchmark problem descriptions........................................................................ 239
4.2. HTR-10 benchmark problem analysis/results ................................................................ 244
4.2.1. China ................................................................................................................... 244
4.2.2. Indonesia ............................................................................................................. 254
4.2.3. Japan.................................................................................................................... 261
4.2.5. Netherlands.......................................................................................................... 271
4.2.7. Turkey ................................................................................................................. 290
4.2.8. France .................................................................................................................. 295
4.2.9. Germany .............................................................................................................. 304
4.2.10 South Africa ....................................................................................................... 314

CHAPTER 5. RESULTS, CONCLUSIONS AND RECOMMENDATIONS ...................... 327
5.1. HTTR reactor physics benchmarks ................................................................................ 327
5.1.1. Collation of codes and models ............................................................................ 327
5.1.2. Collation of benchmark analysis results.............................................................. 328
5.1.3. Discussion of results and general conclusions .................................................... 332
5.1.4. Recommendations ............................................................................................... 333
5.2. HTTR thermal hydraulic benchmarks............................................................................ 333
5.2.1. HTTR vessel cooling........................................................................................... 333
5.2.2. HTTR loss of off-site electric power................................................................... 335
5.3. HTR-10 reactor physics benchmarks ............................................................................. 336
5.3.1. Analysis methods and models ............................................................................. 336
5.3.2. Collation of results .............................................................................................. 338
5.3.3. Analysis and conclusions .................................................................................... 341
5.3.4. Recommendations ............................................................................................... 342
APPENDIX: RELATED IAEA PUBLICATIONS ............................................................... 343
CONTRIBUTORS TO DRAFTING AND REVIEW ........................................................... 345
Chapter 1
INTRODUCTION
Of paramount significance in the development of new high temperature gas cooled
reactor (HTGR) concepts is the predicted capability for this advanced nuclear plant to achieve
a high degree of safety through reliance on passive safety features. Because of this, the
investigation and validation of the safety and operational aspects of the HTGR were the
primary focus for many of the coordinated research programmes (CRPs) initiated by the
IAEA in the 1990s. These included:
the neutronic physics behaviour of the HTGR core,

fuel performance and fission product behaviour, and
the ability of the HTGR to dissipate decay heat by natural transport mechanisms under
accident conditions
The principal tools utilized in these CRPs included scientific research and engineering
development through analytical evaluation of benchmark problems, application of new and/or
existing computer codes and models and utilization of test apparatus and loops for specific
component validation.
The next important step in bringing this advanced nuclear power programme from
concept to actuality is to verify system performance and safety under actual HTGR operating
conditions. It is the need for validation via testing in nuclear reactors that was the stimulus for
the IAEA to initiate this CRP on Evaluation of High Temperature Gas Cooled Reactor
Performance.
The principal facilities utilized in the performance of this CRP included Japans High
Temperature Engineering Test Reactor (HTTR), Chinas High Temperature Test Module
(HTR-10), and Russias ASTRA critical assembly. Selected codes and models associated with
the Gas Turbine Modular Helium Reactor (GT-MHR) and Pebble Bed Modular Reactor
(PBMR) plant designs are also investigated within the scope of this CRP.
The objectives of this CRP include:
1. Validation of analytical codes and performance models to actual operating conditions
of HTGRs,
2. Formulation of research and development code-to-experiment benchmark activities for
inclusion into the test programmes for the HTTR, the HTR-10 and the ASTRA
facilities,
3. Investigation of analytical codes and models associated with future HTGR gas turbine
plants utilizing code-to-code benchmark problems, and
4. Demonstration of HTGR safety characteristics.
Overall, this TECDOC addresses performance based HTGR code and model
verification during startup, steady state and transient operational conditions of the HTR-10
and HTTR test facilities are being evaluated. Also included within the scope of this CRP is
the utilization of national research facilities and computer models to investigate the areas of
core physics, safety characteristics of the HTGR, fission product release and transportation
behaviour, thermal hydraulics, control response and high temperature component
performance.
This TECDOC documents the benchmark problem results obtained by Chief Scientific
Investigators (CSIs) from China, France, Germany, Indonesia, Japan, the Netherlands, Russia,
South Africa, Turkey and the United States for the following areas:
Reactor physics benchmark analysis of the HTTR including initial criticality, control
rod worth, excess reactivity, scram reactivity and temperature coefficient of reactivity,
Reactor physics benchmark analysis of the HTR-10 including initial criticality,

temperature coefficient of reactivity and control rod (including differential) worth,
Selected thermal hydraulic benchmark analysis for the HTTR including vessel cooling
and loss of off-site electric power.
The presentation of this TECDOC is by facility and benchmark type with Chapter 2
devoted to HTTR core physics, Chapter 3 to the HTTR thermal hydraulic benchmark
problems associated with vessel cooling and loss of electric power. Chapter 4 is devoted to
HTR-10 core physics benchmarks. Each of these chapters includes a description of the test
reactor, an overview of the benchmark problems being addressed, individual analysis by each
Member State and a review of the actual test results of each problem as performed on the test
reactor. Chapter 5 provides a collation of the results, general conclusions and
recommendations for code and model improvements determined as the result of participation
in the CRP.
Chapter 2
HIGH TEMPERATURE ENGINEERING TEST REACTOR (HTTR)
REACTOR PHYSICS BENCHMARKS
2.1. HTTR GENERAL INFORMATION
The High Temperature Engineering Test Reactor (HTTR) of the Japan Atomic Energy
Research Institute (JAERI) is a graphite-moderated and helium gas cooled reactor with an
outlet temperature of 950C and a thermal output of 30 MW [2-1].
The major objectives of the HTTR are to establish and upgrade the technological basis
for advanced high temperature gas-cooled reactors (HTGRs) and to conduct various
irradiation tests for innovative high temperature basic researches.
The construction was completed on May 1996. The fuel loading of was started on July
1, 1998 from core periphery. The first criticality was attained in annular type core of 19
columns on Nov.10, 1998. The first full power operation with an average core outlet
temperature of 850C was completed on 7 December 2001, and operational licensing of the
HTTR was approved on 6 March 2002.
2.1.1.
Facility Description
2.1.1.1. Background
In June 1987, the Japanese Atomic Energy Commission (JAEC) issued a revised
Long Term Program for Development and Utilization of Nuclear Energy stressing that
Japan should proceed to develop more advanced reactor technologies in parallel with the
upgrading of existing nuclear reactors. It was recognized in this programme that the HTGR
was not to be incorporated into the present existing power plant system, but the benefits that
could be derived, such as its inherent safety and production of high temperature heat, are
remarkable and should be pursued. Therefore, the promotion of R&D on the HTGR in Japan
is quite significant from the viewpoint of a new nuclear technology frontier.
Within this programme, the early construction of a test reactor in place of the
experimental Very High Temperature Reactor (VHTR) was recommended based on the
estimated length of time (~ 10 years) to construct the plant and perform associated testing.
Based on the conclusions by a special committee investigating the HTGR R&D plan for the
JAEC, demand of nuclear heat applications to ~ 1,000C is expected to become strong in the
early part of the 21st century. The committee requested that a test reactor making the most use
of Japanese technology be designed and built to test and study advanced HTGRs for the
future. Accordingly, the test reactor should have an in-core irradiation region equipped with
the capability to test the threshold of fuel failure and for the irradiation of various materials.
The committees report also stated that the reactor outlet coolant temperature should
be 950C, which is the highest temperature attainable considering the current technology
level. For irradiation tests, a prismatic block type core structure was proposed, with a thermal
power rating of 30MW for securing adequate regions for irradiation tests at high temperature
[2-6, 2-7].
Based on these suggestions, JAERI suspended the previous programme where the
experimental VHTR was defined as an initial step toward nuclear heat application
development and proceed with the design and R&D specifically necessary to prepare the
safety analysis report of the HTTR, with the objectives to:
Establish and upgrade the technology base of the HTGR

Perform innovative basic research in the field of high temperature engineering
Demonstrate high temperature heat applications and utilization achieved from nuclear
heat.
The Japanese government approved proceeding with the HTTR in its 1989 fiscal year
budget. Construction began in March 1991 following submittal by JAERI of the HTTR safety
analysis report and subsequent review by the Science and Technology Agency, and then, by
the Nuclear Safety Commission.
2.1.1.2. General design features of the HTTR
The reactor core is designed to keep all specific safety features within the graphite
blocks. The intermediate heat exchanger (IHX) is equipped to supply high temperature clean
helium for process heat application systems, and the instrumentation and control system is
designed to allow operations that simulate accidents and anticipated operational occurrences
(AOOs). As the HTTR is the first HTGR in Japan and a test reactor with various purposes, it
incorporates specific aspects regarding the safety design. JAERI established the following
safety design principles for the HTTR in reference to the Guidelines for Safety Design of
LWR Power Plants, but taking into account the significant safety characteristics of the
HTGR and the corresponding design requirements as a test reactor:
Coated fuel particles shall not fail during normal operation and AOOs. To satisfy this
principle, the maximum fuel temperature, including systematic and random uncertainties,
shall not exceed 1600C for any AOO.
The reactor shall be shut down safely and reliably during operation using the control rod
system. Furthermore, a reserved shutdown system (RSS) which is independent of the
control rod system shall be provided.
A severe accident resulting from control rod ejection must be avoided.
The residual heat after reactor shutdown shall be removed safely and reliably for any
AOO or accident.
A containment vessel (C/V) shall be provided to prevent fission product release and
excessive air ingress into the core in case of a depressurization accident.
The pressure in the pressurized water cooling system (PWCS) shall be controlled so as to
be lower than that of the primary helium gas to prevent a large water ingress into the core
in case of rupture of a heat transfer tube in the primary pressurized water cooler (PPWC).
The helium gas pressure in the secondary helium cooling system (SHCS) shall be
controlled to be slightly higher than that of the primary helium gas to prevent fission
product leakage from the primary cooling system (PCS) to the secondary due to a crack in
a heat transfer tube in the IHX.
The pressure and heat resisting functions of the structures, where the high pressure and
high temperature coolant is contained, are separated to reduce mechanical loads on the
high temperature metal structures [2-7].
Safe and reliable shutdown of the reactor from any operational condition is achieved
with the control rod system. Furthermore, a reserved shutdown system composed of B4C/C
pellets is provided. The power control and normal reactor shutdown of the HTTR are
achieved with 16 pairs of control rods or 15 pairs when the center column of the core is used
for an irradiation test. The control rod system can achieve subcriticality from any operational
condition and maintain subcriticality under cold core conditions including the postulated
event of a pair of control rods stuck in the operational position. The major design
specifications of the HTTR are shown in Table 2.1. The reactor outlet coolant temperature at
the full power is set at both 850 and 950C. The reactor operational mode at 850C is defined
as "rated operation" and at 950C is "high temperature test operation" because operation of
the HTTR is not allowed at 950C for full life of the initial core. Tests such as the safety
demonstration tests and irradiation tests are allowed only in the rated operation mode. The
high temperature nuclear process heat utilization system will be operated at the high
temperature test operational mode. The design life of permanent structural components in the
HTTR plant is based on 20 years with a load factor of 60 % of full power operation.
The HTGR has excellent safety capabilities with respect to the accidental release of
fission products. Nevertheless, the HTTR is required to have a containment vessel to meet
Japanese safety design guidelines for the light water nuclear power plants.
2.1.1.3. HTTR plant layout and cooling system
The HTTR is located on JAERIs Oarai Research Establishment site which is
approximately 100 kilometers north of the Tokyo metropolitan area and is near the Pacific
Ocean. The plant area is 200 m x 300 m in size. The shortest distance between the HTTR
reactor core and site boundary is about 280 m in the southwest direction. As illustrated in
Figure 2.1, the HTTR plant arrangement is comprised of the reactor building, spent fuel
storage building, a machinery building, cooling towers, exhaust stack, a high temperature
process heat utilization system and other auxiliary facilities. The reactor building of 48 m x 50
m in size is situated in the central area of the plant. The exhaust stack of 80 m in height is
north of the reactor building for the air ventilated from the reactor building to be released to
the atmosphere. The heat utilization system will be constructed south of the reactor building
[2-8].
FIG. 2.1. HTTR plant arrangement.
The reactor building includes five levels with three floors underground (Figure 2.2).
A steel reactor containment vessel of 18.5 m in diameter and 30 m in height is installed in the
center of the reactor building. A refueling hatch is attached to the C/V above the reactor
pressure vessel (RPV). Functions of the C/V are to:
1)
2)
Contain fission products (FPs) as one of the multiple barriers against FP

release into the atmosphere, and
Limit the amount of air ingress into the core in a primary pipe rupture accident.
The reactor pressure vessel (RPV) is formed as a vertical cylinder, with a

hemispherical top and bottom head closures and 31 standpipes. The top head closure is bolted
to a flange of the vessel cylinder. The standpipes include "control rod (CR) stand-pipes",
irradiation stand-pipes", and standpipes for instrumentation. The irradiation standpipes are
utilized to introduce specimens and experimental equipment into the core. A thermal shield is
attached to the inner surface of the top head closure to prevent the closure from overheating in
a depressurization accident such as a primary helium pipe rupture. The RPV is of 2-1/4Cr1Mo
steel normalized and tempered.
FIG. 2.2. HTTR reactor building.

The double containment concept (RPV + C/V) was applied to the HTTR, because the
safety features of the HTGR have not been developed in Japan. The HTTR has multiple
barriers to fission product release, namely, the fuel coatings, the RPV boundary, the C/V and
the reactor building. Most HTGRs being designed in other countries also include these
barriers except the C/V. Some compartments surrounding the C/V in the reactor building
serve as confinement, or service area. The service area is maintained at a slightly negative
pressure to atmosphere by a ventilation and an air conditioning system during normal
operation and accident conditions. The barriers of the C/V and the service area significantly
reduce the off-site radiation dose in an accident condition such as a primary helium pipe
rupture. Major components such as primary cooling system components as well as the RPV
(Figure 2.3) are contained within the C/V.
FIG.2.3. HTTR pressure vessel and internals.
FIG. 2.4. HTTR cooling system.

The flow diagram of the reactor cooling system is shown in Figure 2.4. This system is
composed of a main cooling system (MCS), an auxiliary cooling system (ACS) and two
reactor vessel cooling systems (VCSs). The MCS removes the heat energy from the reactor
core during the normal operation, while the ACS and VCSs are functioned as engineered
safety features and remove the residual heat energy after a reactor scram. As the core restraint
mechanism requires protection against thermal damage from reactor heat during normal
operation and anticipated operational occurrences, the ACS functions as protection by forced
cooling the restraint mechanism.
During a reactor scram, gas circulators of the MCS are shut to protect the heat transfer
tubes of the two pressurized water coolers against overheating. In an AOO and accident
condition when forced cooling of the core is available, the ACS automatically starts in
response to the reactor scram signal. The VCS functions as a residual heat removal system
when forced circulation in the primary cooling system is no longer available due to a rupture
of its piping system. It also operates during normal operation to cool the reactor shielding
concrete wall.
The MCS consists of an IHX, a primary pressurized water cooler (PPWC), a secondary
pressurized water cooler (SPWC) and pressurized water/air cooler. The MCS has two
operational modes; single loaded operation, and parallel loaded operation. The PPWC
functions to remove the reactor heat of 30 MW during the single loaded operation, while
during parallel load operation the IHX removes 10MW and the PPWC removes 20MW. The
SPWC serves the function of removing the heat from the IHX. The heat removed by the
PPWC and the SPWC is transported through the pressurized water at 3.5MPa. The
pressurized water is then cooled down by the air cooler. In the HTTR reactor plant, the reactor
heat of 30 MW is eventually transferred to the atmosphere by the pressurized water and the
air cooler. During normal operation, the pressure of the secondary helium is controlled to
always be 0.1 MPa higher than that of the primary helium at the IHX heat transfer tubes in
order to reduce the pressure load on the tubes and to protect for accidental leakage of
radioactive materials into the secondary helium. The water pressure is controlled so that a
large amount of water can not ingress into the core with a PPWC tube rupture accident.
The auxiliary helium transfers a small fraction of the reactor heat to pressurized water.
Eventually, the reactor heat is dissipated to the atmosphere at the auxiliary water/air cooler.
The ACS consists of an auxiliary heat exchanger (AHX), two auxiliary helium circulators and
an air cooler. At the AHX, the auxiliary helium is cooled by water. During normal operation,
a small flow of auxiliary helium (~ 200 kg/h) passes through the AHX to the primary helium
purification system so as to remove impurities contained in the reactor coolant. With a reactor
scram, while the reactor coolant pressure boundary remains intact, the auxiliary helium
cooling system automatically starts and transfers the residual heat from the core to the
auxiliary air cooler. The AHX has heat transfer capacity of approximately 3.5MW.
Two vessel cooling systems are provided as protection of the reactor core and the RPV
against thermal damage by residual heat after a reactor scram when the ACS cannot, or fails
to, cool the core. Each of these systems is capable of controlling temperatures of the core and
RPV within safe limits and consists of water-cooled panels surrounding the RPV with two
cooling water systems. Cooling tubes with fins form the panels and are arranged so that
adjacent tubes do not belong to the same system and a tube failure will not danger the RPV
and core. The heat removal rate from the RPV to the panels is designed as 0.6 MW so as to
effectively remove heat to meet the requirement for the maximum allowable normal fuel
temperature of 1495C and also 0.3 MW or more with an accident condition where the reactor
core is not cooled by the ACS. The VCS is also an engineered safety feature equipped with
two independent complete sets which are backed up with an emergency power supply. It is
operated even during normal operation in order to cool the biological shielding concrete wall.
The IHX is a helically coiled counter flow type heat exchanger. To minimize
constraints of axial and radial thermal expansion of the helically coiled heat transfer tubes, a
floating hot header with a combination of a central hot gas duct passes through the central
space inside the helix bundle. An assembled tube support allows free thermal expansion of a
helix in the radial direction. The primary helium enters the IHX through the inner pipe of the
primary concentric hot gas duct attached to the bottom of the IHX. It flows up outside the
tubes thereby transferring nuclear heat of 10 MW to the secondary helium and flows back to
the annular space between the inner and outer shells, The secondary helium flows down inside
the heat transfer tubes and flows up through the center as hot gas. A double-walled shell with
thermal insulation attached to the inside surface of the inner shell provides reliable separation
of the heat resisting and pressure retaining functions. Cold helium flowing through the
annulus brings uniform temperature distribution throughout the outer shell which serves the
function of being the pressure retaining member.
2.1.1.4. HTTR core configuration
The annular core is one of the promising core types for future HTGRs because of high
inherent safety characteristics for loss of coolant accidents [2-2, 2-3, 2-4]. The decay heat
removal is enhanced by introduction of the annular core because the heat transfer pass will be
shortened due to thinning of the fuel region. As a result, the fuel temperature on a loss of
coolant accident can be maintained less than the fuel temperature limit of 1600C by the
vessel cooling system (VCS) surrounding the reactor pressure vessel. The decay heat will be
transferred radially through the fuel regions, side reflector blocks and reactor pressure vessel
to the cooling panel of the VCS by heat conduction, radiation and convection without any
active cooling system.
Table 2-1. Specification of the HTTR

Thermal power
Outlet coolant temperature
Inlet coolant temperature
Primary coolant pressure
Core structure
Equivalent core diameter
Effective core height
Average power density
Fuel
Uranium enrichment
Type of fuel
Burn-up period (efpd)
Coolant material
Flow direction in core
Reflector thickness
Top
Side
Bottom
Number of fuel assemblies
Number of fuel columns
Number of pairs of control rods
In core
In reflector
30 MW
950C
395C
4 MPa
Graphite
2.3 m
2.9 m
2.5 W/cm3
UO2
3 to 10 wt%
Pin-in-block
660 days
Helium gas
Downward
1.16 m
0.99 m
1.16 m
150
30
7
9
FIG.2.5. Vertical view of the HTTR.
10
Three different types of cores were formed during fuel loading for start-up core
physics experiments; thin and thick annular cores were achieved at 18 and 24 fuel column
loaded cores, respectively, and the fully-loaded core consists of 30 fuel columns. These three
core types were proposed for benchmark problems within this CRP. The effective diameter
and height of the HTTR are 230 and 290 cm, respectively. The core size of the HTTR
corresponds to about one half of that of the future HTGRs [2-2, 2-3, 2-4]. Moreover, the high
excess reactivity of the HTTR is similar to that of the future HTGRs because the large excess
reactivity is necessary to compensate for power operation effects of temperature, xenon,
burnup, etc.. Therefore, the benchmark problems of the HTTRs start-up core physics
experiments are useful for verification of design codes of the future HTGRs.
The major specifications of the HTTR are given in Table 2-1 [2-1]. The reactor
consists of core components and reactor internals. They are arranged in the reactor pressure
vessel (RPV), which is 13.2m in height and 5.5m in diameter. Figure 2.5 shows the vertical
cross section of the core and reactor internals structure. The core consists of core components
which are prismatic hexagonal blocks 580mm in height and 360mm in width across the flats.
These include fuel assembly blocks, control rod guide blocks, replaceable reflector blocks,
and irradiation blocks. The core components are piled up cylindrically to form the core. The
reactor internals consist of graphite and metallic core support structures and shielding blocks.
They support and arrange the core components within the RPV.
The active core, 290cm in height and 230cm in effective diameter, consists of 30
columns and 7 control rod guide columns. Horizontal cross sections of the HTTR are shown
in Figures 2.6 and 2.7. An additional 9 control rod columns are arranged among the adjacent
reflector graphite columns. The replaceable reflector region adjacent to the active core
consists of 9 control rod columns, 12 replaceable reflector columns, and 3 irradiation
columns, which are surrounded by permanent reflector blocks. Each fuel column consists of 2
top reflector blocks, 5 fuel assemblies, and 2 bottom reflector blocks.
FIG. 2.6. Fuel column loading order and horizontal view of HTTR core.
11
FIG. 2.7. HTTR fuel column name and zone number.
Figure 2.8 shows the structure of the pin-in-block type fuel. A fuel rod consists of a
graphite sleeve containing 14 fuel compacts. The fuel rods are inserted into the coolant
channels of the fuel graphite blocks. Each fuel compact contains about 13,000 coated fuel
particles (CFPs) embedded in the graphite matrix. The number of uranium enrichments is 12.
The highest and lowest enrichments are 9.9 and 3.4 wt%, respectively. Fuels of higherenriched uranium are placed in the upper and outer core regions to reduce the maximum fuel
temperature. Burnable poisons (BPs) made of boron carbide and carbon are inserted into two
of three holes below the dowel pins in the fuel graphite block. The coolant gas flow is
downward through annular channels formed by the graphite block and the fuel rod.
12
FIG. 2.8: Structure of fuel assembly.

The top of the control rod guide columns and irradiation columns are 100mm lower
than that of the fuel columns because the 9th layer block of these columns are 480mm in
height. Each horizontal gap width between two columns is 2mm in average.
Reactivity is controlled by 16 pairs of control rods. A pair of control rods is
individually moved by a control rod drive mechanism located in stand-pipes connected to the
hemispherical top head closure of the RPV. The control rods are inserted into two channels of
the control rod guide columns in the active core and in the replaceable reflector regions.
13
Fuel loading scheme

Before fuel loading, the whole fuel region in the core is filled with graphite dummy
blocks. There are two types of dummy blocks: graphite blocks with three large holes and
those with three small holes. The graphite blocks with three large holes were actually used as
control rod guide columns for out-pile seismic tests of the core structure. The others with
three small holes were newly produced as dummy fuel blocks. The primary coolant system is
filled with helium at normal atmospheric pressure and is not in operation during fuel loading.
The core is at room temperature. Fuel loading is carried out by replacing the dummy blocks
with the fuel assemblies, column by column. The fuel loading scheme is shown in Figure 2.6.
The fuel blocks are loaded from the periphery to the center, and thin and thick annular cores
are made at 18 and 24 fuel-column-loaded core, respectively.
The approach to criticality was observed by monitoring the inverse of the neutron
multiplication fraction. The core is regarded as critical when the neutron density is maintained
constant after removing the temporary neutron source. After the first criticality, the increment
in reactivity was measured by the inverse kinetic method. The excess reactivity of the core
was obtained by adding all increments of the reactivity from the first criticality to the fullyloaded core.
Configuration and composition of components
For the calculations, the detailed data of the active core, the adjacent replaceable
reflector region and the permanent reflector region were necessary. Configuration of the
components in the above mentioned regions is described as follows: The position of blocks in
the core is noted by vertical position number and column number. The vertical number is 1, 2,
. . . , 9 as from the top blocks in the 1st layer to the bottom blocks in the 9th layer, with the
column number named according to Figure 2.9. For example, 4C05 means that the block is
placed at the 4th block from the top, the 2nd ring from the core center, and the 5th block from
the north in clockwise direction. In all, 30 fuel columns are grouped concentrically into 4 fuel
zones as shown in Figure 2.9.
Fuel assembly
A fuel assembly consists of fuel rods, two burnable poison (BP) rods and a fuel
graphite block. Each fuel rod consists of a graphite sleeve and 14 fuel compacts containing
coated fuel particles (CFPs). The fuel rods are inserted into vertical holes of 41mm diameter
in the fuel graphite block and form annular coolant channels between the vertical holes and
the fuel rods. Figure 2.8 shows the structure of fuel assembly. There are two types of fuel
graphite blocks having 31 or 33 fuel rods. The number of different uranium enrichments are
12 in the core. The uranium enrichment of all compacts in a fuel assembly is not changed. The
fuel assembly is classified by the uranium enrichment, the number of fuel rods and the type of
BPs. The fuel assembly arrangement in the core is shown in Figure 2.9.
A CFP consists of a spherical fuel kernel of low enriched UO2 with TRISO coating.
The TRISO coating consists of a low-density, porous pyrolytic carbon (PyC) buffer layer
adjacent to the fuel kernel, followed by high density isotropic PyC layer, a SiC layer and a
final outer PyC layer. The CFPs are embedded in graphite matrix of the fuel compact. The
CFPs are classified into 12 sorts by their uranium enrichments. The form of the CFPs is given
in Figure 2.10.
14
FIG. 2.9. Fuel assembly arrangement in the HTTR core.
15
FIG. 2.10. Coated fuel particle, fuel compact and burnable poison pellet.
The fuel compact consists of CFPs and graphite matrix. The form is a hollow cylinder
of 10mm in inner diameter, 26mm in outer diameter and 39mm in height.
The fuel rod consists of a graphite sleeve containing 14 fuel compacts. It is inserted
into the coolant channel of the fuel graphite block. The form of the fuel rod is given in Figure
2.11. The fuel graphite block is a prismatic hexagonal block 580mm in height and 360mm in
width across the flats. The block has 33 fuel holes in the fuel zones 1 and 2, and 31 fuel holes
in the fuel zone 3 and 4. The fuel zone numbers are defined in Fig. 2.7. The form of the block
is given in Figure 2.12 and Figure 2.13.
FIG. 2.11. Fuel rod.
16
17
FIG. 2.12. Fuel block for 33 pin fuel assembly.
18
FIG. 2.13. Fuel graphite block for 31.pin fuel assembly.
19
FIG. 2.14. Control rod guide block in layers 1 through 5.
20
Burnable poison (BP) rods are inserted into two of three BP insertion holes of a fuel
graphite block as shown in Fig. 2.8. The BP insertion holes are under the 3 dowel pins of the
fuel graphite block and are 15mm in inner diameter and 500mm in length. One hole without
the BP rod is empty. Figure 2.6 shows positions of holes which are filled with BPs. The
diameter of BP rod is 14mm. The BP rod consists of BP pellets and graphite disks as shown in
Table 2.2. The graphite disks are put between the BP pellets. The compositions of the BP
pellets are B4C and C. There are two types of BP pellets: H-I with 2.22wt% of natural boron
concentration and H-II with 2.74wt%. The form of the BP pellets and graphite disks is given
in Figure 2.10.
Control rod, reflector and dummy fuel blocks
There are 16 control rod guide columns each consisting of 9 control rod guide blocks.
The control rod guide blocks have three holes. Two of them are control rod insertion holes.
The third hole is a reserve shutdown system (RSS) hole which is used for emergency. Figure
2-6 shows the position of the control rod insertion holes and RSS holes in the core and the
replaceable reflector region. Top and bottom replaceable reflector blocks are placed above
and below the fuel assemblies. The arrangement of coolant channels (with 23mm in inner
diameter) in the top replaceable reflector blocks corresponds to that of coolant channels (with
41mm in inner diameter) in the fuel assemblies within the same column. The upper bottom
replaceable reflector block (the 8th layer) has the same arrangement of coolant channels as the
top replaceable reflector block. The lower bottom replaceable reflector block (the 9th layer)
has 6 large coolant channels.
FIG. 2.15. Permanent reflector region.
21
FIG. 2.16. Schematic of control rod.
22
FIG. 2.17. Axial control rod position.

Dummy fuel blocks are placed in the active core before fuel loading as substitution for
fuel assemblies. The external form of the dummy fuel block is the same as that of a fuel
graphite block. The dummy fuel blocks contain higher impurity than the fuel graphite blocks.
All dummy fuel blocks are replaced by fuel assemblies. All blocks except the bottom block of
the control rod guide column and the irradiation column are 58 cm in height and 36 cm in
width of across flats. The bottom blocks of the control rod guide and irradiation columns are
48 cm in height and 36 cm in width across the flats. A typical form of the control rod guide
blocks for layers 1 through 5 is shown in Figure 2-14. Boron pins are installed into the lower
parts of blocks in the 9th layer for neutron shielding. For benchmark problems, the neutron
shielding pin region is considered as a black absorber or vacuum in calculation models.
Core components are horizontally surrounded by 12 permanent reflector blocks whose
form is a large polygonal graphite block. The width across the flats of the core,
including permanent reflector blocks, is 4250 mm as shown in Figure 2.15. The permanent
reflector blocks have holes for irradiation tests and neutron detectors. The void fraction is 0.7
% due to the holes. Around the permanent reflector blocks, there are side shielding blocks
consisting of B4C/C. The structures outside of the permanent reflector region were neglected
in the calculation models.
Control Rods
There are 16 pairs of control rods, 7 in the active core and 9 in the replaceable
reflector region. A pair of control rods is inserted into holes of a control rod guide column.
23
Each control rod consists of 10 neutron absorber sections connected with metallic
spines and support rings. Each section contains 5 sintered compacts of B4C and C as neutron
absorber in the annular space. The schematics of the control rod are shown in Figure 2. 16.
The axial positions of the control rods are divided into three types as shown in Figure 2.17.
When the control rods are fully inserted, the lower ends of all control rods are on the same
plane with the bottom face of the 7th layer of a fuel column. In the first approach to criticality
three pairs of control rods at columns E01, E09 and E17 in Figure 2.7 are fully withdrawn.
These are not used to operate the reactor. The rest of 13 pairs of control rods are the
withdrawn during the approach to criticality. At the critical condition, all control rod pairs
except those of columns E01, E09, and E17 are adjusted so that their insertion depth will be
the same. This insertion level is evaluated in the benchmark test HTTR-CR. When the control
rods are fully withdrawn, their upper limit is the upper face of the 1st replaceable reflector
block (over the fuel region) with the exception of those control rods (R2) at columns E03,
E07, E11, E15, E19, and E23 in Figure 2.7. The control rods of these six columns have upper
limits at 725mm below the top of the 1st block.
Further detailed data of the core, fuel assemblies and core internal components have
been provided in [2-5].
Nuclear Instrumentation
Temporary neutron instrumentation consisting of three BF3 counters, three fission
counters (FC), two gamma-ray-compensated ionization chambers (CIC) was used for the
startup core physics tests. Positions of detectors were not arbitrarily chosen like liquid
(water)-moderated reactors. Only the existing holes were available for the solid (graphite)moderated reactor like the HTTR. However, the holes for control rods (CRs) and reserve
shutdown systems could not be used due to safety-related restrictions. Thus the detectors are
located in three irradiation columns as shown in Figure 2.18.
Temporary detectors
of Ch. 1(NS,BF3,FC)
Temporary detectors
of Ch. 3(BF3,FC,CIC)
Irradiation
column
Control rod
Guide column
Fuel
(Loading order)
1
18
2
17
3
16 24 19
25
15 30 28
22
21
13 27 26
29
20
14 23
12
8
11
9
10
4
5
6
7
Replaceable
reflector
Permanent
reflector
Temporary detectors
of Ch. 2(BF3,FC,CIC)
NS : Temporary neutron source
FIG. 2.18. Cross section of the HTTR and fuel loading order.
24
The BF3 is used for monitoring increase in the neutron density. The inverse of the
multiplication factor was evaluated from the change in neutron density. The FC was used for
measurement of the axial neutron flux distribution. The CIC was used for reactivity
measurement by the Inverse Kinetic (IK) method. A temporary neutron source (Am-Be,
1.48104Bq) was inserted in irradiation test column. The vertical position of each detector is
shown in Figure 2.19. The BF3 in Ch.1 is installed about 2 m apart from the neutron source to
avoid the effect of direct-flying neutrons from the source. BF3 and FC in the same channel are
located at the same vertical position in order to have redundancy for 1/M monitoring. FCs
were movable in the vertical direction to measure the axial neutron flux distribution. The
CICs were fixed in the
core. The temporary neutron source was withdrawn from the core to exclude the neutron
source effect on the measurement.
FIG. 2.19. Vertical location of temporary neutron detector system.
2.1.2. Benchmark Problem Descriptions

The benchmark problems addressed within this CRP include HTTR related start-up
core physics tests and thermal hydraulics. General descriptions of these problems are included
below. Details of the methodologies utilized and the results obtained for each problem by the
individual CSIs are presented herein by each Member State.
2.1.2.1. Initial Criticality [HTTR-FC] (Phase 1)
The number of fuel columns are evaluated for the first criticality, with the fuel
columns charged from the outer region of the core. They are loaded clockwise, one by one. A
small excess reactivity at the first criticality is also evaluated.
25
2.1.2.2. Initial Criticality [HTTR-FC] (Phase 2)

The following effects are considered in the Phase 2 to improve the benchmark
problem calculation accuracy:
1) Air in void of graphite
2) Revised impurity contents in dummy block
3) Aluminum in the temporary neutron detector holders.
2.1.2.3. Control Rod Position at Criticality (HTTR-CR)
The control rod insertion depths are evaluated at the critical condition for the
following three cases. All control rod insertion levels are adjusted on the same level except
three pairs of control rods in the most outer region in the side reflectors. These three pairs of
control rods should be fully withdrawn for the calculation.
1) 18 columns (thin annular core)
2) 24 columns (thick annular core)
3) 30 columns (fully-loaded core)
2.1.2.4. Excess Reactivity (HTTR-EX)
The excess reactivity is evaluated for the three cases mentioned above. The room
temperature of 300K is to be assumed as the moderator and fuel temperatures for the
benchmark problem. One atmospheric pressure of helium is to be used as the primary coolant
condition.
2.1.2.5. Scram Reactivity (HTTR-SC)
The Scram reactivity is to be evaluated for the following two cases:
1) All reflector CRs are inserted at the critical condition
2) All CRs in reflector and core are inserted at the critical condition
The core condition for this benchmark problem is as follows:
-Fully-loaded core (30 column fuel core)
-Fresh fuel core
2.1.2.6. Isothermal Temperature Coefficient (HTTR-TC)

Isothermal temperature coefficients for the fully-loaded core are to be evaluated from
the effective multiplication. The critical control rod positions are changed with temperature
elevation in the real reactor operation. However, the control rod position is not to be changed
in the calculation to obtain the reactivity difference. Critical control rod positions are to be
evaluated at temperature of 480K.
26
2.2. REACTOR PHYSICS BENCHMARK ANALYSIS AND RESULTS

2.2.1. Japan
2.2.1.1 Analysis methodology and model description
Diffusion calculation [2-9, 2-10]
The calculations for the benchmark problems were carried out using a nuclear
characteristics evaluation code system which was developed from an HTTR nuclear design
code system [2-11]. The code system consists of the DELIGHT [2-12], TWOTRAN-II [2-13]
and CITATION-1000VP [2-14] codes. An outline of the calculation codes and model is
summarized in Table 2-3. The program structure of the system is shown in Figure 2.20.
Table 2-3. Codes, models and nuclear data library for diffusion calculation
Nucl. Data file

Fuel cell code
Name of country
Name of Institute
ENDF/B-III, IV
DELIGHT
Theory
Collision probability
Model
Pin cell
Cut off energy
2.38eV
No. of groups
40
BP cell code
TWOTRAN-II
Theory
Transport
Model
No. of groups
Control rod cell code
2-D(r-Z)
6
TWOTRAN-II
Theory
Transport
Model
2-D (x-y)
No. of groups
Core cal. Code
CITATION
Model
3-D (Triangle)
No. of groups, (Fast +Thermal)
6, (3+3)
Items
Japan
JAERI
DELIGHT is an one-dimensional lattice burnup cell calculation code that has been
developed in JAERI. TWOTRAN-II is a transport code that was used to provide average
group constants of burnable absorber (BP) in fuel blocks and graphite blocks where control
rods (CRs) are inserted. CITATION-1000VP is a reactor core analysis code. This code was
developed from CITATION [2-15] so that nuclear characteristic analyses could be carried out
with a three-dimensional whole core model of the HTTR in a short calculation time.
27
FIG. 2.20. Program structure of the HTTR nuclear characteristics evaluation code
system.
DELIGHT was used to provide group constants of fuel and graphite blocks for
succeeding core calculations. Resonance, neutron spectrum, neutron flux distribution,
criticality, and burn-up calculations were done sequentially. Nuclear data were based on
ENDF/B-IV except burn-up chain data that were extracted from ENDF/B-III. In the resonance
range, the code employs intermediate resonance approximation and can consider the effect of
a double heterogeneity caused by coated fuel particles (CFPs) and assembled fuel rods. The
average group constants of the whole fuel block were obtained by a fuel cell calculation as
follows; The group constants of the fuel rods were calculated by using a one-dimensional
cylindrical fuel cell model as shown in Figure 2.21. The fuel rods in a block are located in the
inner position of the fuel block, and the outer region of a fuel block is graphite rich region. To
simulate the harder neutron spectrum, the area of cross section of the fuel cell was determined
by the pitch of fuel rods. The pitch of fuel rods is 5.15cm. Therefore, the outer radius of fuel
cell is about 2.7cm.
28
FIG. 2.21. Fuel cell model for DELIGHT code.
The group constants for BP and CR were calculated by TWOTRAN-II code.

BP cell model
The BP of the HTTR is a zebra type configuration. The average group constants of the
zebra type BP was calculated by two-dimensional r-Z model of TWOTRAN-II code. The BP
cell model is shown in Figure 2.22. The BP cell model corresponds to a quarter of a fuel block
which contains half of the BP rod. BP rod consists of BP pellets and graphite pellets which
are surrounded by graphite, the outer region contains homogenized fuel. Macroscopic cross
section sets for the calculation were provided by the DELIGHT code.
In order to obtain the effective microscopic cross section, a homogenized region is
determined to have the same cross section as the BP region in the core calculation model.
Three kinds of a for 10B in the BP rod are evaluated for core calculations. The first one was
for the fuel block with 7.9% enrichment of 235U, 33 fuel pins and 2.0wt% of boron
concentration in the BP rods. This a-set was used for all BPs in the first layer of fuel blocks.
The second one was for the fuel block with 6.3% enrichment of 235U, 33 fuel pins and 2.5wt%
of boron concentration in the BP rods which was used for all BPs with 2.5wt% of boron
concentration. The third one was for the fuel block with 3.9% enrichment of 235U, 33 fuel pins
and 2.0wt% of boron concentration in the BP rods. This kind of a was used for all BPs in the
4th and 5th layer of fuel blocks.
CR cell model
The average group constants of a pair of CRs and of the corresponding graphite
block were obtained with the flux-weighting method. The neutron fluxes were calculated with
a two-dimensional x-y model. The model is based on half of the graphite block where one CR
is inserted.
29
FIG. 2.22. BP cell model by TWOTRAN-II

The average group constant of CRs was obtained by smearing group constants in a
region. For core calculations, two kinds of CR models were developed which are called as
CR-block model and CR-hex model. The smearing regions to obtain average group
constants for CRs are different in the two models. The schematics for each model is shown in
Figure 2.23. The CR-block model was used for all benchmark problems. The CR-hex model
was used for HTTR-SC to check the effect of model.
30
i) CR-block model
The average group constants of CRs were homogenized in a CR guide block region. In
the core calculation for the model, CR guide block was modeled as one region. This model is
the same with the model used in the previous benchmark problem.
ii) CR-hex model
In the model, one CR rod is modeled as hexagonal shape which contains CR and
surrounding graphite. The average group constant of CRs was obtained by smearing a narrow
area in a CR guide block. In the core calculation for the model, the CR guide block is divided
into three regions. One for CRs, one for a reserved shut down system pellet insertion hole, and
one for graphite region. It is possible to consider the position of CRs in a control rod guide
block at core calculations using the model.
In the benchmark problems, CR-block model was used as standard model. The effects
of CR-block model and CR-hex model were evaluated in the benchmark of HTTR-SC.
The CITATION-1000VP is a reactor core analysis code based on the diffusion theory.
This code was improved to enable a full core model calculation of the HTTR by extending the
number of zones and meshes in the original CITATION code and enhancing the calculation
speed by the vectorization of the code. This code was used for the analysis of the effective
multiplication factor. The neutron energy group consists of 3 fast and 3 thermal groups.
FIG. 2.23. Comparsion of CR model concrpt.
31
FIG. 2.24. Configuration of regions in block for core calculation mode
32
A fuel block is divided into 24 triangular meshes horizontally and into 4 meshes
vertically for the three-dimensional whole core calculation to simulate the position of the BP
rods in a fuel block. In the horizontal plane, a fuel block was divided into BP region and fuel
region as shown in Figure 2.24. 10B and 11B in a BP rod were distributed only in the BP
region. Nuclides of the fuel, such as 235U, 238U, Si, were distributed only in the fuel region.
Graphite is distributed homogeneously in the fuel and BP region. For the CR-block model, a
CR guide block was modeled as one region.
Monte Carlo calculation
The whole core is presented with hexagonal lattices consisting of graphite reflector
blocks and fuel blocks. The fuel block is also presented with hexagonal lattices consisting of
fuel rods and BP rods. The fuel rod and BP rod are modeled with cylindrical body
descriptions in hexagonal cells. Effective multiplication factors keff of different cores in fuel
loading were calculated with MVP [2-16, 2-17] using the nuclear data of JENDL-3.2 [2-18].
The thermal neutron treatment for graphite in the JENDL-3.2 is based on S(,) of
ENDF/III. The core temperature of 300K is assumed. The most probable value of keff was
evaluated from track length, collision and analog estimators with the method of maximum
likelihood. Number of histories per batch was 20,000 for all cases. Number of batches was
150. The first 5 batches were neglected for the statistical treatments. Although large number
of scattering reactions must be calculated in one history for HTGR due to the low slowing
down power of graphite, the CPU time is still reasonably small for the whole core analysis.
The computation was carried out on the FACOM VPP-500.
In the geometry description, the hexagonal multiple lattice capability was
used for periodical arrays of fuel rods in a fuel block and of hexagonal blocks in the core.
The cross-sectional model of the 18 column, 24 column and whole core is given in Figures
2.15, 2.26, 2.28 and 2.27, respectively. The outer form of the permanent reflector blocks is
modeled with a cylinder. The cross-sectional model of fuel block is shown in Figure 2-28.
The cylindrical geometry of fuel rods and burnable poison rods were precisely modeled
with body descriptions in hexagonal cells. The voids in control rod guide blocks and
dummy fuel blocks increase neutron leakage because of the neutron streaming through
them. They are also modeled with body descriptions in hexagonal cells. The CFPs in the
compact are treated with the following three models:
(1) Homogeneous Model: All substances in CFPs are smeared with the graphite matrix
in the compact.
(2) Corrected-homogeneous Model: Heterogeneity effect of CFPs, evaluated with
collision probability theory [2-19], is added to the result of (1)
(3) Heterogeneous Model: Heterogeneity effect of CFPs is directory evaluated with
Monte Carlo calculation using statistical geometry model based on MURATA's idea [220].
33
FIG. 2.25. Horizontal cross-section of 18 column core model.
FIG. 2.26. Horizontal cross-section of 24 column core model.
34
FIG.2.27. Horizontal cross-section of 30 column core model (fully-loaded core).
FIG: 2.28. Magnification of analytical model.
35
FIG. 2.29. Vertical cross-section of 30 column core model (fully loaded core).
Total amounts of material such as uranium, oxygen and carbon are conserved in the
compact for each model. About 13,000 CFPs are distributed randomly in a fuel compact.
The keff - calculations for fuel loading steps were carried out in the condition where CRs
were fully withdrawn from the core and reflectors. The vertical cross-sectional model of
core is shown in Figure 2.29.
The vertical length of each component is precisely modeled. The BP is separated in
upper and lower parts in a fuel block. Graphite disks are inserted between upper and lower
parts of BPs. The lengths of BP part and Graphite disk part are 40 and 10 cm, respectively.
36
2.2.1.2. Results of HTTR-FC calculation

The number of fuel columns are evaluated for the first criticality, when the fuels
columns are charged from the outer region of the core. They are loaded clockwise, one by
one. A small excess reactivity at the first criticality is also evaluated. For phase 2, the HTTRFC is re-calculated considering the following effects:
1) Air in void of graphite
2) Revised impurity contents in dummy block
3) Aluminum in the temporary neutron detector holders.
The change in the effective multiplication factor at fuel loading was calculated. The
R2 CRs remain in the top region of the core. The CR model was CR-block model. Fourteen
kinds of microscopic cross section set of fuel were used for the calculations considering the
kinds of 235U enrichment and boron concentration in the burnable absorber rod.
HTTR-FC by diffusion calculation
The change in effective multiplication factor at fuel loading is shown in Table 2-4 and
Figure 2.30. The number of fuel column at the first criticality is 17 columns and the excess
reactivity at the first criticality is about 0.05%k/k.
Table 2-4. Calculation results of HTTR-FC
No. of fuel
column
keff
[-]
Excess
reactivity
[%k/k]
16
0.9921520
-0.79
17
1.0005451
0.05
FIG. 2.30. Change in keff at fuel loading.
37
HTTR-FC by Monte Carlo

The change in effective multiplication factors and excess reactivitys are given
in Figure 2.31 and Table 2-5. The CFP's heterogeneity effect evaluated with the collision
probability theory was 0.012 k for the HTTR fuel. This effect was added to the calculated
keff - values for the corrected-homo. Model. The heterogeneity effect of CFP was evaluated
also from difference in keff of the Homo- and Hetero- models. The results are given in
Table 2-6. Six pairs of control rods in side reflector can not be withdrawn form the top
reflector perfectly, because a sufficient negative reactivity addition rate must be supplied
for reactor scrams. The CR insertion was simulated in the calculation. The fuel number for
the first criticality is estimated for 18 fuel columns (Hetero. model).
Table 2-5. Calculated k eff and excess reactivity to loaded fuel columns
Heterogeneous model
Homogeneous model
Fuel columns
keff*
ex**
keff
ex
12
0.959596 0.00051
-4.21
14
0.97429 0.00056
-2.64
16
0.98461 0.00055
-1.56
0.98052 0.00054
-1.99
17
0.99506 0.00052
-0.50
18
1.00609 0.00047
0.61
1.00266 0.00050
0.27
19
1.01626 0.00046
1.60
21
1.04827 0.00047
4.60
24
1.09968 0.00047
9.06
1.09350 0.00039
8.55
27
1.13399 0.00044
11.82
30
1.14278 0.00039
12.49
1.13258 0.00045
11.71
*: keff is Effective multiplication factor. The variation shows the statistical error.
**:ex is Excess reactivity in %k/k.
Table 2-6. Heterogeneity effects of CFP
Fuel columns
38
ex(heterogeneous)-ex(homogeneous)
(%k/k)
16
0.43
18
0.34
24
0.51
30
0.78
FIG.2.31. Change in excess reactivity with fuel loading.
The experimental results for HTTR-FC were as follows: The fuel loading was started
from the core periphery to core center so that the annular core will be constructed on the way
to the full core, to obtain the nuclear characteristics. Fuel blocks are loaded clockwise in the
core periphery as shown in Figure 2.25. The fuel loading was carried out by replacing dummy
blocks with fuel blocks. The annular core was made when the core was loaded with fuel
columns from 18 to 24. Inverse multiplication factors (1/M) were evaluated at 0, 6, 9, 12, 15,
16, 17 and 18 fuel column-loaded cores to predict the first criticality. The first criticality was
achieved at the 19 fuel-column-loaded core.
2.2.1.3. HTTR-CR
The control rod insertion depths are evaluated at the critical condition for the
following three cases:
1)
2)
3)
18 columns (thin annular core)

24 columns (thick annular core)
30 columns (fully loaded core)
39
All control rod insertion levels are adjusted on the same level except three pairs of
control rods in the most outer region in the side reflectors. These three pairs of control rods
are fully withdrawn for the calculation. (The fully withdrawn position of CRs (C, R1, R3) are
over the top of the replaceable reflector. The fully withdrawn position of the six control
rods of R2 is 725 mm below the top of the 1st replaceable reflector block.)
HTTR-CR by diffusion calculation
The control rod position at criticality for 18-columns loaded core, 24 columns-loaded
core and fully loaded core are shown in Table 2-7.
Table 2-7. Calculation results of HTTR-CR
No. of fuel
column
18
24
30
Control rod position at

criticality [mm]
3035
2055
1665
HTTR-CR by Monte Carlo calculation

The analytical method for HTTR-CR is as same as that for HTTR-FC. The results are
as follows:
Model and column number
CR position (mm)
18 (Corrected homo. model)
2810
24 (Corrected homo model)
2080
30 (Hetero model)
1800
The experimental results for HTTR-CR were as follows: The reactor was made
critical in every fuel loading step after the first criticality to the full core. The steps were 21,
24, 27 and 30 columns. The CR position at every critical condition was measured to evaluate
calculation accuracy. The CRs are inserted from top to bottom. The tops of C-, R1- and R2CRs were kept the same level at critical conditions. The R3-CRs were fully withdrawn. The
CR position is defined as distance from the boundary between the fuel region and the bottom
reflector. The change in the critical CR position is given in the following Table 2-8:
Table 2-8. Measured critical control rod positions
Fuel column
Rod position(mm)*
21
26465
24
22155
27
18995
30
17755
*Distance from the boundary between fuel region and the bottom reflector. Sinking of
CR driving mechanism (14mm) was considered.
40
2.2.1.4. HTTR-EX
The excess reactivity is evaluated for the three cases mentioned in HTTR-CR. The
room temperature of 300K is assumed as the moderator and fuel temperatures for the
benchmark problem. One atmospheric pressure of helium is used as the primary coolant
condition.
HTTR-EX by diffusion calculation
The excess reactivity at 18-columns loaded core, 24 columns-loaded core and fully
loaded core are shown in Table 2-9.
Table 2-9. Calculation results of HTTR-EX
No. of fuel
column
18
24
30
keff
[-]
1.0126343
1.1014290
1.1442246
Excess reactivity
[%k/k]
1.2
9.2
12.6
HTTR-EX by Monte Carlo calculation

The analytical method for HTTR-EX is as same as that for HTTR-FC. The results are
as follows:
Model and column number
Excess reactivity (%k/k)
18 (Hetero. model)
0.61
24 (Hetero. model)
9.06
30 (Hetero model)
12.5
The experimental results are as follows: Fuel addition method was applied for the
excess reactivity measurement. The increment in excess reactivity was measured by IK
method at 21, 24, 27, and 30 column-loaded core. The measurement was affected by the
negative shadowing effect. The measured increments in excess reactivity were revised with
following relation to correct the shadowing effect. The excess reactivity for each core are
given in Table 2-10.
ex =
ex
ex = ikm R
R =
excal
ikcal
ex :Excess reactivity (%k/k)

ex :Increment in excess reactivity (%k/k)
41
ikm :Increment in excess reactivity measured with IK method (%k/k)

excal: Increment in excess reactivity calculated after definition.
All control rods are fully withdrawn from the critical rod position (%k/k)
ikcal: Increment in excess reactivity calculated after IK measurement procedure (%k/k)
R: Revising factor
Table 2-10. Excess reactivity
Fuel columns
21
24
27
30
(eff=0.0065)
ikm (%k/k)
ikm (%k/k) R
2.3 0 .23
5.0 0.50
7.4 0.74
8.5 0.85
2.3 0.23
2.7 0.27
2.4 0.24
1.1 0.11
ex (%k/k)
1.69 0.12
1.40 0.11
1.26 0.15
1.19 0.04
4.0 1.1
3.7 1.0
3.0 0.9
1.3 0.3
ex (%k/k)
4.0 1.1
7.7 2.1
10.7 3.0
12.0 3.3
2.2.1.5. HTTR-SC
The core condition is as follows:
-Fully-loaded core (30 column fuel core)
-Fresh fuel core
The following two cases are investigated the HTTR-SC:
1) Scram reactivity of reflector CRs (R2 and R3)
The Scram reactivity of the reflector CRs is evaluated for the fully-loaded core
as follows.
R =
k Crit . k RCR in
k Crit . k RCR in
Where:
R: Scram reactivity of reflector CRs (k/k)
kCrit.: Effective multiplication factor at critical CR position
kRCR-in: Effective multiplication factor at CR position after scram
The effective multiplication factors is calculated at the critical CR positions and the CR
positions after scram. The CR positions are given in Table 2-11. The reflector CRs are fully
inserted after scram. The positions of the in-core CRs are not changed. The temperature for
core and reflector is 300K.
Table 2-11. Control rod position before and after scram of reflector CRs*
CR Group
Critical position
Position after scram
Remark
(mm)
(mm)
C
1775
1775
Not changed
R1
1775
1775
Not changed
R2
1775
-55**
R3
Full out
-55
*:Case (T4-3005), No neutron source
**:Control rods are inserted slightly into the top of the bottom reflector
42
2) Scram reactivity of all control rods (C, R1,R2 and R3):

The scram reactivity of all CRs is evaluated from the effective multiplication factors as
follows:
A =
k Crit . k ACR in
k Crit . k ACR in
Where:
A: Scram reactivity of all CRs (k/k)
kCrit.: Effective multiplication factor at critical CR position
kACR-in: Effective multiplication factor at CR position after scram
The effective multiplication factors is calculated at the critical CR positions and the CR
positions after scram. The CR positions are given in Table 2-12. The all CRs are fully
inserted after scram. The temperature for core and reflector is 300K.
Table 2-12. Control rod positions before and after scram of reflector
and in-core CRs*
CR Group
Critical positions
Positions after scram
Remark
(mm)
(mm)
C
1775
-55**
R1
1775
-55
R2
1775
-55
R3
Full out
-55
*:Case (T4-3005), No neutron source
**:Control rods are inserted slightly into the top of the bottom reflector
.
The measured scram reactivity was 0.46k/k

HTTR-SC by diffusion calculation
The scram reactivity should be calculated from the criticality condition with CR
position at 1789mm. In the calculation model, it is difficult to set the CR position at 1789mm
because the vertical mesh division is 145mm. The scram reactivity was evaluated by
interpolating the results of 1740mm and 2030mm.
The calculated results of HTTR-SC by the CR-block model and the CR-hex model are
shown in Table 2-13. The results of CR-block model was smaller than that of the CR-hex
model. In the CR-hex model, reflector CRs are nearer to the core than in the CR-block model.
Therefore, scram reactivity becomes larger.
The comparison of measured and calculated value of the scram reactivity are shown in
Table 2-14. For the results of all CRs, the error of CR-block model and CR-hex model are
about 7% and 4%, respectively. The difference of both models is small. Both model show
good agreement with measured results.
For the results of reflector CRs, the error of CR-block model and CR-hex model are
about 31% and 26%, respectively. The CR-hex model shows less error than CR-block
mode, but still shows more than 20% of error. The calculation model should be further
revision.
43
Table 2-13. Calculated results of HTTR-SC

(1)Scram reactivity of reflector CRs(R2 and R3)
CR model
CR-block model
CR position before scram [mm] 1740
2030
Before scram
1.0093164 1.0418115
keff
R2, R3 full in (0mm) 0.9311435 0.9599020
Scram reactivity [%k/k]
8.32
8.19
Scram reactivity from 1789
8.30
mm [%k/k]
(2) Scram reactivity of all CRs
CR model
CR position before scram [mm]
Before scram
keff
All CRs full in
(0mm)
Scram reactivity from 1789
mm [%k/k]
CR-hex model
1740
2030
1.0084132 1.0409116
0.9247989 0.9534538
8.97
8.81
8.94
CR-block model
1740
2030
1.0093164 1.0418115
CR-hex model
1740
2030
1.0084132 1.0409116
0.7069916
0.6980522
42.37
45.46
42.9
44.09
47.19
44.6
Table 2-14. Comparison to the measured value

(1) CR-block model
Measured (E)
Calculated (C)
Reflector CRs 12.1
8.30
All CRs
46.3
42.9
Error [%]
C/E-1
-31.4
-7.3
(2) CR-hex model

Measured (E)
Calculated (C)
Reflector CRs 12.1
8.94
All CRs
46.3
44.6
Error [%]
C/E-1
-26.1.
-3.7
HTTR-SC by Monte Carlo Calculation

The analytical method for HTTR-SC is the same as that for HTTR-FC. The results are
given in Table 2-15:
A two-step scram will be performed in the HTTR to prohibit the high temperature
exposure of the in-core CRs in full power operation. The in-core CRs are inserted when the
core temperature becomes less than 750C or 2400 seconds passes after insertion of reflector
CRs. The simulation of the two-steps scram was conducted in startup core physics. The incore CRs were inserted in the time delay of 20 seconds after insertion of the reflector CRs.
The time delay was determined so that the neutron density signal from the CIC did not
become too small for reactivity measurement. The reactivity was measured continuously by
the IK method. Digital computer was used to evaluate the reactivity from change in neutron
density signal.
44
Table 2-15. Scram reactivity of control rods

Items
Scram Reactivity
1) Scram
reflector CRs
reactivity
of
0.99908 0.00045
kcrit
0.91222 0.00058
kRCR-in
0.0953 0.0007
R(k/k)
2)
Scram reactivity of all CRs
kcrit
0.99908 0.00045
0.68873 0.00064
kACR-in
0.4510 0.0008
A(k/k)
Table 2-16. CR position at two step scram*

CR position [mm]
F
S
B
R
irst
step econd step
efore
scram
scram
scram
1
1
7755
7755
555**
1
1
1
7755
7755
555**
1
2
7755
555** 555**
F
3
ull out
555** 555**
*:Sinking of CR driving mechanism was considered
**: Control rods are inserted slightly below 0 mm at scram.
Table 2-17. Scram reactivity at two-step scram
Scram
Reflector
CRs
All CRs
(eff=0.007)
Error
Measured
scram C/E1
reactivity [k/k]
(%)
0.120.012
-17
0.460.046
10
45
2.2.1.6. HTTR-TC
Isothermal temperature coefficients for fully-loaded core are evaluated from the
effective multiplication factors by the following relation:
n =
k n+1 k n
1
k n +1 k n (Tn+1 Tn )
Where: n: Temperature coefficient between Tn and Tn+1 (k/k/K)

Tn: Core temperature at nth measurement (K)
Tn+1: Core temperature at n+1th measurement (K)
kn: Effective multiplication factor at Tn
kn+1: Effective multiplication factor at Tn+1
The critical control rod positions are changed with temperature elevation in the real
reactor operation. However, the control rod position should not be changed in calculation to
obtain the reactivity difference. The critical positions given in Table 2-18 as CR position are
to be used for calculation of effective multiplication factors. The effective multiplication
factors are calculated for the following temperatures:
280,
300, 340, 380, 420, 460, 480 (K)
Temperature coefficient for following temperatures are evaluated from the effective
temperature coefficients with the above relation.
290
320 360 400 440 470
(K)
Additionally, the critical control rod positions are evaluated at temperature of 480K.
The insertion depths of C, R1 and R2 are at the same level. All of R3 is fully withdrawn.
Table 2-18. Control rod position for temperature coefficient evaluation
Critical positions*
(mm)
C
1777
R1
1777
R2
1777
R3
Full out
*These values are obtained from the critical positions (T4-3005) by correcting
with lifting effect of CR driving mechanism (about -14mm) and temperature
effect from 25 to 27 (about 2mm).
CR Group
HTTR-TC by diffusion calculation

The isothermal temperature coefficients were evaluated using the CR-block model. It
was impossible to calculate isothermal temperature coefficients directly due to mesh division
of the core calculation model. Therefore, the isothermal temperature coefficients at 1776mm
were evaluated by interpolating the results at 1740mm and 2030mm. In the calculation, four
kinds of microscopic cross section set were used. Each layer of fuel block use same
microscopic cross section set. Air in void of graphite and Al of temporary neutron detector
holders were not considered.
46
47
Table 2-20. Criticality CR position at 480K

C, R1, R2
R3
CR position [mm]
1825
4060
HTTR-TC by Monte Carlo Calculation
The analytical method for HTTR-TC is as same as that for HTTR-FC. The results are
given in Table 2-21:
Table 2-21. Temperature coefficients of HTTR core
Items
Temperature coefficients
Effective multiplication factor for

different temperatures
k300
0.99762 0.00054
k340
0.99273 0.00051
k380
0.98621 0.00050
k420
0.97990 0.00050
k460
0.97393 0.00057
k480
0.97221 0.00051
Temperature coefficients (k/k/K)
320
-1.23 10-4
360
-1.66 10-4
400
-1.63 10-4
440
-1.56 10-4
470
-0.91 10-4
Critical control rod position at 480K

C, R1, and R2 (mm)
R3
1789
Full out
The control rod positions at critical condition were measured for three different core
temperatures (T). The core temperature was obtained by averaging core-inlet and -outlet
coolant temperatures. The control rod positions and core temperatures are summarized in
Table 2-22:
48
Table 2-22. Control rod positions and core temperature

CR group
CR position (mm)
T=22.7
T=122.0
T=144.9
C
1751*5
18735
19035
R1
1751*5
18735
19035
R2
1751*5
18735
19035
R3
Full out
Full out
Full out
*:Sinking of CR driving mechanism (14mm) was considered.
**After taking out temporary detectors from core
The temperature effect was evaluated from CR positions, using calculated control rod
worth curve. The isothermal temperature coefficients are given in the following Table:
Table 2-23. Isothermal temperature coefficient
Temperature range ( )
22.7~122.0
122.0~144.9
Iso.temp.coeff. (k/k/T)
-1.34E-4*
-1.42E-4*
49
2.2.2. France
2.2.2.1 General analysis method and model description
Codes and calculation scheme
The French reactor physics code system SAPHYR has been used in the following
HTTR calculations. SAPHYR gathers several codes developed at CEA like APOLLO2 [2-23]
(transport) based on a database produced with THEMIS/NJOY, CRONOS2 [2-24] (diffusiontransport), FLICA4 (3D- thermal hydraulics), ..., which are interconnected. This code system,
initially dedicated to PWR calculations and research & development purposes, seems to be
well adapted for the assessment of the HTGR performances and characteristics. Finally, the
Monte-Carlo code TRIPOLI4 [2-25] has also been used throughout the study.
All the HTTR problems proposed in reference [2-21] have been treated considering
two calculation methods: one based on a Transport Diffusion calculation scheme and a
second one based on a Transport Monte-Carlo calculation scheme. Figure 2.33 illustrates the
general procedure. The standard 172-groups or point-wise cross sections library issued mainly
from JEF-2.2 are used for the calculations.
JEFF2.2
APOLLO2
1D cylindrical cell transport calculation
172 groups - Pij method - Double heterogeneity
172 group-fuel compact cross sections
APOLLO2
Pointwise cross sections
Fuel and control blocks transport calculations

172 groups - Pij method - general geometry
TRIPOLI4
N group-fuel and control block cross sections
Monte Carlo core

calculation
CRONOS2
N groups diffusion core calculation
Core keffectif
Core keffectif
FIG. 2.33. Description of both calculation schemes.

It is noteworthy that as far as the Transport-Diffusion calculations are concerned, two
different approaches have been used in terms of core modelling. The first one, a rough model
with strong hypotheses (homogenised fuel block, no streaming effect, ) led to preliminary
predictions of the HTTR-FC, EX and CR problems (first results). Then, enhancements to
50
improve the modelisation have been identified, assessed and finally implemented in course of
revised calculations using new benchmark data proposed in [2-27]. This has been done for all
the HTTR benchmarks (FC, EX, CR, SC and TC) and constitutes the second prediction (final
results).
Cylindrical calculation of the fuel compact
Knowing that the stochastic geometries calculations (coated fuel particles - CFP randomly distributed in the fuel compact) is not available in the Monte-Carlo code TRIPOLI4,
a 1D-cell calculation has been performed as a first step. It takes into account a precise
spherical description of the particles with all their coatings which themselves fill the annular
part of the cylindrical geometry of the fuel compact (Figure 3.34) with a packing density
around 30 %. The self-shielding of the uranium is calculated during this calculation stage. A
collision probability method is used to solve the transport equation with 172 energy groups.
The critical buckling search allows taking into account the neutron leakage by the addition of
a homogeneous leakage term in the form of DB2. The extra region of the cylindrical cell is
representative of the other fuel cells disposed around with the triangular pitch of the fuel rods
in the assembly.
FIG. 2.34. Cylindrical Fuel Cell Model.
This first stage provides fuel compact-homogenized 172-group cross sections for
Monte-Carlo core calculations (TRIPOLI4). Therefore, in the core calculations performed by
TRIPOLI4, point-wise cross sections are used everywhere in the core
except in the fuel rod region where the multigroup cross sections have been generated with
APOLLO2.
Although that the CFP might be directly considered in the 2D fuel element
calculations, as for TRIPOLI4 the same fuel compact averaged 172-group cross sections
have been used in the 2D transport model described hereafter. This avoids calculating the selfshielding in the 2D configuration and results in a large saving of CPU time without making
severe assumption.
51
Fuel block 2D-calculations

This stage must lead to an averaged flux weighted library of the different fuel blocks existing
in the core. This library of n-group-constants is directly read during the core diffusion
calculations. The energy structures (n = 2 to 20) are given in annexe. A 172 group-collision
probability method is used in the 2D-fuel block geometry (Figure 2.35). The fuel handling
hole zone is modelled by a lower graphite density. As far as the burnable poisons (BP) are
concerned, the B4C and graphite pellets are axially homogenised with the graphite matrix in
order to reduce the number of calculations from 32 to 16 and to simplify the core calculations.
The impact of this axial homogenisation is important and has been evaluated hereafter.
FIG. 2.35. 2D-fuel block geometry.

These 2D calculations were performed in infinite medium with a critical buckling
search. A B1 homogeneous neutron leakage model has been retained as a first step. This led to
isotropic diffusion coefficients and did not allow taking into account the streaming effect.
The Pij model gives a flux evaluation in each region depicted in Figure 2.35. An example of
the power distribution in a fuel block is given in Figure 2.36 (the map is normalised to 100).
FIG. 2.36. Power Distribution in the f343320 fuel block.
52
The 2D-results of the fuel blocks in fundamental mode are gathered in the Table 2-24
below and an example of the average neutron flux obtained for the lower and the higher
enriched fuel blocks are given in Figure 2.37.
Table 2-24. 2D-Transport Calculations on the Fuel Blocks
Fuel Block
f343320
f393320
f433120
f483120
f433325
f523325
f593125
f633125
f633325
f723125
f793125
f673320
f793320
f943120
f993120
Enrichment
[% masse]
3,4
3,9
4,3
4,8
4,3
5,2
5,9
6,3
6,3
7,2
7,9
4,6
7,9
9,4
9,9
Multiplication factor
1,10650
1,14991
1,18391
1,21366
1,15656
1,20380
1,24481
1,25798
1,24962
1,28937
1,30862
1,28423
1,31733
1,36335
1,37213
Migration Area
[cm2]
491,05
487,14
486,43
481,82
476,90
470,53
469,50
468,43
465,50
463,70
457,99
467,60
459,52
458,81
455,45
FIG. 2.37. Neutron Spectrum in the fuel block

Control blocks 2D-calculations
From the previous calculations, the average fluxes of the fuel blocks were used as a
172 group-neutron source placed at the periphery of the control blocks to generate the
averaged weighted library needed for core diffusion calculation. A similar collision
probability method has been used in the 2D-geometry shown in Figure 2.38 for the control
blocks with and without the control rods inserted.
53
As for the fuel blocks, the n-group-constants have been created by spatial
homogenisation on the overall control block geometry. Indeed, among the finite element
meshes available in the CRONOS2 diffusion code, it was first impossible to take into account
the position and the orientation of the fuel and control blocks by an heterogeneous geometric
description of these elements. Moreover, no transport-diffusion equivalence factors have been
considered for the control blocks when the control rods are inserted. The averaged neutron
fluxes obtained in the control blocks are given in Figure 2.39.
FIG. 2.38. 2D-Control Rod Blocks Geometry.
FIG. 2.39. Neutron Spectrum in the Control Rod Blocks.
54
Reflector cross section generation

In the diffusion calculations, the n-group-cross sections of the replaceable reflector
have been evaluated with similar methods to those described in the previous section where a
neutron source is placed at the periphery of the blocks. As far as the permanent reflector
region is concerned, the cross sections come from a 1D-cylindrical core transport calculation
(in its fully loaded configuration) performed with the Sn method.
Core diffusion calculations
An hexagonal 3D-geometry has been used for the core description. Each fuel and
control block is represented by a hexagon with a homogeneous composition. On this
hexagonal mesh, the finite element method provides several polynomial approximations. In
this case, a second order exact integration (standard parabolic Lagrange) has been adopted
horizontally and leads to 19 flux values per mesh (equivalent to a 24 homogeneous meshes in
finite difference method).
FIG. 2.40. Core Axial Description for CRONOS2 code.

Axially, a linear Lagrange method using Gauss Formula has been applied. Each fuel
section (14 fuel compact stack) was divided into 6 meshes and the inter-space between fuel
sections (3.4 cm of graphite) has been described explicitly per Fig. 2.40.
Transport-Monte Carlo core calculations
In spite of the homogenised fuel compacts, the core model developed here is a very
detailed 3D-model of the HTTR with all components modelled explicitly. For these
components, point-wise cross sections based on the JEF2.2 evaluation were used. The fuel
compact was represented by a set of 172-group-constants coming from the previous transport
calculation. It is noteworthy that in this case the burnable poison had the exact axial
description in opposition to the homogeneous axial description used in the core diffusion
calculations.
55
FIG. 2.41. Cross Section of Fully Loaded Core for TRIPOLI4 code.
Only the core configurations with 18 and 30 columns have been considered here.
Figures 2.41 and 2.42 show a cross section view of the fully loaded core and detailed axial
and radial cross section views of the control rod and fuel blocks.
FIG. 2.42. Details of the Geometry used in the TRIPOLI4 code.

Synthesis of the assumption of the initial diffusion calculation scheme
The key factors in modelling the HTTR with the transport-diffusion calculation
scheme are the following:
the BP are axially homogenized
the streaming effect is not taken into account everywhere in the core
homogenised hexagons are considered without equivalent factor in order to respect the
absorption rates of the BP and the control rod (CR) between both transport and
diffusion calculation stages
flux-weighted cross sections are used for the reflector
CR insertion on the top reflector not taken into account
56
The following section provides the results taking into account these assumptions.
Then, their impacts on the results have been assessed. They have been reconsidered in course
of the revised calculations according to the results of these assessments.
2.2.2.2 First Results
HTTR-FC and EX
The calculations performed with the two different methods are gathered in Table 2-25.
It is important to note that all the keff-values of the Table 2-25 are corrected with a k of 0.004 in order to take into account the presence of the control not fully withdrawn to the top
of the reflector and not considered in the core models.
All the preliminary calculations underestimated the number of fuel columns needed to
achieve the first criticality (Diffusion calculations: 10 columns; Monte-Carlo calculations: 17
fuel columns). As it can be seen in Table 2-25, the discrepancy between the calculations and
the experiment at least ranges from k = 0.017 to 0.058 at 18 fuel columns loading and from
k = 0.01 to 0.033 at full core.
It is noteworthy that the observed discrepancies decrease with increasing number of fuel
columns in the core. Due to the large experimental uncertainty at 30 fuel columns loading, the
differences between the calculations and the experiment are within the error bar, whereas at
the thin annular core assembly the discrepancies are significant. Two reasons for the latter
circumstance can be proposed. The first would be that the two steps transport-diffusion
calculation based on the fundamental mode assumption would be less and less appropriate as
one goes toward the annular core configuration. The second would concern the level of the
actual boron impurity in the dummy fuel blocks and of the residual air (instead of helium) in
the graphite pores. As far as the latter is concerned, the impurities of some dummy fuel blocks
have been re-measured by JAERI and revised data [2-27] have been recommended for the
recalculation of the first criticality (HTTR-FC2).
Table 2-25. Experimental, Monte-Carlo and First Diffusion Results
30 col
28
26
24
22
20
19
18 col
16
14
12
10
9
CRONOS2
3D - (first model)
diffusion 8 gr
homog. Fuel block
no streaming
axially homog. BP
1.1698
1.1691
1.1596
1.1399
1.1158
1.0886
1.0745
1.0580
1.0383
1.0284
1.0184
1.0044
.9970
k
[%k/k]
14.5 %
14.5 %
13.8 %
12.3 %
10.4 %
8.14 %
6.93 %
5.48 %
3.69 %
2.76 %
1.81 %
0.44 %
- 0.3 %
TRIPOLI4
3D
Monte-Carlo
172 gr & pointwise
k
[%k/k]
EXPERIMENT
1.14630 0.0009
12.8 %
1.13630 (> 3.6%)
1.01710 0.0009
1.68 %
subcritical
57
In the course of the studies, the following reasons for the above mentioned
discrepancies (especially for the simplified Transport Diffusion calculation scheme) have
been identified and quantified:
A non-adequate treatment of the axial self-shielding in the BP rods,

An underestimation of the neutron streaming (due to large channels of the CR blocks),
The neglect of the detailed structure of the HTTR fuel block in the core calculations.
These main physical effects and their impacts on the core reactivity are briefly depicted
in Figure 2.43 for the case of the annular core configuration. Similar tendencies can be
observed for the full core configuration. Nevertheless, different absolute values (into bracket
Figure 2.43) are obtained for the quantified physical effects due to the harder neutron
spectrum in the fully loaded core. Indeed, the observed weight of the boron absorption in the
BP is different in this case.
Case
Case 33 == case
case 22 ++
homogenised fuel blocks
~ 4 % (2.5 %)
Reference
calculation
(Monte Carlo
Referen
method
calculation
Case
Case 22 == case
case 11 ++
without streaming effect
~ 1.5 %
(2.2 %)
~ 1.8 %
(1.3 %)
Case
Case 11 == Reference
reference+ +
axially
BP
homogenised
BP
Increasing
reactivity
FIG. 2.43. Key factors in modelling and their impact on the reactivity for the 18 fuel columns
core configuration (and full core).
For the thin annular core, starting from a best estimate calculation and neglecting the
fact that the burnable poison was axially a succession of boron and graphite pellets leads to a
predicted reactivity 1.5 % lower than considering the actual heterogeneous composition of
these burnable poisons (3D Monte Carlo estimation). On the opposite, a core calculation that
does not take into account the streaming effect will result in an increase of the reactivity of
about 1.8 %. At this stage, it is interesting to note that by making two strong physical
hypotheses a result not far from the best estimate calculation can be obtained. Finally, a
discrepancy on the order of 4 % can be achieved if an insufficient description of the fuel block
is used to model the high level of radial heterogeneity (2D Monte Carlo estimation).
Therefore, the HTTR-FC2 benchmark has been a good opportunity to implement the
new enhanced methods coming from this analysis and to evaluate the progress considering the
new data.
58
HTTR-CR
Transport-diffusion calculations
The control rod insertion depths have also been evaluated to achieve criticality in the three
configurations recommended by the benchmark problem HTTR-CR (thin and thick annular
core and fully loaded core). The results corresponding to the transport-diffusion method are
given in the following Table 2-26 with the calculational approach illustrated in Figure 2.44.
Table 2-26. Critical Insertion Depths of the Control Rod

Configuration:
(thin annular core) 18
(thick annular core) 24
(fully loaded core) 30
Control rod critical positions from

the bottom of the active core
251 cm
181 cm
151 cm
FIG. 2.44. Effective multiplication factor as a function of the control rod position.
It should be noted that the core model used here does not allow taking into account the
orientation of the control rods in the homogeneous block description, as it is also the case for
the position of the burnable poison in the fuel element.
Transport-Monte Carlo Core Calculations
As far as the second method is concerned (transport-Monte Carlo) it is noteworthy that, as
recommended by the benchmark problem, the structural materials of the control rod have not
been taken into account. Conversely, a detailed axial description of the control rods has been
done as depicted in Figure 2.45.
59
FIG. 2.45. Axial Cross Section View of the Inserted Control Rods.
The calculations aimed to tentatively evaluate the critical insertion depth of the control
rod in the fully loaded core configuration have been performed. Two control rod insertion
depths have been considered. The first one -178.7 cm- is the experimental value and the
second one -170 cm- comes from an estimation of the control rod efficiency obtained in the
diffusion calculations. The results are given in the Table 2-27 below:
Table 2-27. Reactivity for Two Control Rod Positions (Full Core)
Configuration:
Control rod position from the

bottom of the active core
178.7 cm
170.0 cm
keff
1.008500.0009
0.998400.0009
From the results above, the critical rod position can be evaluated to 171 cm. As a
conclusion, one can note that, compared with the Monte Carlo, the control rod worth is
overestimated with the diffusion method (without utilizing equivalence factor).
2.2.2.3 Modification to Model and Assessment of Improvements
New finite elements in the core diffusion model
New finite elements recently implemented in CRONOS-2 have been used. They allow
taking into account the exact position of the burnable poison in the fuel blocks and the fuel
element orientation in the core.
Indeed, from the 2D transport calculations illustrated on the Figure 2.34, the fuel
element was initially homogenised in one hexagonal finite element. Then, with the help of the
new available finite elements, two different meshes were considered to describe the fuel
elements with 24 radial meshes: 24 equilateral triangles (type I) or the cutting out depicted in
Figure 2.46 (type II). Only the last one has been kept in the final model because of the fact that
it is the only one that allows homogenising the poison with its associated graphite without
homogenising partially the fuel compacts. Therefore, the fuel element structure is described by
using three different mediums and the flux is calculated for each point described in Figure
2.46 (61 points for the hexagonal element).
60
Fuel element description

in transport calculations (APOLLO2)
Fuel element geometry
Fuel element description

in diffusion calculations (CRONOS2)
FIG. 2.46. Fuel element modelling in the improved Transport Diffusion calculation scheme.
When such a heterogeneous fuel block geometry is used in the diffusion calculations,
the impact have been evaluated for the three core configurations with 18, 24 and 30 columns,
on the basis of a 2D simplified core with no axial leakage and with an average uranium
enrichment. The diffusion calculations are compared to the Monte Carlo one. The results
obtained for the first configuration are presented in Figure 2.47.
FIG. 2.47. 2D calculation comparison for different energy group structure in the diffusion
calculation - 18 columns core -
61
As previously mentioned, the most important impact is obtained for the 18 columns
core loading. As far as the full core is concerned, the diffusion-Monte Carlo 2D-discrepancies
become quite acceptable with the use of the new finite elements. Moreover, the use of
equivalence factors has been implemented in order to respect the global absorption rate
between the APOLLO-2 transport calculations (172 groups) and the CRONOS-2 diffusion
calculations with few groups. This option has not been considered afterwards because of it
small impact (Figure 2.47) on the finite element of type II.
As a conclusion, a detailed description of the fuel block improves largely the results by
giving a higher weight to the BP absorption in the fuel blocks. It allows getting quite
acceptable values comparing to the reference TRIPOLI4 2D-calculation for the full core but a
remaining discrepancy of about 1% can be observed for the annular core. This could be
attributed to the cross section generation stage where the environment of the BP would not be
representative of the one existing in the annular core configuration. Indeed, in this
configuration the BP is surrounded by much more graphite (reflector) that thus increases the
absorption flux-weighted cross-section. Besides, it would appear that there is actually no
specific trend concerning the energy structure to be retained in the CRONOS-2 calculations.
Streaming modelling in the transport calculations
For improving the Transport Diffusion calculation scheme, the streaming effect has
been taken into account by using anisotropic diffusion coefficients in the core calculations.
These diffusion coefficients have been evaluated in the fuel element transport calculation
performed by APOLLO2. The Benoist method [2-29] available in APOLLO2 (called TIBERE
model) is based on the B1 heterogeneous neutron leakage model. However, it might not be
applicable in the large HTTR channels of the control rod graphite blocks (three large channels
per block). Therefore, another analytical model (Benoist [2-30]) has also been tested on one
control rod block alone and compared to the TIBERE model. With this formulation, the
corrected diffusion coefficient is given by:
Dk
1
m
3
= 1+
Vc
Vt
c
1 +
Qk
m
(1)
where:
Qr
= 11
QZ
= 2
3
BZ c
4
k = r, z (radial or axial)
c is the channel radius, m is the mean free path of the moderator (graphite)
Vc = c 2 is
and: =
the channel volume, Vt is the cell volume, BZ is the axial buckling
+1
2b'
V
1 b' c
Vt
, b' =
1
2
+1
, =
In order to validate these models (TIBERE and analytical model), Monte-Carlo and
diffusion calculations have been performed on the simplified geometry: the control rod block
is surrounded by fuel elements and the axial structure of the geometry is the same as HTTRs
core. The results are gathered in Table 2-28.
62
TABLE 2-28. Streaming Effect Calculated on the Simplified Core

TRIPOLI4
keffectif 3
Homogeneous control block
CRONOS2
2 gr.
8 gr.
1,27958 0,00090
1,26320
1.27397
1,27247 0,00100
1,26040
1,27057
1,25657
1,26672
222
267
526
570
Heterogeneous control block

TIBERE
Analytical formulation
Streaming effect [pcm]
TIBERE
Analytical formulation
560 135
The streaming effect calculated by CRONOS2 is underestimated when the anisotropic

diffusion coefficients are evaluated by the TIBERE model. The underestimation of the
streaming effect is the consequence of a non-adequate calculation of the axial diffusion
coefficient (underestimation of 20 to 25 % on Dz calculation). On the other hand, the use
of an analytical model for the axial diffusion coefficient allows obtaining results in good
agreement with the reference (Monte-Carlo method).
2.2.2.4 New Results
New available data
The new data benchmark (HTTR-FC2) has been defined by JAERI [2-27]. It has been
a good opportunity to implement the new enhanced methods coming from the previous
analyses. Besides, the impact of these new data on the reactivity has been evaluated on the
basis of the Monte Carlo calculations and of the improved core diffusion calculations. These
effects are listed in Table 2-29.
Table 2-29. Analysis of the Benchmark Data Impact on the Results
18
30
Residual air in the graphite porosity and reevaluated impurities in the dummy fuel blocks
0.82 % (1)
0.3 % (1)
Worth of the CR insertion to the top of the

reflector (R2 group)
0.22 % (1)
0.32 % (2)
0.25 % (2)
non evaluated
non evaluated
Temporary neutron detector

(1)
(2)
Evaluated with the Transport Monte-Carlo calculation scheme

Evaluated with the Transport - Diffusion calculation scheme
The overall effect can exceed 1 % according to the core configuration. One
can note that the impact of the residual air in porosities and the impurities in graphite are
much more higher in the thin annular core configuration. This could explain the decrease of
the discrepancies between the experiment and the first calculation results, correlated to the
number of dummy fuel blocks (graphite) discharged during the criticality approach.
63
Diffusion calculation results

As far as the diffusion calculations are concerned, new developments carried out in
APOLLO2 and CRONOS2 take into account:
the exact position of the BP in the fuel block, by using new finite elements mesh in the
core model
the streaming effect, by generating anisotropic diffusion coefficients from both 2D-Pij
calculations and analytical formulation.
The use of the HTTR-FC2 data associated with a complete description of the axial
heterogeneity of the BP poison led to new core diffusion calculation results. This was done for
several energy structures in CRONOS2 without observing a main trend which would allow to
select a reference as energy mesh.
The final results are partially gathered in Figures 2.48 and 2.49. Figure 2.48 illustrates,
with 8 energy groups, the impact of the different model assumptions on the reactivity as a
function of the number of fuel columns loaded into the core. Figure 2.49 shows a streaming
effect ranging from 2.25 % in the 18 columns core configuration to 1.8 % in the full core
configuration. These results highlight also the importance of the used leakage model for
evaluating the neutron streaming in the control rods graphite blocks. Indeed, the first model
(TIBERE model) gave some values varying from 1.8 to 1.5 %.
FIG. 2.48. keff values obtained with different

core models.
FIG. 2.49. Neutron streaming and fuel block

homogenisation effect.
It is noticeable that the number of fuel columns needed to achieve criticality increases
by about 7 in comparison with the first results (Table 2-25) when considering the presence of
the detectors and the CR inserted in the upper reflector. However, at first criticality, a
discrepancy remains between the diffusion and the Monte-Carlo calculations (0.9 % < k/k <
1.7 %). This underscores the limits of a method based on a cross section homogenisation from
a fundamental mode calculation (infinite medium) that is barely pertinent for the 18 columns
core configuration. The actual environment (reflector blocks) should be considered and should
take place instead of the white boundary condition in the 2D APOLLO2 transport
calculations, before homogenising and collapsing locally the cross sections inside the fuel
element.
64
Final results of the HTTR-FC2 problem

All the final results are gathered in Table 2-30. In the case of the Monte Carlo code
TRIPOLI4, the discrepancy between the measurement and the calculation for the 18 fuel
columns configuration is reduced to k/k ~ 0.8 %, when considering the revised data of the
HTTR benchmark.
As far as diffusion calculations are concerned, the discrepancy is now reduced to
k/k ~ 2.7 or 1.7 % (depending on the number of energy groups), when taking into account
the improved models and the revised data. After all, it must be stressed that all the calculation
results obtained for the fully loaded core configuration fit the experiment, especially if one
consider the experimental uncertainties.
Table 2-30. Experimental, Monte-Carlo and Final Diffusion Results

1)
TRIPOLI4
3D Monte Carlo
172 gr & pointwise
1)
1)
CRONOS2
(Final model)
3D Diffusion
4 groups - 8 groups
EXPERIMENT
30 col.
1.13833 0.00090
1.1362 - 1.1451
1.1363 (> 3.6 %)
24 col.
1.1000 - 1.1096
1.0834 (> 2 %)
19 col.
1.02692 0.00043
1.0351 - 1.0432
1.0152 ?
18 col.
1.00855 0.00090
1.0178 - 1.0275
subcritical
detector impact included (k = 0.002) but not modelised
Excess reactivity HTTR-EX

All the excess reactivity are gathered below according to the previous values and the
definition of the excess reactivity given in [2-22].
Table 2-31. Excess Reactivity for Different Core Configurations
Excess Reactivity 1)
(keff 1)/keff [%]
Diffusion
Monte-Carlo
1)
Number of loaded fuel columns

18
19
24
30
1.7 < k/k < 2.7 3.4 < k/k < 4.1 9.1 < k/k < 9,9 12,0 < k/k < 12,7
+ 0.85
+ 2.6
Non evaluated
+ 12,15
detector impact included (k = 0.002) but not modelised
Control rod position for criticality: HTTR-CR

The results presented in Table 2-32 correspond to the fully loaded core configuration.
The control rod positions for obtaining criticality with all the inserted CR are: 178.7 cm with
the diffusion calculations and 177.9 cm with Monte-Carlo code.
65
Table 2-32: kEFF for Different CR Insertion Depths. Full Core

Control Rod Insertion
Depth [cm]
APOLLO2 TRIPOLI4 APOLLO2 CRONOS2
Exp.
178,9
1,00117 0,00024
1,00020
critical
177,6
0,99972 0,00038
0,99840
It is noteworthy that the CR insertion depth is well evaluated with the diffusion core
calculation despite of the discrepancies with the Monte Carlo results observed in this
configuration without the inserted CR. This remark underscores the fact that the CR worth is
overestimated with the diffusion method especially in this case where no equivalence factors
have been used in order to respect either the flux or the absorption rates between the multigroup transport calculations and the broad group diffusion core calculations.
Scram reactivity: HTTR-SC
In this section, the CR worth has been evaluated (Table 2-33) in the fully loaded core
configuration and according to:
kcrit . kControl Rods IN

kcrit . kControl Rods IN
As far as the CR insertion in the reflector is concerned, an unexplained result has been
observed with the Monte Carlo calculation, whilst a good agreement can be observed for the
overall CR worth inserted in the core. Once again, these results highlight the overestimation
of the absorbant in the diffusion calculation when one compares the CRONOS2 results to
those of TRIPOLI4.
Table 2-33. Control Rod Worth

TRIPOLI4 (Monte Carlo)
CRONOS2 (diffusion)
Exp.
CR inserted in the reflector

kcritique
1,00117 0,00024
1,00020
kRCR
0,92215 0,00040
0,90245
RCR [%]
8,56
10,83
12,0 1,2
All the CR inserted in the core
66
kcritique
1,00117 0,00024
1,00020
kRCR
0,68396 0,00030
0,63982
RCR [%]
46,32
56,31
46,0 4,6
Isothermal temperature coefficient in the core : HTTR-TC

The temperature coefficients have been evaluated from the following expression:
12
kT1 kT2
kT1 kT2
1
(T1 T2 )
They have been estimated only from keff of the core diffusion calculations performed at
different temperature (Table 2-34).
Table 2-34. KEFF as Function of the Temperature
Temperature [K]
300
340
380
420
460
480
APOLLO2 CRONOS2
1,00395
0,99724
0,99088
0,98455
0,97823
0,97525
According to the results gathered in Table 2-35, the temperature coefficients range
from - 15 to - 16 pcm/K between 300 to 420 K.
Table 2-35. Isothermal Temperature Coefficients
Temperature coefficients [pcm/K]
320
360
400
440
470
APOLLO2 CRONOS2
16,75
16,09
16,22
16,40
15,62
2.2.2.5 Concluding Remarks

The HTTRs core physics benchmarks have been treated with two different calculation
scheme: a transport-diffusion method and a transport-Monte Carlo one. These benchmarks
constitute the first opportunity in reactor physics to model and benchmark the codes and
methods in thin annular core geometry. This important point led to the limitations of the
classical two-steps core modeling based on a transport-diffusion chained calculations in order
to take accurately into account the core/reflector interface. Moreover, one of the other
characteristics of the HTTR core in its annular configuration was the presence of a large
number of uncommon big channels offering the possibility for the neutrons to leak from the
active zone (streaming effect). In addition to that, the important axial and radial
heterogeneities in the core (burnable poison, many different enrichments) make the HTTR a
real challenge in reactor physics.
Quite acceptable results are obtained in the fully loaded core configuration.
Besides, some discrepancies between the calculations and the experiment appear for
intermediate core configuration (thin annular core) close to criticality. Parts of these
67
discrepancies were reduced to k/k ~ 0.85 % with the Monte-Carlo calculations taking into
account the new benchmark data (air in porosity and impurities in graphite). As far as the
deterministic approach is concerned, the discrepancies were analysed and tackled by different
treatments. However, a difference of at least 1 % remains between the diffusion and the
Monte Carlo calculations in the annular core configuration when considering revised data.
2.2.2.6 Addendum
Several energy structure have been tested in course of this study for the core diffusion
calculations (see Table 2-36): the 2, 4, 8 and 20 group meshes, currently applied in the PWR
studies and the 6 and 13 groups which have already been used in the GT-MHR related
analyses.
Table 2-36. Group Structures for Core Diffusion Calculations
13 gr
6 gr
1
2
3
4
5
6
7
8
9
10
11
12
13
Limite infrieure en
nergie [Mev]
1,8315,10-1
1,0104,10-3
1,6745,10-5
4,1293,10-6
2,13,10-6
1,305,10-6
7,90,10-7
6,2501,10-7
3,91,10-7
3,145,10-7
1,15,10-7
5,9,10-10
1,1,10-10
2
3
4
5
6
The 6 and 13 group structures

8 gr
4 gr
1
2
3
4
5
6
7
8
2 gr
2
1
3
4
Limite infrieure en
nergie [Mev]
9,0718,10-1
2,7324,10-3
5,0045,10-3
2,7679,10-6
1,67,10-6
6,2501,10-7
1,6,10-7
1,1,10-10
The 2, 4 and 8 group structures

20 gr
1
2
3
4
5
6
7
8
9
10
Limite infrieure en
nergie [Mev]
4,493
2,231
1,353
4,979.10-1
1,832.10-1
6,738.10-2
2,479.10-2
9,119.10-3
2,035.10-3
4,540.10-4
20 gr
11
12
13
14
15
16
17
18
19
20
Limite infrieure en
nergie [Mev]
5,560.10-5
4,000.10-6
6,250.10-7
3,500.10-7
2,200.10-7
1,340.10-7
7,700.10-8
3,000.10-8
1,000.10-8
1,1.10-10
The 20 group structure

68
2.2.3. Germany [2-44]

2.2.3.1. Introduction
The calculations described in this report were performed within the framework of the
IAEA Co-ordinated Research Programme (CRP) on "Evaluation of the High Temperature Gas
Cooled Reactor Performance". The benchmark problems of the HTTR's start-up core physics
experiments have been defined by the Oarai Research Establishment, JAERI. The ISR of
Research Centre Jlich joined this CRP with the aim of testing its reactor code system. The
intention is to use the HTTR's start-up core experiments in order to find validation criteria for
the calculational methods hitherto used and the scope within which they can be applied. Pretest and post-test results have been obtained for the benchmark problems HTTR-FC and
HTTR-EX described in detail in [2-34]: the number of fuel columns has been evaluated for
the approach to first criticality when the fuel columns are charged from the core periphery,
and the excess reactivities for 18, 24, and 30 fuel columns have been determined.
2.2.3.2. Computational Methods and Nuclear Data
An overview of the code system used in the criticality calculations is given in Figure
2.50. All calculations performed are based on a 123-group cross section library which was
generated from the JEF-2.2 nuclear data files [2-35]. The NITAWL module of the AMPX-77
system [2-36] was used to calculate shielded cross sections in the resonance region where the
Nordheim integral treatment is employed. The double heterogeneity of the fuel due to the
coated particles (cp) and the fuel rod assembly was taken into account by Dancoff factors
which were calculated by the ZUT code [2-37] and supplied as input data for NITAWL. The
1-d cell calculations were performed by the TOTMOS code [2-38], a one-dimensional
transport-corrected integral spectrum code for eigenvalue calculations and the generation of
homogenized broad group constants. In order to evaluate the effect of the inhomogeneous
distribution of the burnable poison (BP) in the axial direction, additional cell calculations in
two-dimensional r-z geometry were performed by the discrete ordinates transport code DORT
[2-39]. The eigenvalues and flux distributions of the whole reactor were calculated by the
diffusion code CITATION [2-40] in 3-d triangular-z geometry. The OCTAGN module was
used to transform the cell-weighted broad group cross section data into a format such that
they can be used in the CITATION code. To take into account the effect of increased neutron
streaming in the coolant channels and in the large holes of the core and the reflector,
anisotropic diffusion coefficients were used. These were calculated by the MARCOPOLO
code on the basis of the multigroup integral transport theory [2-41]. In order to estimate the
influence of this effect the calculations for the whole reactor were performed with and without
streaming corrections.
The geometric dimensions of the fuel and core components and the atom number
densities of the materials given in Ref. [2-34] were used in the pre-test calculations. In the
post-test calculations residual air in the pores of the graphite, aluminium in the temporary
neutron detector holders and a revised boron impurity in some dummy fuel blocks were taken
into account [2-42].
Dancoff Factors
The NITAWL code considers the double heterogeneity of a system by Dancoff
factors. The Dancoff factor is defined as the probability that a neutron emitted from the
surface of the fuel region of the fuel element under consideration will have its next collision
69
in the fuel region of any other fuel element. Applying this definition to a fuel rod lattice filled
with fuel in the form of coated particles the ZUT code calculates the Dancoff factor as a sum
of the single rod Dancoff factor and the probability that a neutron leaving the first rod will
reach another fuel rod and be absorbed there. The Dancoff factors were calculated for all the
different types of compacts and are listed in Table 1 together with the corresponding U238
resonance integrals. The Dancoff factors remain nearly constant because they are only
dependent on the slightly varying geometric dimensions of the fuel compacts and rods, and on
the density of the graphite matrix. In pre-test and post-test prediction the Dancoff factors are
the same.
FIG. 2.50. Programme Structure of HTTR Calculations.

Table 2-37. Dancoff Factors and U238 Resonance Integrals for the Different Uranium
Enrichments
Enrichment
(wt.%)
Packing
Frac.
Vol. of
Fuel Comp.
(cm3)
Boron
Imp.
(ppm)
Dancoff
Factor
U238
Res.Int.
3.301
3.864
4.290
4.794
5.162
5.914
6.254
6.681
7.189
7.820
9.358
9.810
29.6
30.4
30.5
30.3
30.5
30.3
29.9
30.3
30.8
28.8
29.8
29.3
17.63
17.69
17.70
17.72
17.65
17.70
17.69
17.65
17.69
17.67
17.72
17.71
0.95
0.91
0.90
0.88
0.90
0.51
0.54
0.50
0.85
0.87
0.89
0.90
0.7225
0.7283
0.7284
0.7296
0.7300
0.7280
0.7270
0.7280
0.7258
0.7214
0.7223
0.7245
43.82
44.12
44.17
44.20
44.36
44.67
44.85
44.91
45.09
44.97
45.91
45.89
70
2.2.3.3. Pre-Test Calculations

1-d Cell Calculations
For the 15 types of BP-fuel combinations given in Appendix B of [2-34], onedimensional cell calculations were performed by the TOTMOS code using the following
scheme:
1. 123 cell-weighted group constants of the cp-cell-model were calculated in spherical
geometry using a 123 group cross section library. The cell model consists of 3 zones:
the kernel, the coatings, and the corresponding matrix zone. A white boundary
condition was used at the outer surface of the cell.
2. Subsequently, 123 cell-weighted group constants of the fuel rod cell were calculated
in cylindrical geometry and with the same group structure as used for the cp cell. The
zones of the cylindrical fuel rod cell were: the central hole, the fuel compact, the
graphite sleeve, the coolant channel, and the corresponding graphite block. The fuel
rod cell is shown in Figure 2.51. The group constants of the materials in the fuel zone
were the cell-weighted cross sections resulting from the cp cell calculation. The cross
section of the fuel cell was equivalent to the cross section of the fuel block divided by
the number of fuel rods. A white boundary condition was used at the outer surface of
the cell.
3. In a subsequent third cell calculation, the cylindrical cell model shown in Figure 2.52
consists of a BP rod surrounded by a second zone representing the remaining fuel
block. The group constants of the materials in this second zone were the cell-averaged
123 group constants resulting from the fuel rod cell calculation. The cross section of
the BP cell was the same as the cross section of the fuel block divided by the number
of BP rods. Again a white boundary condition was used at the outer cell surface. In
this last step of the cell calculations, cell-weighted condensation was performed to
four broad energy groups used in the calculations of the whole reactor. The energy
group structure of the four groups is given in Table 2-38.
The k-values of the cp-, fuel- and BP-cell calculations are given in Table 2-39 for all
15 types of BP-fuel combinations.
FIG. 2.51. Pre-Test 1-d Cylindrical Model of the Fuel Rod Cell for TOTMOS.
71
FIG. 2.52. Pre-Test 1-d Cylindrical Model of the BP Cell for TOTMOS
Table 2-38. Group Structure in the Diffusion Calculation
4 Group Set (Pre-Test)
72
26 Group Set (Post-Test)
Group
Upper Energy Boundaries

(eV)
1.492107
1
2
3
4
1.111105
5
6
7
8
9
10
11
12
29.0
13
14
15
16
17
1.492107
7.408106
3.679106
6.721105
1.111105
1.931104
3.355103
1.585103
7.485102
2.754102
1.301102
6.144101
29.0
13.7
8.32
5.04
2.38
1.86
18
19
20
21
22
23
24
25
26
1.29
0.65
0.35
0.20
0.12
0.08
0.05
0.02
0.01
Group
Upper Energy
Boundaries
(eV)
Table 2-39. Pre-Test Results of the TOTMOS Cell Calculations; No BP Adjustment

ID.
No.
Enr.
(wt.%)
k-Values of the
Cp Cell
Fuel Cell
BP Cell
343320
393320
673320
793320
3.4
3.9
6.7
7.9
0.6282
0.6562
0.7782
0.8208
1.4285
1.4604
1.5457
1.5610
1.1309
1.1819
1.3355
1.3716
433120
483120
943120
993120
4.3
4.8
9.4
9.9
0.6771
0.6995
0.8713
0.8854
1.4957
1.5142
1.5996
1.6021
1.2167
1.2514
1.4161
1.4254
433325
523325
633325
4.3
5.2
6.3
0.6771
0.7163
0.7619
1.4790
1.5095
1.5376
1.1887
1.2437
1.2955
593125
633125
723125
793125
5.9
6.3
7.2
7.9
0.7484
0.7619
0.7969
0.8208
1.5476
1.5559
1.5726
1.5802
1.2855
1.3002
1.3351
1.3565
2-d cell calculations and BP adjustment

In the 1-d TOTMOS cell calculations, the axial heterogeneity of the BP is not
explicitly taken into account. Thus, the efficiency of the BP is overestimated since the axial
self-shielding of the BP cannot be considered. In order to find out the influence of the axial
heterogeneity of the BP, 2-d cell calculations were performed by the discrete ordinates
transport code DORT. For this purpose two series of 2-d cell calculations were performed:
1. In a first series of DORT calculations, the BP region is homogenized as in the 1-d cell.
The results are given in Table 2-40 and compared with the corresponding results
obtained by the TOTMOS code. It is seen that there is quite good agreement between
the two computational methods.
2. In a second series of calculations the axial heterogeneity of the BP is explicitly
considered in the 2-d DORT calculations. The geometric model is shown in Figure
2.53 and the results are presented in Table 2-40. It can be seen that the heterogeneity
has quite a large influence on the infinite multiplication constant of the BP cell and
considerably reduces the efficiency of the BP in comparison to the 1-d cell
calculations.
73
In order to transfer this information into the 1-d cell calculation the B10 concentration
was reduced in such a way that the resulting k was the same as that obtained in the two
series of DORT calculations. In all BP-fuel combinations, the B10 concentration had to be
reduced by 22% to 30 %.
Table 2-40. Pre-Test Infinite Multiplication Factors of the BP Cell Obtained by Different
Methods
k-Values of the BP Cell
Case
TOTMOS
Homog.
DORT
Homog.
k
(TOTM.-DORT)
DORT
Heterog.
k
(DORTHet.DORTHom.)
343320
393320
673320
793320
1.1309
1.1819
1.3355
1.3716
1.1347
1.1855
1.3376
1.3737
0.0038
0.0036
0.0023
0.0021
1.1741
1.2225
1.3654
1.3981
0.0394
0.0370
0.0275
0.0244
433120
483120
943120
993120
1.2167
1.2514
1.4161
1.4254
1.2204
1.2548
1.4181
1.4274
0.0037
0.0034
0.0020
0.0019
1.2575
1.2896
1.4417
1.4499
0.0371
0.0348
0.0236
0.0225
433325
523325
633325
1.1887
1.2437
1.2955
1.1929
1.2475
1.2988
0.0041
0.0037
0.0033
1.2343
1.2852
1.3329
0.0414
0.0378
0.0342
593125
633125
723125
793125
1.2855
1.3002
1.3351
1.3565
1.2892
1.3037
1.3383
1.3594
0.0037
0.0036
0.0032
0.0029
1.3264
1.3401
1.3719
1.3908
0.0373
0.0363
0.0335
0.0313
74
FIG. 2.53. Pre-Test 2-d Cylindrical Model of the BP Cell for DORT.
Streaming correction
The presence of the insertion holes in the control rod guide blocks and of the coolant
channels in the fuel and reflector blocks leads to an increased neutron streaming in the axial
direction. A possibility of treating this effect within the diffusion theory is the use of
anisotropic diffusion coefficients. These are determined by the MARCOPOLO code which
calculates anisotropic multigroup diffusion coefficients from the leakages of a heterogeneous
lattice cell. For all types of blocks with insertion holes or coolant channels, 1-d cylindrical
cells are defined which, in most situations, consist of a central hole representative of the block
and surrounded by an associated material zone. As typical examples, the cell models of the
dummy fuel block MB-2 and of the control rod guide block CB-1 are shown in Figures 2.54
and 2.55. The MARCOPOLO code allows the calculation of the buckling-independent
diffusion constants Dgk (k=r,z) in these cells taking into account linear anisotropic scattering.
The group constants used in the MARCOPOLO calculations were provided by the TOTMOS
code. For the control rod guide blocks CB-1 and CB-2 and the reflector block RB-1, the ratio
of the anisotropic diffusion coefficients to the homogeneous diffusion coefficients are given
in Table 2-41. The resulting values for all block types of interest are input into the
CITATION code in the form of these streaming correction factors.
75
FIG. 2.54. 1-d Cylindrical Model of the Simplified Dummy Fuel Block B (MB-2).
FIG. 2.55. 1-d Cylindrical Model of the Control Rod Guide Block CB-1.
Table 2-41. Pre-Test Streaming Correction Factors Obtained by the MARCOPOLO Code
Streaming Correction Factors
Group
1
2
3
4
76
CR Guide Block
CB-1
Dr/Dhom
Dz/Dhom
CR Guide Block
CB-3
Dr/Dhom
Dz/Dhom
1.1403
1.1761
1.1812
1.1877
1.1562
1.1963
1.2016
1.2090
1.5740
1.9333
1.9497
2.0243
1.6317
2.0358
2.0537
2.1369
Repl.Refl.Block
RB-1
Dr/Dhom
Dz/Dhom
1.0199
1.0246
1.0286
1.0307
1.0513
1.0786
1.0836
1.0908
Whole reactor calculations

Using the 4-group cross sections from the NITAWL-TOTMOS cell calculations the
whole HTTR reactor was modelled with the CITATION diffusion code. A 3-dimensional
triangular-z model was chosen. Each block was divided horizontally into 6 meshes and
vertically into 4 meshes. The assembly was modelled by dividing the volume into spectral
zones related to the material compositions. There are 45 different material zones.
Six pairs of control rods in the side reflector of the HTTR cannot be fully withdrawn
to the top of the reflector. The effect of this CR insertion on reactivity is given as k= 0.004
in Ref. [2-34] and is subtracted from the calculated keff-values. Fuel columns are loaded
clockwise from the periphery to the centre according to the loading order given in Ref. [2-34].
A thin annular core is formed at 18 fuel column loading, and a thick annular core is achieved
when 24 fuel columns are loaded. A full core contains 30 fuel columns. According to this
loading scheme four series of diffusion calculations were performed for 9 up to 30 fuel
columns in the core:
without streaming correction of the diffusion constant and without BP adjustment,

with streaming correction, but without BP adjustment,
with BP adjustment, but without streaming correction,
with both corrections.
The keff-values of these four series are shown in Figures 2.56 and 2.57, respectively.
As can be seen, the influence of the streaming correction is nearly independent of the BP
adjustment. With and without BP adjustment the streaming correction causes a difference in
keff from k= 0.02 at a loading with 9 fuel columns down to k=0.015 for the case of a fully
loaded core. This decrease in k can be explained by the fact that dummy fuel blocks with
large holes and a great neutron streaming effect are subsequently replaced by fuel blocks
which exhibit nearly no streaming effect. Moreover, it is found that the neutron streaming in
the coolant channels of the top and bottom replaceable reflector can be neglected, because the
decrease in keff caused by this effect was only k= 0.08 %.
On the other hand, the effective multiplication constants are increased by the BP
adjustment: the "boron adjusted" keff-values are greater than the uncorrected ones, and the
difference in keff increases with the increasing number of fuel rods. But it is evident that the
increase of the effective multiplication constant due to the BP adjustment is not compensated
by the effect of neutron streaming.
When taking into account the neutron streaming in the channels and holes of core and
reflector, the neutron shielding in the BP rods, and when the reactivity of the CR insertion is
subtracted, the first criticality will be achieved at 16 fuel columns loading. The excess
reactivity amounts to k/k=0.42 %. The excess reactivity of the thin, thick, and the fully
loaded core is 2.48 %, 10.27 %, and 13.85 %, respectively. All results are shown in Tables 242 and 2-43.
These results agree very well with the average results of the deterministic calculations
of all CRP-5 participants presented at the 1st Research Coordination Meeting [2-22].
77
Table 2-42. Pre-Test Effective Multiplication Factor and Excess Reactivity at the First
Criticality
No. of
Fuel Columns
keff
[%k/k]
15
16
0.9991
1.0042
-0.90
+0.42
FIG. 2.56. Pre-Test Effective Multiplication Factor as a Function of the Number of Fuel
Columns.
Table 2-43. Pre-Test Effective Multiplication Factor and Excess Reactivity at 18, 24, and 30
Fuel Columns Loading
78
No. of
Fuel Columns
keff
[%k/k]
18
24
30
1.0254
1.1145
1.1607
2.48
10.27
13.85
FIG. 2.57. Pre-Test Effective Multiplication Factor as a Function of the Number of Fuel
Columns.
Discussion of the pre-test results
In the pre-test diffusion calculation the first criticality of the HTTR was achieved with
16 fuel columns in the case of fuel columns loading from the core periphery. This corresponds
to a critical mass of 29.6 kg U235. It turned out that the burnable poison in the fuel blocks and
the neutron streaming in the holes of the CR guide columns had a great influence on the
neutronic characteristics of the HTTR and that an accurate treatment of these effects is
required.
The experiment showed that the HTTR got critical with 19 fuel columns with an
excess reactivity of about k/k=1.5%. The discrepancy between measurement and the FZJ
pre-test calculation amounted to k=0.0287 at the first criticality for a 4 group diffusion
calculation with no leakage iteration; at 30 fuel columns in the core this difference was with
k=0.0261 of the same order as can be seen on Table 2-44. Only at 27 and 30 fuel columns
loading the calculated excess reactivities were in the experimental error intervals, showing
some error compensating effects in the calculations. Six possible reasons for these
discrepancies are given in the following:
a non-adequate modelling of the fuel and BP unit cells,

the neglect of the detailed structure of the HTTR fuel block in the whole core
calculations,
the use of few group homogenized cross sections in the whole core diffusion
calculation without leakage iteration,
79
the consideration of another than the actual boron impurity in some dummy fuel
blocks and of helium instead of air in the graphite pores,
a not yet adequate treatment of the self-shielding in the BP rods,
and an underestimation of the neutron streaming.
Table 2-44. Results of the Pre-Test Diffusion Calculations together with the Experimental
Results
No. of
Fuel
Columns
keff
Calc.
keff
Exper.
k
Calc.-Exp.
[% k/k]
Calc.
[% k/k]
Exper.
9
12
15
16
17
18
0.9596
0.9827
0.9991
1.0042
1.0129
1.0254
0.9282
0.9481
0.9652
0.9701
0.9785
0.9913
0.0314
0.0346
0.0339
0.0341
0.0344
0.0341
-4.21
-1.76
-0.09
0.42
1.27
2.48
-7.7a
-5.5a
-3.6a
-3.1a
-2.2a
-1.0a
19
21
24
1.0439
1.0731
1.1145
1.0152
1.0417
1.0834
0.0287
0.0314
0.0311
4.21
6.81
10.27
1.5
4.01.1b
27
30
1.1469
1.1607
1.1198
1.1346
0.0271
0.0261
12.81
13.85
7.72.1b
10.73.0b
12.03.3b
(CR Insertion considered k=0.004)

from 1/M Measurement [2-43]
b
from IK Method
a
2.2.3.4. Post-Test Calculations

Assessment of differences
The first four possible reasons of the differencies between pre-test prediction and
experiment were analysed in more detail using simplified (1-d and 2-d) models of a thin
annular core with 18 fuel columns:
80
improved modelling of the fuel and burnable poison (BP) unit cells,
consideration of the exact position of the BP rods in the fuel blocks,
use of many group homogenized cross sections or accurate leakage iteration in the
whole core diffusion calculation,
and consideration of the revised boron impurity in some dummy fuel blocks and of
residual air in the pores of the graphite, as proposed by JAERI [2-42].
In the course of these studies it turned out that the change in the unit cell models, the
increase of the boron impurity in some dummy fuel blocks and the replacement of the helium
in the pores of the graphite blocks by air reduced the multiplication constant of the simplified
core models. However, this reduction was more than compensated by the increase of keff in a
26 groups diffusion calculation or a 4 groups diffusion calculation with leakage feedback used
to describe the core/reflector coupling accurately.
In all pre-test calculations, the HTTR fuel blocks were homogenized with only six
triangular meshes per fuel block and the exact position of the BP rods in the fuel blocks as
mentioned in the 2nd item was not considered. When taking into account the detailed BP
positions in the whole core diffusion calculations with 24 horizontal meshes per fuel block the
multiplication constant decreased significantly, as can be seen on Table 2-45. Furthermore, it
can be seen that the increase of the group number from 26 up to 52 in the whole core
diffusion calculation has only a small influence on the keff-values. Thus, a group structure of
26 energy groups seems to be sufficient to get reliable results. The 26 energy group structures
is given in Table 2-38.
Table 2-45. Effective Multiplication Constants and k Values for a Simplified 2-d, Triangular
HTTR Core Model (18 Fuel Columns) Using Different Unit Cell Models and Group
Structures (No Streaming Correction and BP Adjustment; Helium in Graphite Pores and Old
Boron Impurity)
keff
Old Unit Cells

24 Mesh Hom. Model
New Unit Cells

24 Mesh Het. Model
New-Old Model
4 Groups
1.0484
0.9857
-0.0637
26 Groups
1.0840
1.0419
-0.0421
k
26-4 Groups
+0.0355
+0.0563
52 Groups
1.0842
1.0430
-0.0412
k
52-26 Groups
+0.0004
+0.0011
81
Concerning item 5 of the possible reasons for the discrepancies, the axial selfshielding in the BP rod was considered by a B10 adjustment in the 1-d cell calculations of the
pre-test prediction. Now, an improved treatment of the axially heterogeneous distribution of
the BP was performed by 2-d calculations.
As to item 6 of the possible reasons, the comparison with the Japanese Monte Carlo
calculations shows, that the neutron streaming has been underestimated in the previous
calculations. Therefore, modified diffusion constants for treating this effect were determined
on the basis of the Japanese results.
Heterogeneity of the HTTR fuel block
In order to consider the exact position of the BP rods in the whole core diffusion
calculation 24 horizontal meshes per fuel block have to be chosen and each fuel block has to
be divided into three different regions:
one fuel region, corresponding to 18/24 of the whole block,

two BP regions, each corresponding to 2/24 of the whole block,
one "empty BP" region, according to 2/24 of the whole block.
FIG. 2.58. HTTR Fuel Block and Post-Test Model Used in the Diffusion Calculation with 24
Horizontal Meshes per Block.
82
The fuel block with 33 fuel pins is shown in Figure 2.58 together with the modelling
of the different regions in the whole core diffusion calculation with 24 horizontal meshes per
block. Basing on this subdivision of the fuel block it is necessary to create new unit cell
models adapted to this situation.
New 1-d fuel cell model
The new fuel cell presented in Fig 2.59 together with the basic geometry has an outer
radius of 2.70 cm corresponding to the pitch of the fuel rod lattice in the fuel block.
Explicitely modelled are: the inner helium channel, the compact, the sleeve, and the outer
helium gap. The cell-weighted 26 group cross sections generated on the basis of this new fuel
cell model are used in the fuel region of each fuel block in the whole core diffusion
calculation.
p = 5.15
Dimension of the Fuel Lattice Used as Basis

in the New Fuel Cell Model
FIG. 2.59. Post-Test 1-d Cylinder Model of the Fuel Cell.
New 2-d BP Cell Model
In the pre-test prediction, the axial self-shielding of the BP was taken into account by a
reduction of the B10 concentration in the 1-d TOTMOS cell calculations. Now, in the post-test
calculations the axial self-shielding was considered by another method: it was provided that
the absorption rate in the BP-regions of the whole core diffusion calculation is the same as in
the detailed 2-d DORT-cell calculation for these regions.
83
Thus, the axial self-shielding of the BP was taken into account by a more accurate
method compared to the method of reducing the B10 concentration. This reduction
overestimated the efficiency of the BP compared to the use of group constants obtained in a 2d DORT-cell calculation by about k0.0065-0.0077.
The revised 2-d cell model used in the DORT calculations is shown in Fig 2.60. This
BP unit cell consists of three radial zones: the BP rod, surrounded by a graphite zone,
corresponding to the area of 2/24 block, and a third radial zone representing the remaining
fuel block. The group constants in this third radial zone are the cell-averaged group constants
resulting from the fuel cell calculation. The first radial zone, the BP rod, is divided axially
into five zones, representing the BP pellets, the graphite disks, and the upper and lower plugs
of the graphite block. In the 2-d cell calculation the group constants are homogenized over the
two inner radial zones and over the five axial zones of the BP rod and condensed to 26 energy
groups. These 26 zone-weighted group constants obtained by the 2-d DORT cell calculation
are used in the two BP-regions of each fuel block in the whole core diffusion computation.
The cross section of this BP cell corresponds to the cross section of the fuel block
divided by the number of BP rods.
FIG. 2.60. Post-Test 2-d Cyl. Model of the BP Cell for DORT
84
New graphite cell model

The 26 group constants used in the third region of the HTTR fuel block are the zoneweighted cross sections supplied by the 1-d "empty" BP unit cell calculation. This "empty
BP" cell has the same area as the cross section of the fuel block. It consists of three zones: the
inner hole filled with helium instead of BP, surrounded by a graphite zone, corresponding to
the area of 2/24 block, and a third zone, in which the cross sections of the homogenized fuel
cell and of the homogenized replaceable reflector cell RB-2 are taken into account, because
all empty BP holes in the 18 fuel columns assembly are situated at the outer surface of the
annular core. Thus, the neutron spectrum in this "empty BP" unit cell is determined by the
outer reflector and the corresponding fuel block.
In the "empty" BP cell calculation, the group constants are homogenized over the two
inner zones and condensed to 26 energy groups. The k-values obtained by the three different
cell calculations are given on Table 2-46.
Table 2-46. Post-Test Results of the three Different Cell Calculations
Case
Enr.
(wt.%)
k-Values of the
New
Fuel Cell
1-d TOTMOS
New
BP Cell
2-d DORT
Empty
BP Cell
1-d TOTMOS
343320
393320
673320
793320
3.4
3.9
6.7
7.9
1.2808
1.3075
1.3814
1.3943
1.1598
1.2068
1.3464
1.3782
1.4955
1.5419
1.6900
1.6983
433120
483120
943120
993120
4.3
4.8
9.4
9.9
1.3230
1.3374
1.4140
1.4157
1.2415
1.2728
1.4216
1.4297
1.5693
1.5987
1.7288
1.7356
433325
523325
633325
4.3
5.2
6.3
1.3230
1.3490
1.3742
1.2181
1.2677
1.3143
1.5707
1.6164
1.6581
593125
633125
723125
793125
5.9
6.3
7.2
7.9
1.3665
1.3742
1.3893
1.3943
1.3083
1.3217
1.3529
1.3712
1.6477
1.6594
1.6808
1.7011
Modified streaming correction

In the course of the pre-test calculations, it turned out: the streaming effect calculated
by the CITATION code using anisotropic diffusion coefficients determined by the
MARCOPOLO code [2-41] was about 33% smaller than the effect calculated by the Monte
85
Carlo code MVP of JAERI [2-22], in the case of 18 fuel columns in the core. In the case of
the fully loaded core, the streaming effect was smaller and the difference between the
diffusion and Monte Carlo calculation was reduced to about 15%, as can be seen on Table 247.
Table 2-47. Streaming Effects in Pre-Test and Post-Test Calculations
Streaming Effect [% k/k]
No. of
Fuel
Col.
Monte Carlo
Code MVPa
for:
Pre-Test
Diffusion Code
with
Stream. Corr.
for:
Post-Test
Diffusion
Code
with
Modified
Stream.
Corr.
for:
all Coolant Channels and

Holes
of Core and Reflector
CR Guide Columns
Irrad. Columns
18
24
30
a
2.30
1.30
1.53
1.27
1.10
Pre-Test
Post-Test
Diffusion Code
Diffusion
with
Code with
Stream. Corr.
Modified
for:
Stream. Corr.
for:
2.32
1.92
1.68
1.69
1.38
1.18
2.46
2.02
1.75
Results of JAERI-HTTR, presented at the 1st RCM of the CRP-5.
One possibility of getting more accurate diffusion constants is to compare the

corresponding streaming effects obtained by diffusion and by Monte Carlo calculations and to
adapt the anisotropic diffusion coefficients to the results of the Monte Carlo calculations.
Therefore, the neutron streaming coefficients calculated with the MARCOPOLO code
had to be modified. When increasing the neutron streaming correction factors of the CR-guide
and irradiation columns by about 20% (in r- and in z-direction) the streaming effect calculated
by the diffusion code was the same as the reactivity effect resulting from the Monte Carlo
calculation, as can be seen on Table 2-47.
For the control rod guide blocks CB-1 and CB-2, and the irradiation block IB-1 the
ratios of the anisotropic diffusion coefficients to the homogeneous diffusion coefficients are
given in Table 2-48 together with the modified streaming correction factors. They are listed in
the 4 group structure. But in the core calculation 26 group constants were used, which were
interpolated from the 4 group values.
Thus, in the recalculation of the first criticality and of the excess reactivities, these
modified anisotropic diffusion coefficients for all CR-guide and irradiation columns were
used. The anisotropic diffusion constants of the upper and lower replacable reflector, of the
fuel and the dummy fuel blocks remained unchanged because there was no comparable Monte
Carlo calculation.
86
Table 2-48. Streaming Correction Factors Obtained by the MARCOPOLO Code together
with the Modified Factors Deduced from MVP Monte Carlo Calculations
Group
CR Guide Block
CB-1
Dr/Dhom
Dz/Dhom
CR Guide Block
CB-3
Dr/Dhom
Dz/Dhom
Irrad. Block
IB-1
Dr/Dhom
Dz/Dhom
Streaming Correction Factors (Pre-Test)

1
2
3
4
1.140
1.176
1.181
1.188
1.574
1.933
1.950
2.024
1.156
1.196
1.202
1.209
1.632
2.036
2.054
2.137
1.139
1.175
1.180
1.186
1.567
1.923
1.939
2.013
Modified Streaming Correction Factors (Post-Test)

1.368
1.411
1.417
1.424
1
2
3
4
1.889
2.320
2.340
2.429
1.387
1.435
1.442
1.451
1.958
2.443
2.465
2.564
1.367
1.410
1.416
1.423
1.880
2.308
2.327
2.416
Table 2-49. Streaming Effects in the Pre-Test and Post-Test Detailed Core Calculations
No. of
Fuel
Columns
Pre-Test
Benoist
Stream.Correction
keff
Post-Test
Modified
Stream.Correction
keff
9
12
15
16
17
18
-0.0203
-0.0194
-0.0187
-0.0185
-0.0181
-0.0176
-0.0301
-0.0284
-0.0275
-0.0264
-0.0257
-0.0253
19
24
-0.0175
-0.0166
-0.0249
-0.0245
27
30
-0.0159
-0.0153
-0.0237
-0.0229
87
The use of the modified anisotropic diffusion constants decreased the keff-values in the
whole core diffusion calculations by about k0.01-0.007 compared to the pre-test results, as
can be seen on the Table 2-49.
Whole reactor calculations
In the post-test calculations the whole core was modelled with the CITATION
diffusion code using the 26 group cross sections from the cell calculations. A 3-dimensional
triangular-z model was chosen, as in the pre-test-prediction, but each block was divided
vertically into 4 meshes and horizontally into 24 meshes, in order to arrange the BP rods and
the empty BP-hole at their exact positions in the fuel block.
As in the pre-test calculations, four series of diffusion calculations were performed in
order to show the influence of the modified streaming correction and of the axial BP
heterogeneity on the keff-values. Both effects are augmented compared to the pre-test
prediction, as can be noticed on Tables 2-49 and 2-50. The modified streaming correction
reduces the multiplication constants by about k=0.03 at 9 fuel columns loading to k=0.023
at 30 fuel columns in the core. On the other hand, the more accurate treatment of the axial
heterogeneity of the BP increases the multiplication factor by about k=0.024 at 9 fuel
columns in the core and by about k=0.034 at fully loaded core. This stronger effect of the
BP when using the more accurate treatment of the axial heterogeneity is mainly due to the
fact, that the abs-values obtained by the 2-d cell calculation are significantly smaller in the
thermal energy range than those macroscopic absorption cross sections obtained by the 1-d
cell calculation with B10 reduction, as can be seen from Fig 2.61.
FIG. 2.61. Comparison of Macroscopic Absorption Cross Sections Obtained by Different

Methods.
88
Table 2-50. Axial Self-Shielding Effect of the BP in the Detailed Core Calculations
Obtained by Different Methods
No. of
Fuel
Columns
Pre-Test
B -Reduction
keff
Post-Test
2-d BP Cell Calculation
keff
9
12
15
16
17
18
+0.0189
+0.0196
+0.0200
+0.0201
+0.0203
+0.0207
+0.0245
+0.0255
+0.0257
+0.0265
+0.0269
+0.0273
19
24
+0.0213
+0.0242
+0.0281
+0.0314
27
30
+0.0252
+0.0263
+0.0333
+0.0336
10
Table 2-51. Results of the Post-Test Diffusion Calculations together with the Experimental
Results
No. of
Fuel
Columns
keff
Calc.
keff
Exper.
k
Calc.-Exp.
[% k/k]
Calc.
[% k/k]
Exper.
9
12
15
16
17
18
0.9408
0.9642
0.9811
0.9866
0.9959
1.0080
0.9282
0.9481
0.9652
0.9701
0.9785
0.9913
0.0126
0.0161
0.0159
0.0165
0.0174
0.0167
-6.30
-3.70
-1.90
-1.40
-0.41
0.79
-7.7a
-5.5a
-3.6a
-3.1a
-2.2a
-1.0a
19
21
24
1.0263
1.0556
1.0944
1.0152
1.0417
1.0834
0.0111
0.0139
0.0110
2.60
5.30
8.60
1.5
4.01.1b
27
30
1.1261
1.1336
1.1198
1.1346
0.0063
0.0010
11.20
11.80
7.72.1b
10.73.0b
12.03.3b
(CR Insertion and Al in the Neutron Detector Holders considered: k=0.004+0.002)

from 1/M Measurement [2-43]
b
from IK Method
a
89
Table 2-52. Post-Test Effective Multiplication Factor and Excess Reactivity at the First
Criticality
No. of
Fuel Columns
keff
[%k/k]
17
18
0.9959
1.0080
-0.41
+0.79
Table 2-53. Post-Test Effective Multiplication Factor and Excess Reactivity at 18, 24, and 30
Fuel Columns Loading
No. of
Fuel Columns
keff
[%k/k]
18
24
30
1.0080
1.0944
1.1336
0.79
8.60
11.80
FIG. 2.62. New and Old k eff-Values of the Diffusion Calculations in Comparison with the
Experimental Results.
90
When taking into account the modified neutron streaming in the big holes of the core,
the neutron axial self-shielding in the BP rods by the more accurate method, and when not
only the reactivity of the CR insertion of k=0.0040, but also the reactivity of the aluminium
in the temporary neutron detector holders of k=0.0020 is subtracted, the first criticality will
now be achieved at 18 fuel columns loading. The excess reactivity amounts to k/k=0.79%.
The excess reactivity of the thin, thick, and the fully loaded core is 0.79%, 8.63%, and
11.80%, respectively. The results of the post-test calculations are given on Tables 2-51 to 253. In Figure 2.62 the new and the old keff-values of the diffusion calculations are presented
together with the experimental results.
Discussion of the post-test results
When summing up all post-test studies, the intermediate calculations using simplified
whole core models and the final calculations for the detailed core, the analysis yields the
following effects at 18/19 fuel columns in the core compared to the pre-test results:
when considering the detailed structure of the HTTR fuel block in the whole core
calculation, the multiplication factor decreases by about k 0.043,
the description of the detailed energy-dependence of the neutron flux adequately by a
fine energy group structure increases the keff-value by about k 0.035,
when taking into account the residual air in the graphite pores, the higher boron
impurity in some dummy fuel blocks, and the reactivity effect of the aluminium in the
neutron detector holders, the multiplication factor is reduced by about k 0.009,
when considering the axial heterogeneity of the BP by 2-d cell calculations, keff of the
whole core calculation increases by about k 0.0068,
and when treating the neutron streaming effect by modified diffusion constants on the
basis of the Japanese Monte Carlo results, keff is reduced by about k 0.0075.
The number of fuel columns to achieve first criticality increases by about 2 fuel
columns in comparison to the pre-test results presented at the 1.RCM [2-22]. Now, the
discrepancy between measurement and FZJ diffusion calculation amounts to k=0.0111 at the
first criticality (i.e. at 19 fuel columns), as can be seen on Table 2-51.
2.2.3.5. Conclusion
Altogether it turns out that the following procedures seem to be necessary for a better
approach to the experimental results:
detailed heterogeneity of the BP- and fuel-region in the whole core calculation,
use of fine group constants or of broad group constants including detailed leakage
information in the diffusion calculations of the whole core in order to describe the
core/reflector coupling accurately,
treatment of the axial self-shielding in the BP rods by 2-d cell calculations,
consideration of an enhanced neutron streaming, brought about here by an adaption of
the diffusion constants to results of Monte Carlo calculations.
When applying these improvements and regarding the actual boron impurity in some
dummy fuel blocks together with air in the graphite pores the first criticality was recalculated
for 18 fuel columns in the case of fuel loading from the core periphery. This corresponds to a
91
critical mass of 33.1 kg U235. Moreover, it turned out that a 26 energy group structure in the
whole core diffusion calculation seems already to be sufficient to get reliable results when no
detailed leakage information is used in generating the group constants for the whole core
diffusion calculations.
2.2.3.6. Acknowledgement
The authors would like to thank Dr. W. Scherer for very useful discussions and
comments, and Ms. A. Kuhr for supporting all the plots.
92
2.2.4. Indonesia [2-45]

Initial fuel loading and first criticality of the HTTR was accomplished in an annular
core configuration. This core type was chosen because of high inherent safety characteristics
for loss of coolant accidents [2-46]. The vessel cooling system (VCS) surrounding the reactor
pressure vessel is designed to maintain the fuel temperature at less than the limit of 1600C in
a loss of coolant accident. The decay heat will be transferred radially through the regions of
fuel, side removable and permanent reflector blocks and reactor pressure vessel to the cooling
panel of the VCS through heat conduction, radiation and convection without any necessity for
active cooling. Introduction of the annular core type is expected to enhance the heat removal
mechanism, because the heat transfer pass is shortened by thinning of the fuel region.
In the fuel loading for the start up core physics experiment, three different types of
core were considered; a thin and thick annular core with 18 and 24 fuel columns, respectively,
and a fully loaded core that consists of 30 fuel columns. The core size of the HTTR is about
half the size of a commercial HTGR. The high excess reactivity in HTTR is similar to that of
HTGR, which is needed to compensate for the effects of temperature, xenon, burn up, etc.
during power operations.
Data for the core physics calculation is applied by JAERI [2-47]. The data include
specifications for core and its internal components, such as fuel, replaceable reflectors,
permanent reflectors, control rods and dummy fuel elements. Also provided are HTTR details
(see Section 2.1.1) including specifications of the fuel rods, fuel compacts and coated fuel
particles. Generation of cell cross-sections in this study was performed using CELL Module
of the SRAC-EWS Code system [2-48], while the whole core calculation was performed
using CITATION Module of the SRAC-EWS.
The purpose of this study was to examine when the first criticality can be achieved by
using annular thin core loading mechanism. In this loading type, fuel columns are loaded in
the outermost fuel region (18 columns, i.e. annular thin core). Upon completion of the initial
18 columns, fuel loading continued with another 6 fuel columns in the inner fuel region (24
columns, i.e. thick annular core). Finally, loading continued with another 6 fuel columns in
the innermost fuel region, so that the full core of 30 fuel columns was achieved.
2.2.4.1. Calculation Model and Procedures
Calculation for the start up benchmark test was developed and all calculations were
performed for a core temperature of 300K and helium pressure of 1 atm. Figure 2.63 depicts
the flow diagram of the benchmark calculation.
The SRAC-EWS Code system [2-48] was applied throughout the calculations, and the
nuclear data in 107 groups were obtained mainly from JENDL-2 and ENDF/B-V. The
ULIBMAKE module of SRAC-EWS was applied to generate the user library files from the
public library, and to collapse these files into 54 energy groups. The CELL module was then
applied to further condense these into 6 groups (i.e. 3 groups in the fast region and 3 groups in
the thermal region), and to perform the cell calculation in collision probability method. In
order to save computing time, the cell calculations were performed by using the equivalent
cylindrical geometry (i.e. Wigner-Seitzs approximation) with white boundary conditions,
rather than the more exact hexagonal geometry, because the effect on k was minimal, as
pointed out by Jeong et al.[2-49]. The CITATION module was applied for the whole core
diffusion calculations.
93
FIG. 2.63. Flow diagram of cell and core calculation for the HTTR benchmark test.
Cell Module was used to generate group constants of fuel and graphite blocks for
succeeding core calculations. In the fuel region, the double-heterogeneity effects originating
from the structure of fuel rods, in which the coated fuel particles are dispersed within fuel
compacts, is considered by the use of collision probability method. The outer radius R for 33
pin block was 3.249 cm, and for 31 pin fuel block was 3.352 cm.
For the control rod guide tube with three large holes, the collision probability method
was also utilized. Radial and axial directional diffusion coefficients for the unit cell were
calculated by Benoist's formula. Similar to a coolant channel in fuel block, the content of the
holes was helium gas at 1 atm at 300K. An isotropic (white) reflection boundary condition
94
was applied in the outer boundary of the unit cell. The unit cell in the control rod guide
consists of a void region and a graphite region. The inner and outer radius of this region was
6.15 and 10.91 cm, respectively.
Average group constants of the fuel rods were calculated by using a one-dimensional
cylindrical fuel cell model as depicted in Figure 2.64. The neutron flux was calculated by the
collision probability method, which was used to average the group constants in the fuel cell
geometry. The average group constants of the fuel block with BPs were also calculated in
one-dimensional cylindrical BP cell model shown in Figure 2.64. The outer radius of the BP
region was 0.7 cm, while the homogenized fuel region outer radius was 13.44 cm. Variation
in axial composition of material was taken into consideration by averaging the number density
in the axial region, taking into account their volume ratio. Table 2-54 shows the infinite
multiplication factor (k) for each fuel cell and its corresponding BP cell.
The whole core calculation was performed using CITATION Module of SRAC-EWS
in -R-Z geometry. The angular distribution was made such that there were nine angulardivisions in one-sixth of the horizontal cross-section of the core.
FIG. 2.64. Fuel and BP Cell model for the SRAC code calculation (all units in cm.).
95
Table 2-54. Calculated infinite multiplication factors (k)

for Fuel Cell and BP Cells
Fuel Layer
1st layer
2nd layer
3rd layer
4th and 5th layer
Fuel block ID
Infinite mult. factor (k)

Fuel Cell
BP Cell
f673320
1.5446
1.3519
f793320
1.5594
1.3872
f943120
1.5996
1.4309
f993120
1.6013
1.4397
f523325
1.5082
1.2566
f633325
1.5375
1.3092
f723125
1.5721
1.3469
f793125
1.5782
1.3674
f433325
1.4786
1.2012
f523325
1.5082
1.2566
f593125
1.5470
1.2966
f633125
1.5553
1.3116
f343320
1.4259
1.1491
f393320
1.4593
1.1979
f433120
1.4932
1.2340
f483120
1.5129
1.2653
2.2.4.2. Results of HTTR-FC Calculations

Multiplication factors (keff) and excess reactivity for first criticality where all control
rods are assumed to be fully withdrawn, are provided in Table 2-55. It indicates that the first
criticality is to be achieved when 18 columns of fuel are inserted. The excess reactivity at
first criticality was determined be 0.577 % k/k. At 17 columns, the keff was determined to be
exactly one with zero excess reactivity, therefore the first criticality was estimated at 18th
column.
96
Table 2-55. Effective multiplication factor and excess reactivity at

first criticality.
Number of fuel
column
Effective mult. factor

[keff]
[% k/k]
16
0.9979
-0.002
17
1.0000
0.000
18
1.0058
0.577
2.2.4.3. Results of HTTR-EX Calculation

Multiplication factors (keff) and excess reactivity for 30 column, 24 column and 18
columns of fuel loaded into the core are shown in Table 2-56. The control rods were assumed
to be in the fully withdrawn condition for all calculations.
Table 2-56. Effective multiplication factor and excess reactivity
at 18, 24 and 30 column loaded in the core.
Number of fuel
column
Effective mult. factor

[keff]
[% k/k]
18
1.0058
0.577
24
1.0692
6.472
30
1.0931
8.517
97
2.2.5. Netherlands
Both NRG and IRI are taking part in the benchmark of start-up core physics of the
HTTR. To compare the performance of the SCALE based IRI code package with that of the
WIMS/PANTHER code package of NRG, a calculational intercomparison has been
performed. This report first describes the NRG and IRI efforts in the first Benchmark (Phase
1) using the Monte Carlo code KENO Va (3-D) and the diffusion theory codes BOLD
VENTURE (2-D) and PANTER (3-D). In the second phase of the Benchmark, only KENO
calculations are performed, for the scram reactivities of the core and reflector control rods and
the isothermal temperature coefficients are given for phase 2. NRG/IRI also participated in
the start-up measurement of reactivity and reactor noise measurements.
2.2.5.1. Computational Methods and Associated Data
The computational tools used at IRI for the cross section generation have been described
previously [2-51]. In short it contains, as a branch, the SCALE-4 code system with master
libraries produced by NJOY from the JEF-2.2 basic nuclear data files. The used reactor codes
for this study are: KENO-Va and BOLD VENTURE. At ECN, the WIMS-7B code system has
been used for this study which has libraries also based on JEF-2.2. The reactor code used at
ECN is PANTHER-5.0.
Cross sections for use in KENO [2-50]
In KENO, only the coated fuel particles (CFPs) in the fuel compacts are homogenised
with the graphite matrix of the fuel compacts; all other reactor components can be modelled
explicitly. As the fuel also contains the only two resonant nuclides (235U and 238U) present in the
core model, the only problem is the generation of cross sections for the homogenised fuel
compacts. The general CFP and compact data are given in Table 2-57.
Table 2-57. General CFP and Compact Data
3
radius (m) density (g/cm )
fuel kernel
298.5
10.79
1st coating
358.5
1.14
material
UO2
PyC (low
dens.)*
PyC
SiC
PyC
2nd coating
389.5
1.89
3rd coating
418.5
3.20
4th coating
464.5
1.87
*PyC: Pyrolitic graphite
Compact dimensions: i/o diameter = 1.00/2.60 cm, height = 3.91 cm.
Since the problem is similar to the generation of cross sections for the fuel pebbles of a
pebble-bed type HTR, the following scheme was adopted from the analysis work for HTRPROTEUS [2-51]:
1. First only the coated fuel particles inside a fuel rod are considered. An infinite closepacked hexagonal CFP lattice is calculated by BONAMI, NITAWL and XSDRNPM.
XSDRNPM is run in spherical geometry for a white boundary elementary cell of the
CFP lattice. This elementary cell contains two regions: a sphere of 0.0597 cm diameter
which contains the fuel kernel of UO2 surrounded by the homogenised mixture of the
98
coating layers and graphite matrix in the fuel compact. The matrix graphite contains
some natural boron to represent impurities in the graphite. A cell-averaged weighted
library, WGH(1), is produced which takes the self-shielding of the fuel in the Caps into
account.
2. An infinite fuel-rod lattice is treated by BONAMI and NITAWL to obtain working
library WRK(1). The unit cell with cylinder geometry has three regions. The innermost
region is a channel filled with helium (0.5 cm radius). This region is surrounded by a
cylinder of 1.3 cm radius with the fuel. The outermost region surrounding the fuel
contains fuel block graphite (r = 3.29 cm). A triangular lattice is assumed with a pitch of
6.2668 cm, consistent with 1/33rd block for the 33-rods fuel block. This step is required
because it provides the unweighted data for the materials outside the fuel region. The
overall Dancoff factor for the core has been deduced from the Dancoff factors for a
lattice of CFPs in a fuel compact and for a lattice of fuel rods in a fuel block [2-51].
3. The library WRK(1) cannot be used for the fuel-rod lattice cell calculation as it would
not take into account the self-shielding in the CFPs. Therefore the WGH(1) and
WRK(1) libraries are merged. All fuel-region materials are taken from the weighted
library WGH(1), the other materials from WRK(1). The resulting library is called
WRK(2).
4. XSDRNPM is run with working library WRK(2) for the unit cell of the infinite fuel-rod
lattice. This unit cell of cylindrical geometry has five radial zones: 1. Channel with
helium (r 0.5 cm). 2. Fuel zone (r 1.3 cm). 3. Graphite sleeve of fuel rod (r 1.7
cm). 4. Fuel hole in fuel block filled with helium (r 2.05 cm). 5. Fuel block graphite
with reduced density to take the fuel handling hole into account. The radius is of this
zone is 3.2903 cm (1/33rd fuel block). If no axial dimensions are used, this run yields
the k of the fuel rod lattice. XSDRNPM is run with a buckling search option to get a
critical system (by the addition of a leakage term in the form of DB2). The weighted
library WGH(2) with zone-averaged cross sections is produced.
5. In order to obtain a working library for KENO, WGH(2) and WRK(1) are merged. The
cross sections for the nuclides inside the fuel compact are taken from WGH(2), and the
cross sections for all nuclides in the other components (He, C, 10B, and 11B) are taken
from WRK(1). The resulting library is denoted as WRK(3).
No group collapsing is done in any of these steps. All libraries contain cross section data
for 172 energy groups! A simpler scheme would have been possible if no comparison had to be
made for two-groups cross sections.
Two-group cross sections
In total, five two-group cross-section libraries have been generated for comparison
purposes:
1. GRLAT2GR: XSDRNPM output of step 1 but with condensation
2. RODLAT_K: output of XSDRNPM k-calculation using WRK(1) as input (step 2)
3. HTTR_K: output of step 4 XSDRNPM, k-calculation and using WRK(2) as input
4. HTTR_B2: output of step 4 XSDRNPM, buckling search using WRK(2) as input
99
5. RODLAT_B2: output of XSDRNPM, buckling search and using WRK(1) as input (step 2)
The order of the numbers in Tables 2-58 and 2-59 correspond to this order. These five
sets enable the assessment of the effects of step 1, the separate treatment for the coated particles,
and of the spectrum used for weighing (buckling search versus k-calculation). Table 2-58 lists
the microscopic total (MT = 1), absorption (MT = 27), and transport (MT = 1000) cross section
for the nuclides in the fuel compact. For the uranium isotopes 235U and 238U also the total
number of fission neutrons (MT = 452), and the fission (MT = 18) and capture (MT = 101)
cross-section are specified in Table 2-59.
Table 2-58. Two-group cross sections for nuclides in the fuel compact (5.2 w% enrichment)
nuclide
10
B
11
Si
235
238
100
tot (b)
group 1
26.86
47.49
49.40
45.26
44.19
4.239
4.267
4.274
4.223
4.225
4.128
4.157
4.164
4.114
4.116
3.754
3.752
3.751
3.736
3.740
2.669
2.612
2.605
2.652
2.651
22.93
28.80
29.00
27.53
27.60
14.77
17.71
15.86
15.47
17.27
group 2
1132.4
2147.2
2157.8
2095.3
2096.0
4.871
4.941
4.945
4.941
4.937
4.752
4.805
4.808
4.805
4.803
3.889
3.943
3.912
3.910
2.941
2.113
2.176
2.178
2.174
2.173
177.5
366.3
363.6
352.2
356.8
10.08
10.87
10.79
10.75
10.84
abs (b)
group 1
24.49
45.16
47.07
42.92
41.85
6.414 E-5
8.961 E-5
9.209 E-5
8.706 E-5
8.562 E-5
1.295 E-4
1.567 E-4
1.576 E-4
1.673 E-4
1.647 E-4
1.231 E-3
1.293 E-3
1.295 E-3
1.437 E-3
1.407 E-3
2.644 E-3
3.548 E-3
3.621 E-3
3.570 E-3
3.511 E-3
12.52
18.20
18.39
17.08
17.12
1.775
3.779
2.879
2.670
3.541
group 2
1130.3
2145.0
2155.6
2093.2
2093.8
1.620 E-3
3.073 E-3
3.088 E-3
2.999 E-3
3.000 E-3
1.000 E-3
1.891 E-3
1.900 E-3
1.845 E-3
1.846 E-3
5.561 E-5
1.062 E-4
1.054 E-4
1.023 E-4
1.036 E-4
5.057 E-2
9.593 E-2
9.640 E-2
9.361 E-2
9.363 E-2
163.2
351.5
349.0
337.6
342.1
0.853
1.545
1.534
1.492
1.510
tr (b)
group 1
39.62
34.45
29.67
34.56
34.59
2.796
2.686
2.608
3.441
3.436
2.808
2.693
2.617
3.373
3.369
7.760
2.701
2.654
3.189
3.185
2.050
2.382
2.399
2.356
2.353
88.03
22.22
19.18
21.45
21.96
64.94
22.11
14.31
12.65
15.24
group 2
1446.0
3074.8
3080.0
2214.5
2246.6
3.970
4.757
4.762
4.666
4.666
3.988
4.857
4.862
4.727
4.729
9.647
3.840
3.794
3.757
3.790
1.807
2.187
2.190
2.133
2.134
733.5
543.2
534.0
374.3
385.8
26.97
11.52
11.38
10.79
10.92
Table 2-59. Total fission neutrons and fission and capture cross section of uranium isotopes
nuclide
U
235
238
group 1
2.440
2.437
2.437
2.438
2.438
2.736
2.740
2.741
2.742
2.741
group 2
2.439
2.438
2.438
2.438
2.438
2.489
2.489
2.489
2.489
2.489
fis (b)
group 1
8.457
11.82
11.92
11.12
11.16
4.581 E-2
4.628 E-2
4.610 E-2
5.093 E-2
5.023 E-2
group 2
137.8
299.9
297.7
288.0
291.8
3.577 E-6
6.640 E-6
6.591 E-6
6.407 E-6
6.486 E-6
capt (b)
group 1
4.066
6.378
6.462
5.962
5.967
1.729
3.732
2.833
2.619
3.491
group 2
25.42
51.63
51.25
49.67
50.31
0.853
1.545
1.534
1.492
1.510
FIG. 2.65. Neutron spectrum in the centre of the compact
101
The spectrum in the grain lattice is much harder than in the fuel-rod lattice, which
explains the lower values of the cross sections. The extra leakage term in the buckling search
is seen to slightly reduce the cross sections, because of the greater leakage of low energy
neutrons compared to high energy neutrons. Figure 2.65 shows the spectrum in the centre of
the fuel-rod lattice, as calculated with step 4 of the cross-section generation procedure.
Cross sections for BOLD-VENTURE [2-50]
In BOLD VENTURE the core region is represented by five rings, containing the A, B,
C, D, and E labelled columns, respectively. The material in each ring is completely
homogenised. In order to maintain the reaction rates, the cross-section generation procedure for
KENO was extended.
The first three steps are identical to the procedure for KENO. The fourth step is similar,
but now a cell weighting is performed instead of a zone weighting. Subsequent steps are new.
4. XSDRNPM is run with working library WRK(2) for the unit cell of the infinite fuel-rod
lattice. This unit cell of cylindrical geometry has five radial zones: 1. Channel with
helium (r 0.5 cm). 2. Fuel zone (r 1.3 cm). 3. Graphite sleeve of fuel rod (r 1.7
cm). 4. Fuel hole in fuel block filled with helium (r 2.05 cm). 5. Fuel block graphite
with reduced density to take the fuel handling hole into account. The burnable poison
rods are not taken into account. The radius of this zone is 3.2903 cm (1/33rd fuel block).
XSDRNPM is run with a buckling search option to get a critical system (by the addition
of a leakage term in the from of DB2). The weighted library WGH(2) with cellaveraged cross sections is produced.
5. Unweighted cross sections for the materials outside the fuel blocks (i.e. inside the
control rod guide blocks and reflector) have to be added to WGH(2). These unweighted
cross sections of C, 10B, and 11B, were taken from WRK(1). The resulting library is
called WRK(3).
6. XSDRNPM is run with library WRK(3) for a 1D-model of the reactor. This model
contains six radial zones. The first five represent the five rings of the core region, the
outermost zone represents the permanent reflector. The radii of the zones were
calculated to be 19.0064 cm, 50.2861 cm, 82.8468 cm, 115.6112 cm, 148.4444 cm, and
214.9814 cm. With these radii, the area of the rings is identical to the true area of the
columns (the pitch in the core region is taken to be 18.1 cm, hence the space between
the blocks is taken into account). The material within each zone is completely
homogenised. The atomic densities in the homogenised zones can be found in the
appendix. Note that the burnable poison rods are not taken into account. XSDRNPM is
run with a buckling search option and with zone weighting, producing weighted library
WGH(3). For the homogenised KENO model the 172 groups were not condensed, for
BOLD VENTURE the groups were condensed to 13 broad groups, like for HTRPROTEUS [2-52].
Two-group cross sections
Two-group cross section data is obtained by condensing the 172 fine groups to 2 broad
groups in step 6 of the procedure in section 2.2.1. The results are summarised in Tables 2-60
thru 2-63.
102
Table 2-60. Two-group cross sections for the uranium isotopes in the fuel compact (5.2 w%
enrichment) in the radial zones B, C, and D.
nuclide
235
U B
C
D
238
U B
C
D
tot (b)
group 1
group 2
abs (b)
group 1
group 2
tr (b)
group 1
group 2
27.79
29.13
27.34
15.71
16.08
15.59
368.3
378.6
380.8
10.47
10.50
10.50
17.18
18.39
16.78
2.716
2.929
2.651
354.2
364.5
366.7
1.544
1.581
1.589
21.30
23.47
21.04
12.34
12.11
12.26
372.4
372.8
385.3
10.45
10.45
10.48
Table 2-61. Two-group cross sections for the uranium isotopes in the fuel compact (5.2 w%
enrichment) in the radial zones B, C, and D.
nuclide
group 1
U B
C
D
238
U B
C
D
2.438
2.438
2.439
2.742
2.741
2.742
235
group 2
fis (b)
group 1
2.438
2.438
2.438
2.489
2.489
2.489
11.18
11.92
10.93
5.486 E-2
4.884 E-2
5.667 E-2
group 2
capt (b)
group 1
group 2
302.3
311.1
313.0
6.640 E-6
6.805 E-6
6.841 E-6
6.000
6.469
5.846
2.661
2.880
2.595
51.94
53.44
53.72
1.544
1.581
1.589
Table 2-62. Two-group cross sections for the non-fissionable nuclides in the fuel compact (5.2
w% enrichment) in radial zones B, C, and D.
nuclide
O
Si
C
10
11
B
C
D
B
C
D
B
C
D
B
C
D
B
C
D
tot (b)
group 1
3.807
3.814
3.805
2.730
2.672
2.749
4.177
4.220
4.163
45.86
49.71
44.63
4.287
4.332
4.273
group 2
abs (b)
group 1
3.775
3.773
3.773
2.106
2.108
2.108
4.639
4.636
4.635
2177.5
2233.4
2245.6
4.778
4.778
4.778
1.543 E-3
1.371 E-3
1.590 E-3
3.694 E-3
3.707 E-3
3.683 E-3
1.779 E-4
1.647 E-4
1.813 E-4
43.47
47.33
42.23
8.849 E-5
9.315 E-5
8.700 E-5
group 2
fis (b)
group 1
group 2
1.063 E-4
1.090 E-4
1.096 E-4
9.728 E-2
9.978 E-2
1.003 E-1
1.917 E-3
1.966 E-3
1.977 E-3
2175.3
2231.3
2243.5
3.116 E-3
3.197 E-3
3.214 E-3
3.350
3.394
3.351
2.500
2.455
2.501
3.515
3.609
3.510
34.01
39.40
33.07
3.593
3.690
3.589
3.620
3.620
3.619
2.059
2.059
2.061
4.566
4.568
4.564
2199.1
2201.1
2266.9
4.498
4.497
4.501
103
Table 2-63. Two-group cross sections for the nuclides in the graphite of the blocks in
all radial zones
nuclide
C
A
B
C
D
E
reflector
10
B A
B
C
D
E
reflector
11
B A
B
C
D
E
reflector
tot (b)
group 1
4.290
4.130
4.183
4.113
4.347
4.524
57,31
46.77
50.72
45.50
65.21
97.35
4.404
4.239
4.294
4.221
4.462
4.640
group 2
abs (b)
group 1
4.810
4.844
4.846
4.847
4.815
4.818
2483.0
2316.6
2380.7
2393.3
2933.7
3251.3
4.976
4.995
5.003
5.004
5.020
5.052
1.178 E-4
1.440 E-4
1.351 E-4
1.463 E-4
1.161 E-4
1.274 E-4
54.99
44.43
48.40
43.16
62.91
95.12
1.017 E-4
8.824 E-5
9.325 E-5
8.666 E-5
1.116 E-4
1.519 E-4
group 2
fis (b)
group 1
group 2
2.185 E-3
2.039 E-3
2.095 E-3
2.106 E-3
2.580 E-3
2.858 E-3
2480.7
2314.4
2378.5
2391.1
2931.5
3249.1
3.554 E-3
3.316 E-3
3.407 E-3
3.425 E-3
4.199 E-3
4.654 E-3
3.442
3.418
3.528
3.412
3.500
3.971
38.79
34.60
40.15
33.63
38.77
79.34
3.515
3.495
3.609
3.490
3.583
4.066
4.772
4.774
4.776
4.779
4.786
4.838
2600.2
2344.3
2345.5
2424.5
2788.0
3168.6
4.701
4.705
4.705
4.718
4.721
4.755
Cross sections for PANTHER

Cross sections for the reactor code PANTHER have been generated by means of the
code suite WIMS-7B. Apart from service modules for group condensing and material
homogenisation, two collision probability modules were used to calculate the flux weighted
cross sections of the fuel cell (PROCOL) and for the fuel blocks or assemblies, control guide
blocks and reflector blocks (PIJ).
In order to avoid much extra work, densities, impurities and sizes of graphites, CFPs and
coatings, weighted means of these parameters where appropriate, have been derived to be used
all over the reactor. This leads to the following standardised parameters for the CFPs (Table 264).
Table 2-64. Parameters for CFPs

density (g/cm3)
material
radius (m)
fuel kernel
297.95
10.774
UO2
1st coating
358.80
1.127
PyC (low density)
2nd coating
389.45
1.896
PyC
3rd coating
418.35
3.225
SiC
4th coating
464.20
1.866
PyC
Compact dimensions: i/o diameter = 1.00/2.60 cm, height = 3.91 cm.
104
Table 2-65. Parameters for Graphites

density (g/cm3)
matrix
sleeve
fuel / control block
repl. reflector
perm. reflector
1.690
1.770
1.770
1.760
1.732
impurity
Bnat)
0.82
0.37
0.40
0.37
1.91
(ppm
PROCOL [2-50]
In the WIMS-suite a cell module PROCOL, based on collision probabilities, exists to
calculate fluxes in systems with spherical grains packed in a matrix with an annular geometry.
A cell radius of 3.29 cm has been used, consistent with a lattice of a 1/33rd part of a fuel
block or assembly, in which explicitly modelled: the inner gas channel (r = 0.50 cm), the
compact (r = 1.30 cm), the gas gap (r = 1.3125 cm), sleeve (r = 1.70 cm) and the fuel hole
drilling in the fuel block (r = 2.05 cm). Using this model, flux weighted cross sections are
obtained for homogenised CFPs + matrix + gas gap, to form the compact material with cross
sections in the 69 neutron energy groups structure of the library.
The spectrum in the centre of the inner gas channel in the compact with 5.2 w%
enrichment is shown in Figure 2.65. Comparison with the spectrum as obtained with the KENO
cross sections is very good. Differences are only due to the resolution of the spectrum with the
number of energy groups used in the calculations (KENO: 172 vs. WIMS: 69).
Accordingly obtained cross sections were condensed to 16 neutron energy groups for
subsequent use in the WIMS assembly module PIJ, which calculates collision probabilities in
multi-pin assembly systems.
For comparison purposes microscopic cross sections for the nuclei present in the
compacts were condensed to two group cross sections. In WIMS only microscopic absorption
and fission cross sections are easily available, but for some elements transport and total cross
sections could be deduced from macroscopic cross sections. Values for an enrichment of 5.2
w% are given in the Tables 2-66 and 2-67 and can be compared with those values given in
Tables 2-58 and 2-59. Agreement is in general rather good which can be confirmed by the
spectrum comparison of Figure 2.65 and the calculated neutron multiplication factors: kinf =
1.499 for the KENO-cell and kinf = 1.493 for the PROCOL-cell.
Table 2-66. Comparative two group cross-sections (5.2 w% enrichment)
tot(b)
abs(b)
tr(b)
nuclide group 1
group 2
group 1
group 2
group 1
group 2
10
B
42.20
2095
C
4.086
4.805
1.882E-4 1.856E-3 3.348
4.702
O
1.527E-3 1.029E-4
Si
3.623E-3 9.397E-2
235
U
1.692E+1 3.397E+2
238
U
2.747
1.499
105
FIG. 2.66. Axial composition of the fuel block in PIJ.
106
Table 2-67. Uranium isotope fission neutrons and fission and capture cross-sections
(5.2 w% enrichment)
group 1
2.438
2.742
235
U
U
238
group 2
2.438
2.489
fis(b)
group 1
1.096E+1
5.768E-2
group 2
2.897E+2
6.437E-6
capt(b)
group 1
5.968
2.689
group 2
49.91
1.499
PIJ [2-50]
For modelling in PIJ the fuel assembly has been adapted in the following way:
1.
2.
3.
4.
The stack of compacts has been moved to the top such that the upper rim of the
upper compact is flush with the fuel block,
The upper graphite plug and buffer plate has been moved to the bottom,
The stack with the burnable poison (BP) pellets and graphite disks has been
moved to the top as in 1, while the length of the upper section of the BP stack
has been changed from 20 cm to 15 cm and the lower section to 25 cm
The fuel handling hole has been simplified by taking an effective diameter of
4.017 cm and a length of 25.0 cm.
This way four layers in the assembly can be created (See Figure 2.66):
1.
2.
3.
4.
First layer of 15 cm height with compacts, fuel handling hole (FHH) and BP
pellets,
Second layer of 10 cm with compacts, FHH and graphite disks,
Third layer of 25 cm with compacts, graphite for FHH and with BP pellets,
Fourth layer of remaining 8 cm with a mix of 4.75 cm of compact, 2.35 cm of
graphite and 0.9 cm of void at the fuel positions and graphite at the FHH and BP
positions.
The void has been modelled in the empty BP insertion leg.

For each layer a model of the fuel assembly has been laid out in which the hexagonal
perimeter has been replaced by an equivalent circle (radius 19.01 cm). Within this circle the fuel
positions (comprising: inner gas space, compact, sleeve and outer gas space), FHH and BP
insertion holes are modelled at the exact positions and filled with the materials in conformance.
This circle in turn is surrounded by another circle (radius 38.01 cm), divided into 12 segments,
to accommodate the matching surrounding materials for the fuel assembly under study (Figure
2.67). It makes a total of 206 material regions per assembly layer.
The coolant bearing reflector blocks in the 1st, 2nd and 8th reactor layer are modelled in
the same way but with empty fuel holes and of reduced diameter.
To reduce the number of materials, the PIJ model is finally divided into seven regions:
one central region comprising the FHH position and the six inner fuel positions, and the six
surrounding segments (Figure 2.67). Materials within a region are homogenised or smeared to
one material. Finally the seven materials for the four layers are smeared, according to their
height, to seven final materials for one assembly having flux weighted cross sections in 16
neutron energy groups.
107
FIG. 2.67. Radial composition of the fuel and control guide block in PIJ.
The procedure for the control guide blocks (Figure 2.67) and reflector blocks is similar;
also divided into seven regions but with only two layers, with and without FHH.
The advantage of the sub-division in seven regions is that the anomalies in a block, like
BP stacks, absent fuel pins, control guide holes, control rods, etc. are confined to only one
region a piece and are not smeared over the entire block. This allows for more pronounced local
absorption and/or streaming, which form major problems for modelling this kind of reactor
cores.
108
For all 48 different block configurations (enrichments, block types, surroundings, etc.)
two runs with PIJ were done; first a run without control rods (unrodded) and a second run with
control rod material modelled in the control guide holes and using rodded material in that sector
of the surrounding where present (rodded). Afterward all cross sections were condensed to two
energy groups (Eth = 2.1 eV) and organised in such a way that it can be used in the reactor code
PANTHER, leading to 336 different materials in as well a rodded state as an unrodded state.
By making use of the modules PROCOL and PIJ the double heterogeneity formed by
the CFPs and the fuel rods has been modelled explicitly and therefore no Dancoff factor has to
be introduced.
2.2.5.2. Results of HTTR first criticality (Phase 1) [2-53]
The KENO model is a very detailed model of the HTTR in which practically all
components are modelled explicitly, with the following exceptions:
1 As mentioned before, the coated fuel particles were homogenised with the graphite
matrix of the fuel compacts.
2 It is not possible to model hexagonal blocks in KENO-Va. Therefore, the permanent
reflector was approximated by a cylinder of 214.98 cm radius which preserves the
volume of the actual reflector. Furthermore, the hexagonal blocks in the core and in the
replaceable reflector were represented by cylinders of 36 cm diameter (the distance
between the parallel faces of the blocks). These cylinders (which contain all fuel rods
and the two burnable poison rods or all coolant channels) were placed in a large
cylinder of graphite with a radius of 162.9 cm (Figure 2.68).
In the BOLD VENTURE model the HTTR is represented by an R-Z model. It contains
six zones in the radial direction, and nine in the axial direction, one for each layer. The six
radial zones are the:
1.
2.
3.
4.
5.
6.
Central control rod guide column (column A)

First fuel zone (the six B columns)
Second fuel zone (the 12 C columns: 6 fuel columns and 6 control rod columns)
Third and fourth fuel zone (the 18 D columns)
Replaceable reflector (the 24 E columns)
Permanent reflector
The height of each layer is 58 cm, except layer 9 (42.9 cm). The radii of the zones are:
19.01, 50.29, 82.85, 115.61, 148.44 and 214.98 cm. Calculations were performed with a 2 cm
mesh, both in axial as radial direction and the BP rods were simulated by adding boron to the
radial zones B, C and D to such an extend that a reactivity change, as determined by auxilary
KENO calculations, was reached.
For PANTHER a 3-D model has been developed in a hexagonal representation, taking
a cluster of seven sub hexes (size: 13.68 cm flat-to-flat) per hexagonal reactor assembly
position in the radial direction and 5 layers per assembly in the axial direction. This leads to
937 radial reactor channels with an equivalent radius of 220 cm and 45 axial layers of 11.6
cm.
Control rods, those left partially inserted in the E-column ring, reached only till the
bottom level of the upper block (464 cm level).
109
FIG. 2.68. Fuel assembly (upper left), the KENO model for the critical reactor (upper right,
and the KENO model for scram with all control rods (C, R1, R2 and R3) (lower).
110
Materials defined and prepared in the WIMS data generation phase has been laid down
according to proper compositions and orientations of the reactor assembly blocks in the
reactor. For the simple core all enrichments were set at 5.2 w%.
In PANTHER the assemblies which carry control rods are represented by two sets of
nuclear data: one set for the part where is no control rod inserted (unrodded) and a set for the
rodded part. The control rod insertion depth for a certain control rod bank determines whether
PANTHER uses the set for the rodded material or for the unrodded material in a particular
mesh, thus enabling to drive a control rod.
The results of the calculations by the different codes are presented in Table 2-68.
Good agreement can be found between the KENO and PANTHER results, the higher keff
values for BOLD VENTURE can be attributed to neutron streaming in the control rod guiding
holes
.
Table 2-68. Comparison of the results
KENO
keff simple core
keff fully loaded core
- rods withdrawn
- rods inserted
critical insertion
- above bottom core
1.1278 0.0005
1.1584 0.0005
0.6983 0.0005
BOLDVENTURE
1.1592
PANTHER
1.1974
1.1595
0.7510
0.685 0.010
161.5 cm
178.9 cm
170.5 cm
Measured
1.1251
2.2.5.3. Results HTTR-CR and HTTR-TC for Phase 2 [2-53]

The second benchmark of the HTTR core physics (Phase 2), defined by JAERI
(Yamashita et.al.,1999b), were also calculated by using the Monte-Carlo code KENO-Va
(V4.3) for a fully loaded core with 30 fuel elements and are presented in the Working
Material of the IAEA meeting (CRP-5, 1999 and Trkcan et.al.,1999).
In Phase 2 the following answers were requested by the organisers:
- Scram reactivities (control rod worth) for the control rods at critical position and
after a scram of the reflector control rods (HTTR-SR) and after a scram with all the
control rods (HTTR-SA), both at a temperature of 300K.
- Isothermal temperature coefficients for a fully loaded core from 280K to 480K in six
temperature steps (HTTR-TC). Where the control rod settings (C, R1 and R2) have
slightly different settings due to a temperature elevation and the critical insertion of
the control rods C, R1 and R2, (R3 stays fully out) at 480K.
For the new benchmark calculations, new cross-sections were prepared for seven
different temperatures using the aforementioned procedure. The resulting cross section
libraries contain data for 172 neutron energy groups. The geometry input for KENO is revised
to be able to calculate the questions of the benchmark. The KENO model for a critical reactor
and with all control-rods inserted is shown in Figure 2.69. The results of the benchmark-phase
2 will be summarised.
111
The scram reactivity of control rods (k/k) is defined by:

k k
R = Crit RCR in
kCrit .k RCR in
With: kCrit. . : Effective multiplication factor at critical CR position
and kRCR-in: Effective multiplication factor at CR position after scram.
Table 2-69 provides the scram reactivity for two different scram conditions. Also
given are the calculated and measured scram reactivities (Fujimoto et.al.,1999).
Table 2-69. Scram Reactivities (Using KENO)
CR Group
C
R1
R2
R3
keff
(average)
calc
meas
Critical position
(mm) HTTR-Crit
1789
1789
1789
Full out
1.0093 0.000

(mm) HTTR-SR
1789
1789
-41
-41
0.9178 0.0005

(mm) HTTR-SA
-41
-41
-41
-41
0.6809 0.0005
0.0988 0.0007
0.120 0.012
0.4778 0.0007
0.46 0.04
The results using PANTHER calculations include the following:

Keff at critical position = 1.0088; Keff after scram HTTR-SA = 0.7317;
= 0.375
This scram reactivity is too low but can be explained by neutron streaming in the control
rod holes and is to be recalculated by means of anisotropic cross section
In the second question of the benchmark, the isothermal temperature coefficients
(HTTR-TC) for a fully loaded core between temperatures 280K to 480K (in six steps) were
asked, where the control rods C, R1, and R2 have slightly different settings due to
temperature elevation (13 mm).
The effective multiplication factors should be calculated for the following
temperatures: 280, 300, 340, 380, 420, 460 and 480 Kelvin and the isothermal temperature
coefficients should be calculated at: 290, 320, 360, 400, 440 and 470 Kelvin. The insertion
depth of C, R1, R2 is the same at level =1776 mm and R3 again is fully withdrawn. Also the
critical position for those control rods at 480K, with R3 fully out, is requested for the
benchmark.
The following relation should evaluate the isothermal temperature coefficients for a
fully loaded core from the effective multiplication factors:
k k
1
n = n+1 n .
kn +1.kn (Tn +1 Tn )
n : Temperature coefficient between Tn and Tn+1 ( k/k/K)
Tn : Core temperature at nth measurement (K)
Tn+1 : Core temperature at n+1th measurement (K)
kn : Effective multiplication factor at Tn
kn+1 : Effective multiplication factor at Tn+1.
112
Results of the calculations are shown in Figure 2-69. The calculated isothermal
temperature coefficient (average between 320K and 440K) is 14.7 (pcm/oC), while the
measured value equals 14.2 (pcm/oC) on the average. The calculated critical control rod
position at 480K is 1879 mm, while for the measurements at T= 395K; 1873 mm and at T=
418K; 1903 mm are found.
The PANTHER value for iso = 15.2 (pcm/K), and the control rod position at 480K
is calculated as 1934 mm.
k-eff for different temperatures
Isothermal Temperature Coefficient

0
1.015
Isothermal Temperature
Coefficient (pcm/K)
1.010
k-eff
1.005
1.000
0.995
0.990
calculated
interpreted
y = 14.78
-5
-10
-15
0.985
0.980
-20
250
300
350
400
Temperature (K)
450
250
500
300
350
400
Temperature (K)
300
350
400
temperature (K)
450
500
k-eff for different conrol rod settings
1.32
1.000
neutron generation time (ms)
1.005
T = 480 K
k-eff
0.995
0.990
0.985
0.980
0.975
0.970
1700
1.30
1.28
1.26
1.24
1.22
1.20
1.18
1750
1800
1850
C, R1, R2 (mm)
1900
1950
250
450
500
Figure 2-69. The effective multiplication factor (upper left) and isothermal temperature
coefficient at different core temperatures, the multiplication factor for the different control
rod settings of C, R1, R2 while R3 fully out (lower left) and the neutron generation time
versus core temperatures (lower right).
2.2.5.4. Reactivity and reactor noise measurements of the HTTR during the start-up
Cores [2-53]
During the start-up phase of the HTTR at different core configurations, reactivity and
the reactor noise measurements were carried out in parallel with measurements of the HTTR
Physics group. For these measurements two temporary compensated ionization chambers
CIC-A and B were used. Figure 2.70 shows the horizontal cross-section of the core and the
positions of the detectors. Measurements were carried out by using the signal processing
system DSA-2 (Trkcan, 1993) in real-time. During the on-line reactivity experiments, the
measured DC signals were digitised and the reactivity is calculated by using the Inverse
Kinetics Method (IK).
During the first critical approach after the loading of the 19th fuel assembly, the
source criticality at very low power is achieved. For criticality, first the neutron source is
removed and then by moving the central control rod (C) to compensate for the reactivity until
the first criticality of the reactor is reached on Nov. 10 1998. Figure 2.71 shows the result of
reactivity measurements during this approach to criticality.
113
FIG. 2.70. HTTR horizontal neutron detector positions.
FIG. 2.71. Approach to criticality.

After each new few fuel loading, the reactivity value of the control rods were
measured successively with the IK-method by the HTTR physics group and by DSA-2
system. As an illustration the experimental result is given in Figure 2.72 for full core with 30
fuel assemblies. By the move of a control rod with a small step, the neutron flux is increased
while the reactor power is kept in the same power range by compensating the reactivity effect
by another control rod. For each action where the reactivity is constant over about 80 seconds
the calculated reactivity is averaged. This way the average reactivity worth of the control rod
is determined for this stepwise change.
FCh. 1
F Ch. 2
R Ch. 1
0.1
-0.1
1
0
0
250
500
750
1000
R.Ch 1&2 ($)
0.2
5
F. Ch 1&2 (V)
R Ch. 2
-0.2
1250
Time (s)
Fig. 2.72. Measured reactivity at HTTR full core. Two measured signals of Ch1 (A) and Ch2
(B) and the computed reactivity (blocks match each other, scale on right).
114
For the reactor noise measurements, the same neutron detector signals were used and
the reactor at very low power is kept as stable as possible especially for this measurement.
Signals were conditioned for the noise measurements by using high- and low-pass filters with
a gain amplifier.
Inherent to this type of reactor, the neutron generation time is long and was calculated
to be: 1.1730.001 ms. The prompt neutron decay constant is quite close to the decay of the
fastest delayed neutrons, therefore no intermediate plateau can be recognized in the measured
spectral functions such as the Normalized Auto and Cross Power Spectral Density (NAPSD
and CPSD).
Figure 2.73 is an example of the measurements on the 21 fuel element critical core at
very low power. Our investigations indicated a shortcoming of the bandwidth of the used
current amplifiers due to the large cable capacity of about 100-m of cable between the
detectors and the amplifiers. This situation was not possible to change during the
measurements.
FIG. 2.73. The NAPSD(f) and the NCPSD(f) functions measured 21st fuel loading. The cross
spectra do not give clear break frequency.
FIG. 2.74. Calculated Transfer functions for 18 and 30 fuel elements loading.
115
The measured coherence between the two neutron detector signals was 0.8 at 0.1 Hz
and gradually decreased to 0.2 at 1 Hz and the phase between them practically zero. The
calculation of the zero-power transfer function (Figure 2.74) for different number of fuel
loading, the transfer function shape did not changed considerably.
2.2.5.5. Summary of results [2-53]
On the level of cell calculations a good agreement has been obtained between the
cross sections and the spectra as prepared by the SCALE-system and as prepared by WIMS.
Calculations with detailed geometry converged to very good agreement between the results of
PANTER and the results of KENO with an exact geometrical model. In the second phase,
KENO results gave very well the measured values of the scram reactivities as well as the
estimation of the isothermal temperature coefficient within the requested temperature interval.
Therefore, we can conclude that the Benchmark calculations of the start-up physics
calculations were successful and that the results of the reactivity measurements and the
reactor noise analysis done at the HTTR, using the DSA-2 system, resulted with a good
agreement with the results of the HTTR Physics Group.
116
2.2.6 Russian Federation

2.2.6.1. General analysis method and model description
For benchmark calculations in the diffusion approximation, the computer codes
WIMS-D/4 and JAR [2-54, 2-55] were used. MCU [2-56] and MCNP [2-57] codes,
implementing the Monte Carlo method were used as reference.
WIMS-D/4 code [2-54] has been used for computation of few-group constants for fuel
and the reflector blocks. The code is provided by 69 groups system of nuclear constants in
neutrons retardation and thermalization range on the basis of estimated nuclear data in
ENDF/B6 format. The characteristic feature of WIMS is a two-stage approach to solution of a
task on spatial-energy neutron distribution in a reactor cell. In the first place, a detailed
spectrum in 69 groups for every zone, typical for the cell is calculated: fuel, shell, coolant and
moderator. Then cross-sections are convoluted to a specified few-group approximation, where
detailed distribution of neutrons in geometric zones of the unit cell is calculated. The S4method was used in the first and second stages of the calculation. In calculations of spatial
resonance screening the equivalence theorem is employed, with the help of which the
heterogeneous problem comes to an effective homogeneous. Detailed matrices of scattering
cross-sections for main moderators are employed in the neutron thermalization calculations.
For calculating of fuel cells with double heterogeneity of fuel allocation at the fuel
particles and fuel compacts level in the cell, containing burnable poison rods, Segev algorithm
[2-57] was used to calculate macro-sections for a cell, associated with a fuel compact. When
using the Segev formalism to calculate parameters of fuel compact cell containing fuel
particles, it is necessary to carry out calculation of three cell variants in cluster PIJ or PIJPERSEUS option:
a) a cell associated with fuel particle is calculated, whereby a kernel is represented as a
rod with radius rc= 9/16 rsph, on which matrix moderator and shells , related to one
kernel is spread. Radii of the rod and cell are chosen so that collision probability for
neutron, born in fuel, would be the same in cylindrical of module and spherical
geometry. At cell calculation by WIMS at the given stage a value of Dankoff factor for
inner region of the cell () is found;
b) a cell associated with annular fuel compact, in which the fuel and the compact
matrix graphite are homogenized, is calculated. Panniculus equivalent of serves the
shell from graphite, taking into account of gas, related to one compact, serves a layer
of external moderator. At this stage, a value of Dankoff factor for outer region of the
cell () is determined;
c) then input is calculated:
input =
1
Vm m
1 + (1/ 1)
1/ L
Vm m +
1/ A + 1/ 1
where Vm- volumetric fraction;

m- moderator scattering cross-section;
L=M1/3.lcell;
M - number of fuel particles;
lcell - mean chord;
A - Bell factor.
(1)
(2)
117
The value of input is set in WIMS and calculation as per point b is repeated.
Resulting from the last calculation macroscopic cross sections and multiplication factor for a
fuel compact cell with double heterogeneity are acquired. It is worth noting that value of input
weakly depends on fuel temperature and isotopic content in case of fixed geometrical
parameters of a cell.
Basic results of diffusion calculations were obtained in two-group approximation with
thermal and epithermal neutrons cut-off energy of 0.625 eV, selected from the condition of
obtaining values of cross-sections for transition to higher energy close to zero, that is defined
by the calculation model of scattering in the reactor code, that takes account of energy
transitions only in direction with energy decrease.
In the case of the cell with burnable poison rods calculation, the same option of WIMS
code was used, but poison rods in the cell was surrounded by effective fuel composition
accounting shielding factors.
The macro-constants received at WIMS calculation stage were then accepted as input
data for reactor calculations by JAR code. The JAR code [2-55] is based on the nodalization
method and makes it possible to carry out three-dimensional analysis of reactor in few-group
diffusion approximation. The reactor core and the reflectors in a calculations model are
represented as hexagonal prisms, broken down on triangles and divided into axial zones.
For the diffusion calculations the effect of nonuniformity of poison in burnable rods
through its height was not accounted. Burnable pellets and graphite pellets were homogenized
in the vertical direction. The construction of control rods absorber through its height is
considered as uninterrupted.
The MCU code [2-56] based on Monte-Carlo methodology employs a combination of
DLC/MCUDAT-1 data libraries. MCU is used to compute a continuous spectrum of neutrons
moderation within energy range 10-5 eV to 20 MeV, and to solve the neutron transport
problem with an external source or criticality problems. To describe cross-sections in the field
of unresolved resonance the subgroup dividing method is used, while for resolved resonance
range the detailed description of cross-sections is possible on the basis of special data
libraries.
MCNP code (Version 4A) [2-57] has been used to check results of the diffusion
calculations. Nowadays the MCNP code is one of the most efficient computer codes, which
employ the Monte Carlo methodology for analysis of high temperature gas-cooled nuclear
reactors. Possibility of the MCNP code application for analysis of HTGRs with double
heterogeneity of fuel allocation on the basis of comparison with other codes, was studied
before. The heterogeneous model by MCNP was described so that each coated fuel particle
was located in unit lattice by lattice geometry in fuel compact (not randomly).
Among essential features of MCNP, a capability to compute continuous spectrum of
neutrons moderation within energy interval 0 to 20 MeV in approximation of point-by-point
representation of nuclear data, should be pointed out. Point cross sections of neutrons
interaction with nucleus were prepared by using of NJOY code in ENDF /B6 format.
In criticality analysis a multiplication factor keff was estimated by three ways for larger
reliability: viz.: estimation on collisions (col), estimation on absorption (abs), estimation on
free track length (trk). Moreover, a combined estimation of keff based on the first three
estimates was made.
118
For the Monte-Carlo calculations the detailed geometry was taken into account, but
the presence of spacers on the fuel rod surface was not into account. Beside in the MCNP-4A
code version the location of coated particles in graphite matrix is described in the form of
cubic lattice.
Design model of the reactor (Figure 2.75) consists of nine layers of hexagonal fuel
blocks and graphite blocks of reflector, divided into four subzones each. The reactor core
includes reactivity compensation rods viz.: control rods and channels for the reserve shutdown
system. It is supposed in the analysis, that the poison in the reactivity compensation rods is
distributed axially without gaps, which actually are available in the rods and are about 6.45 %
from the overall length of rod.
To calculate few-group cross sections and multiplication factor, the fuel block is
modeled by an annular cell, area of which corresponds to 1/3 area of a fuel block; a burnable
poison rod, a hole for it or a graphite disk are set in the center and surrounded by fuel,
medium content of which corresponds to 1/3 of the block. The fuel content of the surrounding
zone is homogenized taking into account a double heterogeneity of fuel arrangement and from
condition to preserve multiplication factor value and migration parameters in the zone
compared with a cell of the annular fuel compact (Figure 2.76).
Macro-cross sections of graphite reflector cells, including absorber rods or gas cavities
for their location, were also calculated using the WIMS code. A calculation model of the cell
is the following: in the center of the cell a cavity or an absorber rod (with detailed description
of its geometry) is set, which is surrounded by a graphite ring with outer of radius, determined
from the condition of surrounding graphite volume conservation. A layer, containing
homogenized fuel content is used as an outer layer of the cell (for shaping spectrum).
Information about the codes, calculation models and nuclear data are summarized in
Tables 2-70, 2-71 and 2-72.
Table 2-70. Codes, model and library of nuclear data for Diffusion calculations
Items
Nuc. data file
Fuel cell code
Theory
Model
Cut - off energy
No. of groups
BP cell code
Theory
Model
No. of groups
Control rod cell cal.
The theory
Model
Number of groups
Core cal. Code
Model
No. of groups (Fast + Thermal)
Name of Country
Russia
Name of Institute
B
ENDF/B6
WIMS-D/4
S4
Cylindrical cell
0,625 eV
69
WIMS-D/4
S4
Cylindrical cell
69
WIMS-D/4
S4
Cylindrical cell with the central absorbing zone
69
JAR-3D
Triangular lattice (6 points / blocks)
1+1
119
120
Cavity for reserve

shutdown system
Fuel blocks with

different enrichment on
uranium-235
Control rods
Fuel block with

burnable poison rods
Bottom reflector block

a) reactor in plan
b) fuel column along height
FIG. 2.75. Reactor calculation model.
Graphite
disks
BP pellets
Fuel kernel
(UO2)
Multilayer
coating
Annular fuel compact
Fuel matrix
graphite related to
one fuel particle
Burnable poison rod
Graphite sleeve
Gas gap and graphite spacing ribs
Homogenized fuel
Fuel block graphite related to a fuel compact
) plan view
Fuel particle
Fuel compact
b) axial view
Fuel cell with burnable poison rod
121
FIG. 2.76. Fuel cell calculation mode.
Table 2-71. Codes, models and nuclear data library for Monte-Carlo calculations
Name of Country: Russia
Items
Code
Nuc. data file
Energy structure
Coated fuel particles
History
Batches
Skipped batches
Name of Institute: RRC KI

MCU
DLC/MCUDAT-1.0
Continuous
Fuel kernels are not smeared
200
5000
1
Name of Country:
Russia
Name of Institute:
IBRAE
MCNP 4A
ENDF/B6
NJOY
Continuous
Detailed account
2000 (up to 16000)
1000
10
Table 2-72. Methodical errors

Items
Calculation refining
Taking into account

Monte-Carlo
Diffusion
calculation
calculation
(MCNP)
Forms and dimensions

Detailed description of spacer of
homogenized
sleeve
Detailed description of gaps
+
between sleeve and fuel compact
2 Graphite block
Detailed description of coolant
+
holes, gaps between dowel pin and
dowel socket, chamfering and spot
facing
3 Handling hole of graphite Detailed description of handling
+
block
hole in all types of blocks
4 Position of burnable
Horizontal locations of burnable
+
absorber
absorbers are displaced on 3 mm to
outside in fuel blocks
5 Shock absorber of
Detailed description of shock
+
control rod (CR)
absorber
6 Heterogeneity of coated Statistical treatment of CFP
cubic cell
fuel particle (CFP)
location in fuel compact
Composition and density
7 Fuel compact
Fuel compact load drift from
nominal
8 Air in graphite block
Residual air in graphite block
9 Impurity in dummy block
Impurity revising from 2,5ppm
+
to 3,lppm
10 Partial insertion of R2-CR Partial insertion of R2-CR
+
11 Composition of graphite
Different composition of a part
+
block
of permanent reflector and
dummy block
1 Graphite sleeve
122
homogenized
+
homogenized
+
determined
location
+
+
+
2.2.6.2. First Criticality [HTTR-FC]

Analysis of keff dependence on sequence of fuel columns loading was conducted,
the results of which are depicted in Figure 2.77. In connection with this a consecutive loading,
considered accordingly the specification JAERI-memo 10-005 and symmetrical loading, is
illustrated in Figures 2.78, 2.79 and 2.80. First results of HTTR-FC calculations were
presented at IAEA Research Coordination Meeting [2-59]. The results on phase 2 of
Benchmark Problems of HTTR were presented at IAEA CRP-5 meeting in China, October
18-22, 1999 [2-60, 2-61].
Characteristic
Diffusion calculations
OKBM
16
1.005
0.498
Organization
keff
Reactivity excess, %k/k
Value
Monte Carlo calculations
IBRAE
RRC KI
16
17
1.0060.0016
1.0040.0012
0.596
0.398
In the annular reactor core of small diameter with high neutron leakage value of
multiplication factor is very sensitive to disposition of fuel columns, and to ensure a power
distribution symmetry, while control rods in upper most position it makes sense to load the
fuel columns symmetrically throughout the core.
1.15
1.10
1.05
Keff
1.00
0.95
0.90
0.85
0.80
0.75
6
10
12
14
16
18
20
22
24
26
28

thin core, loading as per specification document
thin core, symmetric load (fig.4a)
thin core, symmetric load (fig.4b)
thick core, symmetric load (fig.5)
30 collumn core, symmetric load (fig.6)
MCNP-code
30
FIG. 2.77. keff versus number and arrangement of fuel columns.
123
keff = 0.758
keff = 0.770
keff = 0.938
keff = 0.944
keff = 1.028
keff = 1.028
a)
b)
FIG. 2.78. keff versus fuel columns arrangement (thin annular core 18 fuel columns).
124
keff = 0.945
keff = 1.038
keff = 1.035
keff = 1.079
keff = 1.108
FIG. 2.79. keff versus fuel columns arrangement (thick annular core 24 fuel columns).
125
keff = 0.924
keff = 1.039
keff = 1.125
FIG. 2.80. keff versus fuel columns arrangement (completely loaded core 30 fuel columns).
126
The Diffusion and Monte-Carlo calculations for the benchmarks HTTR-CR, HTTREX, HTTR-SC and HTTR-TC are provided in Tables 2-73, 2-74, 2-75 and 2-76, respectively.
2.2.6.3. Control Rod Position at Criticality [HTTR-CR]
Table 2-73. Results of Diffusion and Monte-Carlo Calculations for the HTTR-CR Benchmark
Problem
Characteristic
Organization
Depth of control rods insertion, cm
18 columns
24 columns
30 columns
OKBM
271
196
166
Value
IBRAE
RRC KI
259
195
170
306
201
154
2.2.6.4. Excess Reactivity [HTTR-EX]

Table 2-74. Results of Diffusion and Monte-Carlo Calculations for the HTTR-EX Benchmark
Problem
Characteristic
Organization
Reactivity excess, %k/k
18 columns
24 columns
30 columns
OKBM
2.68
9.73
11.14
Value
IBRAE
RRC KI
2.70
10.83
13.55
1.70
9.80
13.40
2.2.6.5. Scram Reactivity [HTTR-SC]

Table 2-75. Results of Diffusion and Monte-Carlo Calculations for the HTTR-SC Benchmark
Problem
Characteristic
Value
IBRAE
RRC KI b)
Organization
OKBM
Worth of reflector rods
1.02650.0010
1.00980.0015
keffkrit a)
1.0023
after drop
0.92050.0017
0.93490.0010
0.9242
keff
0,09550.0014
0.09610.0020
0.0843
(k/k)
Worth of all rods
1.02650.0010
1.00980.0015
1.0023
keffkrit
after drop
0.71720.0015
0.67460.0010
0.6573
keff
0.50810.0014
0.40400.0019
0.5237
(k/k)
a
Calculated value at critical position of rods obtained by experiment.
b
Results were obtained accordingly the JAERI-memo 10-005 including graphite impurity data.
127
2.6.6.6. Isothermal Temperature Coefficient [HTTR-TC]

Table 2-76. Results of Diffusion and Monte-Carlo Calculations for the HTTR-TC Benchmark
Problem
Characteristic
Organization
Multiplication factor of reactor for
design temperatures
k280
k300
k340
k380
k420
k460
k480
Temperature coefficients (k/k/K)
290
320
360
400
440
470
OKBM
Value
IBRAE
RRC KI
1.0023
0.9930
0.9844
0.9768
0.9699
0.9665
1.00320.0002
0.99540.0002
0.98860.0002
0.98220.0002
0.97540.0004
0.97260.0002
1.02790.0010
1.02320.0010
1.01600.0010
1.01240.0010
1.00520.0010
1.00260.0010
- (2.3310-4)
- (2.1910-4)
- (1.9710-4)
- (1.8210-4)
- (1.8110-4)
- (1.950.10)10-4
- (1.730.10) 10-4
- (1.650.10) 10-4
- (1.770.16) 10-4
- (1.480.32) 10-4
-(1.10,4)10-4
-(1.70,4)10-4
-(0.90,4)10-4
-(1.80,4)10-4
-(1.30,7)10-4
2.6.6.7 Accuracy analysis

The accuracy analysis of obtained results is presented as follows:
Table 2-77. Excess reactivity [HTTR-EX]
18 columns
Average value
1.6
(in accordance with RCM4 results), %k/k
a)
Average calculated value / experimental value,
%
WIMS-JAR value / average calculated value, %
+ 67
a)
WIMS-JAR value / experimental value, %
Monte - Carlo calculations
Average value
1.87
a)
%
MCNP value / average calculated value, %
+ 44
MCU value / average calculated value, %
-9
a)
MCNP value / experimental value, %
a)
MCU value / experimental value, %
a)
Experimental result is absent.
128
24 columns
30 columns
8.8
12.4
+ 14
+ 3.3
+ 10
+ 26
- 10
-7
10.1
13.0
+ 31
+ 8.3
+7
-3
+ 40
- 27
+ 4.2
+3
+ 13
+ 12
Table 2-78. Control rods worth [HTTR-SC]

Reflector control rods
All control rods
Average value
9.2
- 23
%
-8
- 30
Average value
%
48.0
+ 4.3
+9
+ 14
9.2
44.8
- 23
-2.6
+ 4.4
+ 3.8
- 20
- 20
- 10
+ 13
- 12
+ 10
Table 2-79. Isothermal Temperature Coefficient at the temperature range 345 407 K
[HTTR-TC]
Average value
(in accordance with RCM4 results), %k/kK
Average calculated value / experimental value, %
- 1.6310-4
Average value
(in accordance with RCM4 results), %k/kK
Average calculated value / experimental value, %
- 1.3210-4
+ 21
+ 28
+ 54
-2
+ 28
-4
+ 25
-4
Calculated
accuracy
analysis,
presented
above,
demonstrates
that
2-groups diffusion approximation gives significant error at the excess reactivity calculation
for annular core with one ring of fuel assemblies. Multigroups approximation is necessary for
calculation of this type core.
Large spread in excess reactivity values, obtained by MCNP and MCU, from average
calculated and experimental values demonstrates the necessity to revise calculational initial
data (it may be moisture content in graphite), on the one side, and also to verify nuclei data
library, on the other side.
129
Errors analysis of control rods worth estimation shows the necessity to use
multigroups diffusion approximation for calculations of the side reflector control rods worth
that is estimated with the most errors. Deviation of Monte-Carlo calculated control rods worth
from experimental one in the range about 20 % should be also referred to the effect of initial
data because deviation in calculated values is not large.
To estimate reactivity temperature coefficients, it is necessary additional analysis
characterized by accurate modeling of geometry and fuel particle distribution, detail modeling
of burnable poison position by Monte-Carlo method, and for diffusion calculations it is
necessary to model axial distribution of burnable poison, neutrons streaming for adequate
accounting of leakage and to use multigroups approximation.
Significant dependence of the results versus chosen nuclei data libraries and its
compiled programs defines the necessity to choice basic library for HTGRs calculation
especially for description of thermalization effects.
130
2.2.7 Turkey [2-63]

2.2.7.1 Introduction
This study was performed by the Nuclear Engineering Department of Haceettepe
University as a contribution to the IAEA CRP on Evaluation of HTGR Performance. MCNP4B has been utilized throughout this study. This is a multi-purpose Monte Carlo n-particle
(neutron, photon, and electron) transport code. This code also performs criticality analysis.
Reactor configuration
The HTTR core consists of vertically arranged prismatic hexagonal blocks with a
height of 580 mm and a width of 360 mm accross flats. Active core height and effective
diameter are 390 cm and 230 cm, respectively. Reactor core is made of 30 fuel colums and 7
control rod guide columns. The core is enveloped by 24 prismatic hexagonal reflector blocks.
Among them, there are 9 control rod guide columns, 12 replaceable reflector columns, and 3
irradiation colums. The whole structure is surrounded by a permanent graphite side reflector.
Each fuel column consists of 2 top reflector blocks, 5 fuel assemblies, and 2 bottom
reflector blocks. There are two types of fuel assemblies; containing 31 and 33 fuel rods. Each
fuel block contains three burnable poison (BP) insertion holes, but only two of them are filled
with B4C-C BP materials (containing 2.22 or 2.74 wt. %.natural boron). Fuel rods contain 14
annular fuel compacts. These compacts are manufactured with inner and outer diameters of
are 10 and 26 mm, respectively, and a height of 39 mm. UO2 coated particles are embedded
into graphite matrix to form the fuel compact.
Each assembly contains only one type of fuel. However, there are 12 different fuel
types with enrichments between 3 to10 wt. % of U-235.
Control rods are made of B4C pellets sorrounded by a 3.5 mm thick sleeve of Alloy
800H. The total height of control rods is 3094 mm. Inner and outer diameters are 65 mm and
113 mm, respectively. The detailed geometry and composition of all components are provided
by a JAERI report [2-5].
MCNP-4B model and simulations
MCNP-4B is a very versatile computer code for particle transport simulation purposes
[2-62]. It has extended features in geometrical modeling in three dimensional space. Every
single geometrical detail is defined to generate full HTTR core model. The complexity of the
reactor core structure makes this detailed modeling necessary. Horizontal cross sectional
views of the HTTR core model with MCNP-4B are given in Figures 2.81 and 2.82.
The primitive element of the model is a coated particle. It is generated with five
concentric spherical volumes. Then, these coated particles are uniformly distributed into a
hexagonal prismatic lattice which is enclosed by an annular cylindirical graphite. Thus, a fuel
compact model containing 13 000 CFPs is obtained. 14 fuel compacts are then stacked into a
structre made of graphite. Hence, a fuel assembly is constructed by placing this structure in a
hexagonal lattice. Two types of fuel assemblies with 31 and 33 fuel rods are generated in the
model as mentioned earlier.
131
Figure 2.81. Top cross sectional view of HHTR core in MCNP model.
Figure 2.82. Cross sectional view of HTTR core in MCNP model with fuel and control
assemblies.
Control rod guide colums are also modeled by leaving appropriate space for two
control rod locations and a location for reserve shut down rod system.
MCNP provides a facility to visualize the geometry described in the model. Hence, the
model is geometrically verified by taking views from different vertical and horizontal cross
sections.
Criticality calculations are performed with 5 000 source particles generated in each
cycle. 150 cycles are considered in each run. ENDF/B-VI cross section library with
continuous energy is utilized throughout the calculations. Cross sections are evaluated at
respective temperatures as necessary for graphite provided by thermal neutron libraries
TMCCS.
132
Figure 2.83. MCNP model for Coated Fuel Particles and their arrangement in fuel compact.
2.2.7.2. Results of benchmark problems
In this work, all benchmark problems expect HTTR-FC (Phase 2) are studied.
Benchmark Problem I:
HTTR-FC (Phase 1)
The number of fuel columns necessary to make the reactor critical is calculated in this
problem. Results of criticality calculations with different number of fuel colums are shown in
Table 2-80. The reactor is estimated to be critical with 15 fuel colums.
Table 2-80. Variation of effective multiplication
factor with loaded fuel colums
Fuel Colums
12
13
14
15
16
17
18
19
Benchmark Problem 2:
k-eff
0.98703
0.99252
0.99923
1.00503
1.01226
1.02079
1.03073
1.04698
Error
0.00102
0.00102
0,00103
0.00108
0.00103
0.00105
0.00097
0.00091
HTTR-CR
Control rod insertion depths for three different core loading configurations are
evaluated for criticality condition. Control rod positons are evaluated from the top of the 8th
layer (the top of the bottom reflector).
133
Table 2-81. Control rod depths for criticality and calculated k-eff
Fuel Columns Control Rod Depth (cm)
18
285
24
210
30
164
k-eff
0.99859
1.00864
0.99901
Error
0.0016
0.0013
0.00152
HTTR-EX
Excess reactivities are calculated for three different cases mentioned in Benchmark
Problem 2 assuming 300K for moderator and fuel temperatures.
Table 2-82. Excess reactivities for three different fuel l
oading conditions
HTTR-SC
There are two cases defined in this problem for evaluating scram reactivity. The first
one is specified with all reflector control rods are inserted when the reactor is critical. The
second one is evaluaeted with all control rods, reflector and core, are inserted again with the
critical configuration for fully loaded core.
Table 2-83. Scram reactivities with two different cases
Scram Reactivity (%)
All reflector CRs in
-7.75
All reflector & core CRs in
-37.96
HTTR-TC
Isothermal temperature coefficients are calculated for fully loaded core at the critical
condition. Critical control position is also evaluated at a temperature of 480K to be 190 cm..
Table 2-84. Critical control rod position at 480K
and corresponding k-eff
CR Position (cm)
190
134
k-eff
0.99838
Error
0.00220
As mentioned earlier cross sections for graphite at different temperatures are evaluated
by means of thermal neutron libraries TMCCS. There are only 4 sets of data at temperatures
of 300K, 600K, 800K, and 1200K available in our institution. Therefore, k-eff evaluations are
performed for these temperatures and the results are shown in Figure 2.84. As it can be seen
from the figure, the slope of the curve change with temperature. Therefore, it is rather difficult
to make an accurate prediction of isothermal temperature coefficient. However, it is estimated
for the first two datum points to be 1.20 10-4 1/K around the midpoint of 450K.
Figure 2.84. Effective multiplication factor as a function of temperature.

2.2.7.3 Conclusions
Results of HTTR start-up core physics benchmark problems are presented. Calculations are
performed by MCNP-4B code and with continuous energy cross sections. Phase 2 of the first
problem (HTTR-FC) is not taken into consideration. Therefore, all problems are performed
with the original data. Geometrical modeling is done such a way that geometrical details are
considered as much as possible. Isothermal reactivity calculation presented in this study is not
very reliable due to the lack of relevant cross section data at rather close temperatures.
135
2.2.8. United States of America

2.2.8.1. Computational Methodology [2-64]
The Monte Carlo code MCNP4a was used because the geometry of the coated fuel
particles, fuel compacts, burnable poisons, fuel element and reflector blocks, and the control
rods and guide blocks can be modeled explicitly without any approximations.
MCNP4a uses nuclear data from the ENDF/IBV Version 0 cross-section library. The
thermal neutron scattering matrix for the graphite is based on S(,) of ENDF/BV. A core
temperature of 300 K is assumed. MCNP4a uses track length, collision and analog estimators
in determining the most probable value of keff. The number of histories per batch is 10,000.
The number of batches is 30 with the first 5 batches being discarded for statistical reasons.
The core is modeled with hexagonal lattices consisting of control rod guide blocks,
fuel blocks, and permanent reflector blocks. The fuel rods and the burnable poison rods are
modeled using cylindrical body descriptions. The fuel block consists of several smaller
hexagonal lattices containing one fuel rod or burnable poison rod. This is shown in Figure
2.85. In the geometry description, all hexagonal lattices were assigned to a universe number.
The universe numbers were then input into an array to describe the location of the hexagonal
blocks in the reactor core. The cross sectional model of the 30 column reactor core is shown
in Figure 2.86. The outer boundary of the permanent reflector was modeled as a cylinder.
The fuel compacts are modeled with cylinders and the coated fuel particles (CFPS) are
modeled with spherical body descriptions. In the control rod insertion heterogeneous excess
reactivity calculations all five layers of the CFP are modeled. There are about 13,000 fuel
particles defined for each fuel compact corresponding to a packing density of 30 percent with
each fuel rod containing 14 fuel compacts. Therefore, for plotting simplicity the fuel particles
are not shown in the cross sectional view of the fuel hexagonal lattice. The fuel particle
representation is smeared in the homogeneous excess reactivity calculation.
The burnable poison rods are modeled as two different types. The geometry does not
change but the material composition of the burnable poison changes. There are graphite disks
that separate the upper and lower parts of the burnable poison rods. There are only two of the
three burnable poison rod positions filled for each fuel hexagonal lattice. The third hole
contains helium coolant. The configurations of the burnable poison rods change throughout
the reactor core and are shown in Figure 2.86.
There are three pairs of control rods in the outer ring that are not used in the approach
to criticality. All other control rod insertion depths change as one bank. The control rods in
the outer ring can only be removed to 72.5 cm from the top of the active core because a
sufficient negative reactivity addition rate must be available for reactor scrams. The control
rods in the inner rings can be fully withdrawn. Therefore, in the excess reactivity model, the
control rod holes contain coolant in the inner rings.
136
FIG. 2.85. Cross section of fuel assembly.
FIG. 2.86. Cross section of fully loaded core.
137
2.2.8.2. Results [2-65]

Monte Carlo methodology was used by the US in the initial HTTR criticality
benchmark problem (MCNP-4a), and the calculations were limited to the case of the full core
(30 elements loaded). The calculations made for the initial problem consisted of k-effective
calculations vs. control rod (C/R) position in cm from the bottom of the active core, evaluated
at two different uniform core temperatures (300K and 800K), and are shown in Figure 2.87
and detailed in Table 2.85.
FIG. 2.87: HTTR K-eff at 300 and 800K.

Table 2-85. MCNP4a Summary of results
Core
Temperature,
K
Critical Rod
Position (cm)*
K-eff,
(Rods Full In)
K-eff,
(Rods Full Out)
300
159
0.6899 (0.005)
1.1400 (0.004)
800
194
0.6525 (0.006)
1.1118 (0.005)
* Distance withdrawn (from bottom of the active core)
Homogeneous
Model,
K-eff,
(Rods Full Out)
1.1336 (0.005)
1.10598 (0.004)
As specified in JAERIs initial benchmark problem description, the outer ring of

reflector C/Rs were modeled as fully withdrawn, and all other rods were moved in tandem,
and their position with respect to the bottom of the active core denoted as C/R position.
There were no additional MCNP calculations done for Part 2 of the benchmark; hence
only simplified and approximate versions of the new benchmark problem calculations were
done using the original MCNP results. Hence for the scram reactivity estimates, scram from
critical results are for the case of all but the outer ring reflector rods inserting, rather than the
configurations specified in the new benchmark. The calculations of temperature coefficient of
reactivity also use just the original k-effective calculations at 300K and 800K rather than the
more detailed calculations specified in the new benchmark.
138
HTTR-SC
For the HTTR-SC benchmark problem, the scram reactivity from critical was
calculated from the expression noted in the benchmark definition (difference between keffective at critical [1.0] and k-effective with rods fully inserted divided by the product of the
two k-effective values). The results are shown in Figure 2.88 as a function of core
temperature (curve labeled Scram from Critical.). Hence the scram reactivity for the core
temperature specified in the benchmark (300K) is 0.45. The scram reactivity for the case
where all rods drop would be somewhat greater. Note that at higher temperatures, the scram
reactivity would be greater, as expected. Figure 2.88 shows a scram reactivity of 0.53 for a
uniform core temperature of 800K. Also shown in Figure 2.88 is the reactivity worth of the
rods (again excluding those rods in the outermost outer reflector ring) for full travel.
FIG. 2.88. Scram reactivity
FIG. 2.89. Temperature coefficient of reactivity vs. k-eff
139
HTTR-TC
The calculated temperature coefficient of reactivity (for uniform core temperatures) is
plotted vs. average k-effective in Figure 2.89. The point at which the curve intersects k-eff =
1.0 is taken as the value of interest. That value is 7.5E-5 dk/k per degree C, evaluated at the
average of the two temperatures at which k-eff was calculated (300K and 800K), i.e. 550K. It
is interesting to note that Figure 2.89 shows that the coefficient becomes considerably more
negative at smaller values of k-eff. One interpretation of this result would suggest that if the
shutdown reactivity is calculated using the conventional estimated temperature coefficient of
reactivity (at k-eff = 1.0), a non-conservative (low) value of shutdown reactivity would result.
140
REFERENCES TO CHAPTER 2
[2.1.] JAPAN ATOMIC ENERGY RESEARCH INSTITUTE, Present Status of HTGR
Research and Development, JAERI, Oarai, Japan, 1996.
[2.2.] YAMASHITA, K., A new HTR concept concerning a modular system with blocktype fuel elements and its comparison with a pebble-bed-type HTR, Jl-1842,
Kernforschungsanlage Jlich GmbH, 1983, (in German).
[2.3.] KUNITOMI, K., et. al., Development of New Type of HTGR, (Proc. 73rd JSME Fall
Annual Meeting), JSME, Japan, 1995.
[2.4.] NEYLAN, A.J., SILADY, F.A., and BAXTER, A.M., Gas Turbine Module Helium
Reactor (GT-MHR): A Multipurpose Passively Safe Next Generation Reactor,
Proceedings of The ASME/JASME 3rd International Conference on Nuclear
Engineering, ASME, 1995.
[2.5.] JAPAN ATOMIC ENERGY RESEARCH INSTITUTE, Benchmark Problems Data
for The HTTRs Start-up Core Physics Experiments JAERI-memo 10-005, Oarai,
Japan, 1998.
[2.6.] BREY, H.L., (Ed), Current Status and Future Development of Modular High
Temperature Gas Cooled Reactor Technology, IAEA-TECDOC-1198, Vienna, 2001
[2.7.] WEISBRODT, I.A., Status of Gas-Cooled Reactor Technology, Private
communication, Vienna, March 1997.
[2.8.] INTERNATIONAL ATOMIC ENERGY AGENCY, Design and Evaluation of Heat
Utilization Systems for the High Temperature Engineering Test Reactor, IAEATECDOC-1236, Vienna, August 2001.
[2.9.] FUJIMOTO, N., YAMASHITA, K., JAERIs Benchmark Calculation of HTTR Core
Physics-Phase 2, JAERI, (Proc. Second RCM of CRP on Evaluation of HTGR
Performance, held at Beijing), IAEA, Vienna, 1999
[2.10.] FUJIMOTO, N., et. al., JAERIs Benchmark Calculation of Start-up Core Physics
with HTTR Nuclear Characteristics Evaluation Code System, JAERI memo 11-030,
Oarai, Japan, pp. 80-106, 1999.
[2.11.] YAMASHITA, K., et. al., Nuclear design of the High-Temperature Engineering Test
Reactor (HTTR), Nucl. Sci. Eng. pp.122, 212-228, 1996.
[2.12.] SHINDO, R., YAMASHITA, K. and MUARTA, I., DELIGHT-7; One-dimensional
fuel cell burnup analysis code for High Temperature Gas-cooled Reactors (HTGRs),
JAERI-M 90-048, 1990, (in Japanese).
[2.13.] LATHROP, K. D. and BRINKLEY, F. W., TWOTRAN-II: and interfaced exportable
version of the TWOTRAN code for two-dimensional transport, LA-4848-MS, Los
Alamos, USA, 1973.
[2.14.] HARADA, H. and YAMASHITA, K., The reactor core analysis code CITATION1000VP for High Temperature Engineering Test Reactor, JAERI-M 89-135, 1989, (in
Japanese).
[2.15.] FOWLER, T. B., VONDY, D. R. and CUNNINGHAM, G. W., Nuclear reactor core
analysis code, CITATION, ORNL-TM-2496, Oak Ridge, USA, 1971.
[2.16.] MORI, T. and NAKAGAWA, M., "MVP/GMVP:General Purpose Monte Carlo
Codes for Neutron and Photon Transport Calculations based on Continuous Energy and
Multi-Group Methods", JAERI-Data/Code 94-007, 1994.
[2.17.] MORI, T., NAKAGAWA, M. and SASAKI, M., "Vectorization of Continuous
Energy Monte Carlo Method for Neutron Transport Calculation" , J. Nucl. Sci.
Technol., 29 [4], pp. 325, 1992.
141
[2.18.] SHIBATA, K., et al.: Japanese Evaluated Nuclear Data Library, Version-3,
JAERI-1319, 1990.
[2.19.] OKUMURA, K., et al., SRAC95; "General Purpose Neutronics Code System,"
JAERI-Data/Code 96-015, Japan Atomic Energy Research Institute, 1996.
[2.20.] MURATA, I., MORI, T. and NAKAGAWA, M., "Continuous Energy Monte Carlo
Calculations of Randomly Distributed Spherical fuels in High-Temperature GasCooled Reactors Based on a Statistical Geometry Model", NSE, 123, pp. 96-109, 1996.
[2.21.] N. NOJIRI, M. NAKANO, N. FUJIMOTO, K. YAMASHITA, M. TATEUCHI,
S.FUJISAKI, S. YOSHIMUTA, K. SAWA, M. SATO, M. ISHIHARA, T. IYOKU, H.
SAWAHATA, Benchmark problems data for the HTTRs start-up core physics
experiments, JAERI, Oarai, Japan, January, 1998
[2.22.] K. YAMASHITA, N. FUJIMOTO, M. NAKANO, U. OHLIG, IAEA Benchmark
calculation results of the HTTRs start-up core physics tests, Japan Atomic Energy
Research Institute memo 11-030, 1999
[2.23.] R. SANCHEZ, A. HEBERT, Z. STANKOVSKI, M. COSTE, S. LOUBIERE, C VAN
DER GUCHT and I. ZMIJAREVIC, APOLLO2 Twelve Years Later, Mathematics
and Computation, Reactor Physics and Environmental Analysis in Nuclear
Applications, Madrid, September, 1999
[2.24.] J.J. LAUTARD, S. LOUBIERE, C. MAGNAUD, CRONOS, A Modular
Computational System for Neutronic Core Calculations, (Proc. IAEA Specialists
Meeting), France, September, 1990
[2.25.] J.P. BOTH, Y. PENELIAU, The Monte Carlo code TRIPOLI4 and its first
benchmark interpretations, International Conference PHYSOR 96, Mito, Ibaraki,
Japan, September, 1996
[2.26.] INTERNATIONAL ATOMIC ENERGY AGENCY, Report of Second RCM of CRP
on Evaluation of HTGR Performance (held in Beijing, October 18-22, 1999), IAEA,
Vienna, 1999
[2.27.] N. FUJIMOTO, Data for re-calculation of HTTR-FC, Japan Atomic Energy
Research Institute, November, 2000
[2.28.] F. MOREAU, R. SANCHEZ, S. SANTANDREA, Contribution to the NEA C5G7
MOX benchmark, International Conference PHYSOR 2002, Seoul, South Korea,
October, 2002
[2.29.] PETROVIC, P. BENOIST, BN Theory: Advances and New Models for Neutron
Leakage Calculation, Advances in Nuclear Science and Technology, Plenum
Publishing Corporation, New York, vol. 24, 1996
[2.30.] P. BENOIST, J. MONDOT, I. PETROVIC, Calculational and experimental
investigations of void effect. A new model for leakage treatment of heterogeneous
assemblies, Nuclear Science and Engineering, 118, 197, 1994.
[2.31.] N. NOJIRI, N. FUJIMOTO, K. YAMASHITA, HTTRs Benchmark Calculation of
Start-up Core Physics Tests, JAERI, Japan, 1998.
[2.32.] N. FUJIMOTO, K. YAMASHITA, M. NAKANO, M. TATEUCHI, S.FUJISAKI,
Experimental Results of HTTRs Start-up Core Physics Tests, JAERI, (Proc. Second
RCM of CRP on Evaluation of HTGR Performance, Beijing, October 2000), IAEA,
Vienna, 2000.
[2.33.] N. FUJIMOTO et al., Control Rod Position and Temperature Coefficients in HTTR
Power-Rise Test, JAERI Tech 2000-091 (2000).
[2.34.] NOJIRI, N., et al., Private Communication, JAERI, Japan, (Jan. 1998)
[2.35.] PHILIPPEN, P., Private Communication, (1994)
142
[2.36.] GREENE, N.M., et al., AMPX-77: A Modular Code System for Generating Coupled
Multigroup Neutron-Gamma Cross-Section Libraries from ENDF/B-IV and/or
ENDF/B-V, Oak Ridge National Laboratory, ORNL/CSD/TM-283, (Oct. 1992)
[2.37.] TEUCHERT, E., HAAS, K.A., ZUT-DGL-V.S.O.P.:Programmzyklus fr die
Resonanzabsorption in heterogenen Anordnungen, FZ-Jlich, Interner Bericht IRE-701, (1970)
[2.38.] BROCKMANN, H., TOTMOS: An Integral Transport Code for Spectrum
Calculations, FZ-Jlich, ISR, Germany, (Dec. 1995)
[2.39.] RHOADES, W.A., et. al., DORT-TORT, Two- and Three-Dimensional Discrete
Ordinates Transport Code, Version 5.13.14, ORNL CCC-543, Oak Ridge, USA, (Jan.
1992)
[2.40.] FOWLER, T.B., et al., Nuclear Reactor Core Analysis Code: CITATION, ORNLTM-2496, Oak Ridge, USA, Rev.2, (July 1971)
[2.41.] YANG, CH., BENOIST, P., Scattering anisotropy and neutron leakage in reactor
lattices, Nucl. Sci. Eng., 86, pp. 47-62, (1984)
[2.42.] FUJIMOTO, N., Private Communication, JAERI, Japan,(Nov. 2000)
[2.43.] FUJIMOTO, N., Private Communication, JAERI, Japan, (March 2001)
[2.44.] BROCKMANN, H., OHLIG, U., Calculational Results for the Benchmark Problems
of the HTTR Start-up Core Physics Experiments, ISR, FZJ Research Centre-Jlich,
Germany, October 15, 2001.
[2.45.] AZIA, F., LASMAN, A. N., and BAKRI, A., Results of Benchmark Calculation on
Start-up Core Physics of High Temperature Engineering Test Reactor, (Presented at
the First RCM of CRP on Evaluation of HTGR Performance, Vienna, August 1998),
BATAN, Indonesia, 1998.
[2.46.] RONEN, Y., LEIBSON, M. J., Nuclear Technology, 80, pp. 216-224, 1988.
[2.47.] NOJIRI, N., et al., Benchmark Problems Data for the HTTR Start-up Core Physics
Experiments, JAERI Memo 10-005, 1998.
[2.48.] KUGO, T., TSUCHIHASHI, K., TAKANO, H. and AKIE, H., An EWS Version of
SRAC-Code: SRAC-EWS, JAERI, 1994.
[2.49.] JEONG, C. J., OKUMURA, K., ISHIGURO, Y. and TANAKA, K., Accuracy of Cell
Calculation Methods Used for Analysis of High Conversion Light Water Reactor
Lattice Journal Nuclear Science & Technology, 27, pp. 515-523, 1990.
[2.50.] DE HAAS, J.B.M., WALLERBOS, E.J.M., Analysis of the HTTR with Monte-Carlo
and Diffusion Theory, An IRI, ECN Intercomparison, ECN-I-98-056, the Netherlands,
September 1998.
[2.51.] WALLERBOS, E.J.M., HOOGENBOOM, J.E., IRI Results for the benchmark
problems of start-up core physics of the high temperature engineering test reactor
(HTTR), IRI-131-98-007, Interfaculty Reactor Institute, Delft, 1998.
[2.52.] WALLERBOS, E.J.M., HOOGENBOOM, J.E., and VAN DAM, H., IRI results for
LEU HTR-PROTEUS cores 5, 7, 9 and 10, IRI-131-97-005, Interfaculty Reactor
Institute, Delft, 1997
[2.53.] DE HAAS, J.B.M, TURKCAN, E., HTTR Criticality, Physical Parameters
Calculations and Experimental Results, NRG/IRI, the Netherlands, Presented at the
ICENES-2000 Meeting, 2000.
[2.54.] ASKEW, J.R., et.al., WIMS-D/4, A general description of the lattice code WIMS code for unit cell calculation and preparation of few energy group macrosections.
ENDF/B6 nuclear cross sections library is used. JBWES, pp. 564, Oct. 1966.
[2.55.] ZIZIN, M., SHISHKOV, L., YAROSLAVTSEVA, L., JAR, Test neutron-physics
calculations of nuclear reactors. - code for 3D reactor calculation in few energy group
approximation , Moscow, Atomizdat, pp. 9, 1980.
143
[2.56.] LIMAN, G., MAYOROV, L., YUDKEVICH, M., MCU, Code package for solution of
radiation problems in reactors by Monte Carlo method. Collected volume: Questions of
Atomic Science and Technique, Series: Physics and Technique of Nuclear Reactors,
Iss.7, p.27, 1985 - code for unit cell and 3D reactor calculations in multi energy group
approximation. The original cross section data library collection is used: ABAGIAN,
L., ALEKSEEV, N., BRYZGALOV, V., et al., MCU-PFFI A Software with
DLC/MCUDAT-1.0 Nuclear Data Library. Certificate number of PC 61 of Russian
Federal Agency for Supervisions and Radiation Safety, Moscow, 1996
[2.57.] MCNP, A General Monte Carlo Code for Neutron and Photon Transport, LA-7396-M,
1981 - code for 3D reactor calculations. Point by point interactions of isotopes with
neutrons required by MCNP were prepared with the help of NJOY program package:
JOY Nuclear Data Processing System, LA-9303-M, 1982 - on the basis of ENDF/B6
format. MCNP- 4A option was used, Los Alamos, USA, 1981.
[2.58.] SEGEV, M., An Equivalence Relation for Doubly Heterogeneous Lattice, Nucl. Sci.
Eng. 81, pp.151-160, 1982.
[2.59.] KUZAVKOV, N., MAROVA, E., SUKHAREV, YU., (OKBM), MITENKOVA, E.,
NOVIKOV, N., (IBRAE), BRYZGALOV, V., GLUSHKOV, E., GOMIN, E., (RRC
KI), HTTRs Start-up Core Benchmark Physics Calculation, JAERI-memo 11-030,
March 1999
[2.60.] GLUSHKOV, E., GOMIN, E., BRYZGALOV, V., LOBYNTSEV, V., (RRC KI),
Benchmark Problems of HTTR Core Physics (Phase 2). Monte-Carlo Calculation
Results, (Proc. Second RCM of CRP on Evaluation of HTGR Performance, Beijing,
1999), IAEA, Vienna, 1999.
[2.61.] KUZAVKOV, N., MAROVA, E., SUKHAREV, YU., (OKBM), MITENKOVA, E.,
NOVIKOV, N., (IBRAE), Benchmark Problems of HTTR and HTR-10 Core PhysicsPhase 2, (Proc. Second RCM of CRP on Evaluation of HTGR Performance, Beijing,
1999), IAEA, Vienna, 1999
[2.62.] LOS ALAMOS NATIONAL LIBRARY, MCNP- A General Monte Carlo Particle
Transport Code, LA-1265-M, Los Alamos, USA, 1997.
[2.63.] SEKER, V., COLAK, U., KADIROGLU, O.K., Results of HTTR Start-up Core
Physics Benchmark Problems, Hacettepe University, Ankara, Turkey, 2001
[2.64.] BENTZINGER, D. and BALL, S., ORNLs Benchmark Calculation of the HTTRs
Startup-Core Physics with Continuous Energy Monte Carlo Code MCNP4a, JAERI,
Memo 11-030, ORNL, USA, March 1999
[2.65.] BENTZINGER, D. and BALL, S., HTTR Benchmark Part 2, Temperature
Coefficient of Reactivity and Scram Reactivity Calculations, ORNL, USA, November,
1999.
144
Chapter 3
HIGH TEMPERATURE ENGINEERING TEST REACTOR
THERMAL HYDRAULIC BENCHMARKS
3.1. THERMAL HYDRAULIC BENCHMARK INFORMATION
This chapter includes a review of Member State investigation into the HTTR related
thermal hydraulic benchmark problems of vessel cooling (VC) at 30 MW and the transient
behaviour of loss of off-site electric power (LP) at 15 and 30 MW.
3.1.1. Description of HTTR Systems Related to Vessel Cooling and Loss of Power
Included in this section are specific descriptions of the HTTR systems associated with
the vessel cooling and loss of off-site electric power benchmark problems. A description of
the general HTTR facility has been provided herein as Section 2.1.1.
3.1.1.1 HTTR Vessel Cooling [3-1]
This benchmark problem includes predicting the heat removal capacity of the vessel
cooling system (VCS) at 30 MW power operation and the temperature profile on the surface
of the side panel. These predictions are to be compared with the actual measured heat
removal capacity of the VCS and the associated temperature profile.
Table 3-1 shows
reference operation conditions at 30MW power operation.
Table 3-1. 30MW operating conditions of the HTTR
Coolant temperature at reactor inlet/outlet
Mass flow rate
Thermal power
Water inlet temperature of VCS
395C/850C
12.4kg/s
30MW
4MPa
25C
Vessel cooling system description

The reactor and VCS are installed in the reactor cavity surrounded by the primary
radiation shielding. The VCS consists of upper, side and lower panels. It has two completely
independent sets of cooling systems to assure minimum heat removal to keep fuel and RPV
integrity during depressurization accident. They are designated as A and B systems.
Cooling panels are located on the inner surface of the reactor cavity as shown in Figure 3.1.
Figures 3.2 and 3.3 represent the heat transfer schematic and expected cooling panel flow
path, respectively. Table 3-2 includes VCS specifications.
145
FIG. 3.1. HTTR cooling panel schematic.

Table 3-2. Major Specifications of the VCS
Number of set
Minimum requirement for heat removal
Nominal water flow rate
Total
Lower panel
Side panel
Bottom panel
Allowable temperature for components
2 set (A & B systems)

0.3MW
86 ton/hour/each set
6.5 ton/hour/each set
72 ton/hour/each set
6.5 ton/hour/each set
90
The upper panel is on the lower surface of the upper radiation shielding as shown in
Figure 3.1. It consists of 96 cooling tubes in a steel casing. Each system has 48 cooling tubes.
Water is distributed to six inlet headers from the inlet ring header and flows to the outlet ring
header through cooling tubes and outlet headers. As an example, the flow diagram is depicted in
Figure 3.2. Flow rate of the upper panel is 6.5 t/h and average flow rate per cooling tube is
3.76e-2 kg/s as shown in Table 3-3. There are 31 penetrations for the standpipes in the upper
panel.
146
FIG. 3.2: Schematic of heat transfer from RPV to side panel

Table 3-3. Water flow rate for each independent cooling system
Side panel
Flow rate
(Total)
72 t/h
Lower
panel 6.5 t/h
(side)
Upper panel
6.5 t/h
Lower
panel 6.5 t/h
(bottom)
Number of Outer diameter

tubes
Thickness
108
25.4mm
3.5 mm
96
25.4mm
3.5 mm
48
15.9mm
3.2 mm
8 or 3
25.4mm
3.2 mm
Flow rate
(Each tube)
1.85e-1 kg/s
3.76e-2 kg/s
3.76e-2 kg/s
(See Table 4)
The side panel consists of twelve units. Water flows into each unit through a ring header,
removes heat and goes out through another ring header. A reference unit of the side panel has a
total of 18 cooling tubes, 9 for the A system and 9 for the B system. Flow rate for the side
panel is 72 t/h and average flow rate for each tube is 1.85e-1kg/s. A steel plate connects adjacent
two tubes. There are two thermal reflector plates between EL.17.35M and EL.30.99M in the side
panel (Table 3-4). They are stainless steel and carbon steel plates. Two more carbon steel plates
exist between EL.19.025M and EL.27.175M. There are nine horizontal gaps to absorb axial
thermal expansion of the thermal reflector plates. These provide flow passages for air natural
circulation in the side panel. The skirt panel on the RPV skirt removes heat from the RPV skirt
directly to prevent concrete heat up under the RPV skirt.
147
The lower panel is on the bottom and side surface of a space under the RPV as shown in
Figure 3.1. Water flows through the lower panel (bottom) and lower panel (side) in turn and
returns to the water cooler. The bottom part is similar to the upper panel and its cooling tubes are
installed between two plates. It has four headers and that are connected with 3 or 8 cooling tubes.
In the A system, water comes from the water cooler flows through cooling tubes from header 1
to header 4 and goes out to the inlet ring header of the lower panel (side). Average flow rates are
shown in Table 3-4. In the B system, water flows from the header 4 to header 1. The lower
panel (side) is similar to the side panel and consists of twelve units. Each unit has 16 cooling
tubes, 8 for the A system and 8 for the B system. Water is distributed to each unit from the
inlet ring header. The water then goes upward through 4 cooling tubes, and returns to the outlet
ring header through the other 4 cooling tubes. Average flow rates are shown in Table 3-4.
FIG. 3.3. Assumed flow passage in side panel

Table 3-4. Number of cooling tubes of lower cooling panel (bottom)
Total
Between Header 1 and Header 2 16
148
A system
B system
8
3
8
8
3
8
Average
flow rate of
one tube
2.25e-1 kg/s
6.02e-1 kg/s
2.25e-1 kg/s
Table 3-5. Thermal radiation reflector plate

Panel
Side panel
Numbers
3 (Carbon steel)
1(Stainless steel)
1 (Carbon steel)
1 (Stainless steel)
1 (Carbon steel)
1 (Stainless steel)
Lower panel (side)
Elevation
From EL.19.025M to EL.27.175M
There are four ring plates of 6 mm thickness in reactor cavity. Their elevations are
EL.27.6M, EL.25.45M, EL.22.2M and EL.19.0M. They separate the reactor cavity into five
spaces. A ring plate is also in the space under the RPV. Its elevation is 15.15M.
The material properties, thermal conductivity and emissivity of components, the
correlation of heat transfer coefficient of helium and power distribution in the core are as follows:
(1) 2.25Cr-1Mo steel (RPV, standpipe, core support gird, core support plate)
Thermal conductivity T < 282.22(C)
282.22(C) < T < 676.67(C)
676.67 (C) < T
Emissivity
0.8
32.0(kcal/mhC)
-7.861e-3T + 36.245 (kcal/mhC)
-1.459e-2T + 44.652 (kcal/mhC)
(2) Stainless steel (Casing of shielding block, thermal shield, thermal reflector)
Thermal conductivity
Emissivity
0.0107T + 9.888 (kcal/mhC)

0.48
(3) Carbon steel (Thermal reflector, ring plate)

Table 3-6. Thermal conductivity of Carbon Steel
Temperature (C) Thermal
conductivity
(kcal/mhC)
0
44.6
100
43.6
200
41.4
300
42.8
400
36.0
149
Emissivity
0.8
(4) PGX blocks (Permanent reflector, hot plenum, lower plenum, bottom)
Table 3-7. Thermal conductivity of PGX graphite
Thermal
Temperature (C) (kcal/mhC)
Radial
100
100.74
200
92.28
300
84.62
400
77.71
500
71.54
600
66.01
700
61.18
800
56.89
900
53.22
1000
50.06
1100
47.39
1200
45.14
1300
43.37
1400
41.95
1500
40.82
1600
40.04
Emissivity
Conductivity
Axial
74.52
67.54
61.34
55.87
51.05
46.87
43.31
40.25
37.66
35.5
33.73
32.33
31.18
30.28
29.59
29.02
0.8
(5) IG-110 (Core components)

Table 3-8. Thermal conductivity of IG-110 graphite
Temperature (C) Thermal
conductivity
(kcal/mhC)
400
68.15
600
56.1
800
47.75
1000
42.15
1200
38.35
1400
35.75
Emissivity
150
0.8
(6) ASR-0RB (Carbon block)

Table 3-9. Thermal conductivity of Carbon block
Temperature
(C)
100
200
300
400
500
600
700
800
900
1000
1100
Thermal
Conductivity
(kcal/mhC)
7.452
7.884
8.244
8.532
8.82
9.036
9.216
9.36
9.468
9.54
9.612
(7) B4C/C (Neutron absorber)

Thermal conductivity
16.2 (kcal/mhC)
(8) Correlation of heat transfer coefficient in the annual flow passage in the reactor
Nu = 0.023 Re 0.8 Pr 0.4 (Di/Do) 0.16 (Tw/Tb) 0.5
Where;
Nu
Re
Pr
Di
Do
Tw
Tb
Nusselt number
Reynolds number
Prandtl number
Inner diameter of annual flow passage
Outer diameter of annual flow passage
Wall temperature
Bulk temperature
(9) Power distribution in the core

Table 3-10. Power distribution
Axial fuel layer
1
2
3
4
5
Fuel zone number

1
0.031944
0.060489
0.060698
0.033081
0.021392
2
0.033058
0.062328
0.057048
0.034271
0.022164
3+4
0.096679
0.170170
0.158590
0.095860
0.062224
151
3.1.1.2 HTTR loss of off-site power event description [3-3]

The benchmark problems concerning loss of off-site electric power simulation are
designated as HTTR-LP (15MW, 30MW). During normal HTTR operation (referred as parallel
loaded operation), the intermediate heat exchanger, primary and secondary pressurized water
coolers are operated simultaneously. In HTTR-LP (15MW), analytical simulation of transient
behavior of the reactor and plant during the loss of off-site electric power from normal operation
under 15MW thermal power is evaluated. Similar evaluation of transient behavior of the reactor
and plant during the loss of off-site electric power from the normal operation under 30MW
thermal power is conducted as HTTR-LP (30MW). In both simulation cases, estimation items are
as follows; transition of (1) hot plenum block temperature, (2) reactor inlet coolant temperature,
(3) reactor outlet coolant temperature, (4) primary coolant pressure, (5) reactor power, (6) heat
removal of auxiliary heat exchanger. Estimation duration is for 10hr. from the beginning of the
loss of off-site electric power.
__________________________________________________________
(1) Loss of off-site electric power of HTTR caused by failure of
power transmission line or HTTR electrical equipment
(2) Coast down of all primary and secondary helium circulators as well as water pump
for pressurized water cooling system
(3) Decrease of flow rate of primary and secondary helium as well as water
5s after (1)
(4) Decrease to 92% of total primary helium flow rate of intermediate heat exchanger
3.2s after (4)
(5) Reactor scram by reactor protection system
1s after (5)
(6) Initiation of breaking stop of all primary and secondary helium circulators
50s after (1)
(7) Startup of auxiliary cooling system by supplying electricity through emergency power feeder
40min after (7)
(8) Coast down of one of two auxiliary helium circulators
Fig. 3.4. Event scenario for the analyzed transient of loss of off-site electric power from 30MW.
152
The loss of off-site electric power is caused by failure of the power transmission line or
the HTTR electrical equipment. All primary and secondary helium circulators as well as the
water pump for the pressurized water cooling system coast down immediately after the loss of
off-site electric power. Accordingly, flow rates of primary and secondary helium as well as water
reduce.
Figure 3.4 depicts the event scenario for the analyzed transient during the loss of off-site
electric power from 30MW. The flow rate of primary helium, which is deflected at a hot header
and discharged around the heat transfer tubes of the intermediate heat exchanger, decreases to
92% of normal flow rate in 5s after the loss of off-site electric power. In 3.2s after 92% flow rate
is reached, the reactor is scrammed by the reactor protection system. In 1s after the reactor scram,
breaking stop of all the primary and secondary helium circulators is initiated. The auxiliary
cooling system starts up in 50s after the loss of off-site electric power by electricity supplied
through the emergency power feeder. The auxiliary cooling system mainly consists of an
auxiliary heat exchanger, two auxiliary helium circulators, air cooler and two water pumps.
Helium flow rate of the auxiliary cooling system reaches about 1.2kg/s in 20s after the startup of
the auxiliary cooling system. One of the two auxiliary helium circulators is coasted down in
40min after the startup of the auxiliary cooling system to prevent the graphite blocks, composing
the reactor core, from overcooling. Then helium flow rate of the auxiliary cooling system
decreases to about 0.8kg/s.
HTTR plant conditions at 15MW and 30MW
Initial conditions at 15MW and 30MW power before startup of auxiliary cooling system
are shown in Table 3-11. These include the reactor core (thermal power, reactor inlet and outlet
temperatures, primary coolant pressure and flow rate as well as hot plenum block temperature).
Tables 3-12, 3-13, 3-14 and 3-15 show the major process conditions (temperature, pressure and
flow rate) of the intermediate heat exchanger, primary and secondary pressurized water coolers as
well as the air cooler for the pressurized water cooling system, respectively. Helium and water
flow rates of the auxiliary cooling system in standby are fixed to 0.036kg/s and 5.5kg/s,
respectively, to prevent the auxiliary heat exchanger from the thermal shock. During the standby
of the auxiliary cooling system, the two auxiliary helium circulators are not operated and one of
the two water pumps for the auxiliary cooling system is driven. Tables 3-16 and 3-17 show the
major process conditions (temperature, pressure and flow rate) of the auxiliary heat exchanger
and air cooler for the auxiliary cooling system, respectively.
Table 3-11. Major process conditions of reactor core
Items
Thermal power
Reactor inlet coolant temperature
Reactor outlet coolant temperature
Primary coolant flow rate
Hot plenum block temperature
HTTR-LP
(15MW)
15MW
About 241C
About 470C
About 3MPa(abs)
12.4kg/s
About 490C
HTTR-LP
(30MW)
30MW
395C
850C
4MPa(abs)
12.4kg/s
About 890C
153
Table 3-12. Process conditions of intermediate heat exchanger

Items
Primary helium inlet temperature
Primary helium outlet temperature
Primary helium pressure
Primary helium flow rate
Secondary helium inlet temperature
Secondary helium outlet
temperature
Secondary helium pressure
Secondary helium flow rate
HTTR-LP
(15MW)
About 468C
About 238C
About 3MPa(abs)
4.1kg/s
About 154C
About 431C
HTTR-LP
(30MW)
850C
395C
4MPa(abs)
4.1kg/s
About 241C
About 783C
About 3.1MPa(abs)
3.6kg/s
4.1MPa(abs)
3.6kg/s
Table 3-13. Process conditions of primary pressurized water cooler

Items
Helium inlet temperature
Helium outlet temperature
Helium pressure
Helium flow rate
Water inlet temperature
Water outlet temperature
Water pressure
Water flow rate
HTTR-LP
(15MW)
About 468C
About 242C
About 3MPa(abs)
8.3kg/s
About 89C
About 110C
About 2.6MPa(abs)
115kg/s
HTTR-LP
(30MW)
850C
395C
4MPa(abs)
8.3kg/s
About 135C
About 175C
3.5MPa(abs)
115kg/s
Table 3-14. Process conditions of secondary pressurized water cooler

Items
Helium pressure
Helium flow rate
Water pressure
Water flow rate
154
HTTR-LP
(15MW)
About 430C
About 154C
About 3.1MPa(abs)
3.6kg/s
About 89C
About 110C
About 2.5MPa(abs)
60kg/s
HTTR-LP
(30MW)
About 782C
About 240C
4.1MPa(abs)
3.6kg/s
About 135C
About 175C
3.4MPa(abs)
60kg/s
Table 3-15. Process conditions of air cooler for pressurized water cooling system
Items
Water pressure
Water flow rate
Air inlet temperature
Air outlet temperature
Air pressure
Air flow rate
HTTR-LP
(15MW)
About 110C
About 58C
About 2MPa(abs)
70kg/s
About 33C
About 58C
0.1MPa(abs)
605kg/s
HTTR-LP
(30MW)
About 175C
About 80C
3MPa(abs)
70kg/s
About 33C
About 81C
0.1 MPa(abs)
605kg/s
Table 3-16. Process conditions of auxiliary heat exchanger

Items
Helium pressure
Helium flow rate
Water pressure
Water flow rate
HTTR-LP
(15MW)
About 452C
About 50C
About 3MPa(abs)
0.036kg/s
About 35C
About 39C
About 2MPa(abs)
5.5kg/s
HTTR-LP
(30MW)
About 812C
About 62C
4MPa(abs)
0.036kg/s
About 36C
About 45C
About 3MPa(abs)
5.5kg/s
Table 3-17. Process conditions of air cooler for auxiliary cooling system
Items
Water pressure
Water flow rate
Air pressure
Air flow rate
HTTR-LP
(15MW)
About 39C
About 35C
About 1MPa(abs)
5.5kg/s
About 33C
About 34C
0.1MPa(abs)
106.5kg/s
HTTR-LP
(30MW)
About 45C
About 36C
About 2MPa(abs)
5.5kg/s
About 33C
About 42C
0.1 MPa(abs)
106.5kg/s
155
Conditions after startup of auxiliary cooling system include helium flow rate of the
auxiliary cooling system reaches about 1.2kg/s in 20s after the startup of the auxiliary cooling
system. Helium flow rate of the auxiliary cooling system decreases to about 0.8kg/s in 40min
after the startup of the auxiliary cooling system. Water flow rate of the auxiliary cooling system
is fixed to about 18.3kg/s all the time after the startup of the auxiliary cooling system.
3.1.2. HTTR Thermal Hydraulic Benchmark Problem Descriptions
Two sets of thermal hydraulic benchmark problems associated with the HTTR are investigated in
this section. These include:
The prediction of the amount of heat removed by the Vessel Cooling System (VCS) at 30
The analytical simulation on transient behavior of the reactor and plant during the loss of
off-site electric power for HTTR operation at 15 and 30 MW.
MW power and the associated temperature profile on the surface of the side panel, and
The calculations formulated by the participating Member States will then be compared
with actual experimental values achieved during testing on the HTTR.
156
3.2 THERMAL HYDRAULIC BENCHMARK ANALYSIS/RESULTS

3.2.1. Japan
3.2.1.1. HTTR-VC Test and analytical results of heat removal capability [3-1]
Heat removal of the Vessel Cooling System (VCS) was calculated using the
SSPHEAT code which was developed to analyze the temperature distribution in the in-core
structure test section (T2) of HENDEL with complicated passages of helium flow [3-2]. The
helium flow was simulated by a thermal-flow element, which is a uniaxial element in threedimensional space with the ability to conduct heat and transmit fluid between its nodal points.
The element has two parameters, temperature and pressure, at each nodal point. In the
computer code, the model is solved by the finite element method (FEM).
Analytical conditions for full power operation are shown in Table 2-18. Reactor
power, reactor inlet temperature and flow rate are 30MW, 395C and 12.4kg/s, respectively.
Heat removal of 0.3MW is minimum requirement for the VCS to keep fuel and RPV integrity
at the accident conditions.
Table 3-18. Analytical conditions and result at full power

Reactor power
30 MW
Reactor inlet temperature
395 degree C
Coolant flow rate in reactor
12.4kg/s
Cooling water temperature of VCS
25 degree C
Reactor outlet temperature
850 degree C
VCS heat removal
0.77MW
The calculated VCS heat removal at full power of 30MW is 0.77MW. The analysis
shows the VCS satisfies this requirement. The calculated VCS heat removal is larger than the
maximum design value of 0.6MW. It is because air natural circulation occurs in the side
panel and the air transfers heat to cooling tubes through thermal reflector plates which are
provided to reduce radiant heat from the reactor.
Figure 3.5 shows the temperature profile for the HTTR-VC analysis and actual
measurements at 9MW power. In a comparison of test and analytical results at 9 MW
operation of HTTR the VCS heat removal at steady state was ~ 0.22MW. Table 3-19 shows
the test and analytical conditions. Table 3-20 and Figure 3.5 show the comparison of test and
analytical results at 9MW power.
157
Table 3-19. Test and analytical condition at 9MW operation

Reactor power
9 MW
175 degree C
Coolant flow rate in reactor
12.4kg/s
28.5 degree C
312 degree C
Calculated heat removal with no air circulation in the side panel was 0.13MW and
calculated temperatures on the side panel do not agree with the measured temperatures.
Calculated heat removal increases to 0.2MW when taking into consideration the air
circulation in the side panel and temperature distributions agree well with the measured
temperatures as shown in Figure 3.5.
Table 3-20. Comparison of test and analytical results
at 9MW operation
Test
Calculation
(No air circulation in side panel)
Calculation
(Air circulation in side panel)
0.22 MW
0.13 MW
0.20 MW
Table 3-21 shows the test condition and result at 30MW operation. Reactor inlet and
outlet temperatures and coolant flow rate are 392C, 844C and 12.3kg/s, respectively. The
VCS heat removal reached about 0.81MW which compares favorably with the analytical
estimate of 0.77MW. Figure 3.6 shows the temperature profile on the RPV and side panel.
Test results showed that RPV temperatures at EL.19-27m are 340~360C and about 20C
lower than the analytical results. Side panel temperatures also showed that test results are
lower than analytical results.
Table 3-21. Test condition and result at 30MW operation
158
Reactor power
30MW
392C
844C
Coolant flow rate
12.3kg/s
33.2C
Test
0.81 MW
FIG 3.5. Temperature profile for the HTTR-VC analysis and actual measurements at 9MW power.
159
FIG. 3.6. Temperature Profiles on the RPV and Side Panel at 30 MW.
3.2.1.2. HTTR-LP Test Results
Pre-estimation results by the ACCORD code [3-4]
The pre-estimation results of the benchmark problems concerning the loss of off-site
electric power simulation of the HTTR by the ACCORD code include the transition of the
hot plenum block temperature, reactor inlet and outlet coolant temperatures, primary coolant
pressure, reactor power and heat removal of the auxiliary heat exchanger. The estimation
duration is for 3600s from the beginning of the loss of off-site electric power.
The analyzed transient of loss of off-site power from 30MW has been described
previously in Section 3.1.1.2. The maximum heat removal of the auxiliary heat exchanger
during the operation of the two auxiliary helium circulators under the loss of off-site electric
power from 15 and 30MW will be approximately 1.5 and 3MW, respectively.
Results of loss of off-site electric power tests from 15 and 30 MW operations [3-4]
The transient experienced upon tests of the loss of off-site electric power from 15 and 30 MW
operations are shown in Figures 3.7 and 3.8, respectively. All of the primary and secondary
helium circulators as well as the water pumps for the pressurized water, and the auxiliary and
the vessel cooling systems coasted down immediately after the loss of off-site electric power.
In 1.6 s after the loss of off-site electric power, the reactor scrammed by the
160
161
At the test from 15 MW operation, all the control rods were inserted simultaneously into the
reactor core by gravity within the design criterion of 12 s. At the test from 30 MW operation,
the outer nine pairs of the control rods were inserted into the replaceable reflector region of the
core within 12 s at first, and 40 min later the other inner seven pairs fell into the fuel region
within 12 s.
In about 2 s after the reactor scram, braking of all the primary and secondary helium
circulators was initiated by flowing electric current from their associated batteries. The
braking stop times, when the rotations of the helium circulators became to zero, were 10 s or
less after the open of the reactor scram breaker.
In about 50 s after the loss of off-site electric power, the auxiliary cooling system and
the vessel cooling system started up by supplying electricity from two emergency power
feeders. In 40 min after the startup of the auxiliary cooling system, one of the two auxiliary
helium circulators stopped by design to reduce thermal stresses of the core graphite
components.
Table 3-22 shows the thermal power, the reactor inlet and outlet coolant
temperatures, the primary coolant pressure and flow rate as well as the hot plenum block
temperature before the loss of off-site electric power tests from 15 and 30 MW operations.
Temperatures of the hot plenum blocks are measured at representative three points. Because
no significant temperature difference among the hot plenum blocks appeared, mean
temperature among the hot plenum blocks was used.
Tables 3-23, 3-24, 3-25 and 3-26 show the temperature, pressure and flow rate of the
intermediate heat exchanger (IHX), the primary and secondary pressurized water coolers as
well as the air cooler for the pressurized water cooling system before the loss of off-site
electric power tests from 15 and 30 MW operations, respectively.
Table 3-22. Conditions of reactor before LP tests
Items
Thermal power
Reactor inlet coolant temperature
Reactor outlet coolant temperature
Primary coolant flow rate
Hot plenum block temperature
HTTR-LP (15MW)
15MW
242C
468C
2.9MPa(abs)
12.4kg/s
391C
HTTR-LP (30MW)
30MW
392C
828C
4.0MPa(abs)
12.4kg/s
677C
Table 3-23. Conditions of IHX before LP tests

Items
Primary helium inlet temperature
Primary helium outlet temperature
Primary helium flow rate
Secondary helium inlet temperature
Secondary helium outlet temperature
Secondary helium flow rate
162
HTTR-LP (15MW)
468C
235C
2.8MPa(abs)
4.1kg/s
158C
420C
2.9MPa(abs)
3.6kg/s
HTTR-LP (30MW)
828C
383C
4.0MPa(abs)
4.1kg/s
248C
751C
4.0MPa(abs)
3.6kg/s
Table 3-24. Conditions of primary pressurized water cooler before LP tests

Items
Helium pressure
Helium flow rate
Water pressure
Water flow rate
HTTR-LP (15MW)
468C
235C
2.8MPa(abs)
8.2kg/s
79C
100C
2.3MPa(abs)
111kg/s
HTTR-LP (30MW)
828C
390C
4.0MPa(abs)
8.2kg/s
120C
160C
3.5MPa(abs)
113kg/s
Table 3-25. Conditions of secondary pressurized water cooler before LP tests

Items
Helium pressure
Helium flow rate
Water pressure
Water flow rate
HTTR-LP (15MW)
420C
152C
2.9MPa(abs)
3.6kg/s
79C
100C
2.3MPa(abs)
57kg/s
HTTR-LP (30MW)
751C
243C
4.0MPa(abs)
3.6kg/s
120C
158C
3.5MPa(abs)
59kg/s
Table 3-26. Air cooler conditions for pressurized water cooling system before LP tests
Items
Water pressure
Water flow rate
Air pressure
Air flow rate
HTTR-LP (15MW)
100C
45C
2.2MPa(abs)
65kg/s
7C
62C
0.1MPa(abs)
270kg/s
HTTR-LP (30MW)
160C
66C
3.3MPa(abs)
76kg/s
8C
88C
0.1MPa(abs)
370kg/s
Tables 3-27 and 3-28 show the temperature, pressure and flow rate of the auxiliary
heat exchanger and the air cooler for the auxiliary cooling system before the loss of off-site
electric power tests from 15 and 30 MW operations, respectively. The auxiliary cooling
system is in standby during the normal operation. During the standby, the two auxiliary
helium circulators were not operated, however, a small amount of helium of 0.031 kg/s from
the primary helium purification system was provided into the auxiliary cooling system to keep
the auxiliary heat exchanger from thermal shock at the startup of the auxiliary cooling system.
Furthermore, water pressure of the auxiliary cooling system was regulated through its
controlling system, and one of the two water pumps for the auxiliary cooling system was
driven. The water flow rate in standby was fixed at 5.6 kg/s.
163
Table 3-27. Conditions of auxiliary heat exchanger before LP tests

Items
Helium pressure
Helium flow rate
Water pressure
Water flow rate
HTTR-LP (15MW)
426C
74C
2.8MPa(abs)
0.031kg/s
59C
62C
1.3MPa(abs)
5.6kg/s
HTTR-LP (30MW)
754C
97C
4.0MPa(abs)
0.031kg/s
63C
71C
2.3MPa(abs)
5.6kg/s
Table 3-28. Conditions of air cooler for auxiliary cooling system before LP tests
Items
Water pressure
Water flow rate
Air pressure
Air flow rate
HTTR-LP (15MW)
62C
59C
0.5MPa(abs)
5.6kg/s
7C
8C
0.1MPa(abs)
41kg/s
HTTR-LP (30MW)
71C
63C
1.5MPa(abs)
5.7kg/s
8C
9C
0.1MPa(abs)
118kg/s
Figures 3.9 and 3.10 show transient behaviors of the rotation and the flow rate of the
auxiliary helium circulators during the loss of off-site electric power tests from 15 and 30
MW operations, respectively. In about 1 s after the startup of the auxiliary cooling system, the
rotations of the two auxiliary helium circulators reached about 3000 min-1. This rotation was
kept for about 2 s. After that, the rotation of the two auxiliary helium circulators rose up to
about 8500 min-1 at the constant rate by the automatic frequency control. In about 14 s after
the startup of the auxiliary cooling system, helium flow rates of the auxiliary helium
circulators (A) and (B) reached about 0.73 kg/s and about 0.75 kg/s respectively at the test
from 15 MW operation, while they achieved about 0.85 kg/s and about 0.89 kg/s at the test
from 30 MW operation. In 40 min after the startup of the auxiliary cooling system, helium
flow rates of the auxiliary helium circulator decreased to about 0.78 and about 0.87 kg/s at the
tests from 15 and 30 MW operations, respectively. On the contrary, all the time after startup
of the auxiliary cooling system, water flow rate was fixed at about 20.0 kg/s by operating the
two water pumps.
Estimation items on the benchmark problems are as follows; transition of (1) hot
plenum block temperature, (2) reactor inlet coolant temperature, (3) reactor outlet coolant
temperature, (4) primary coolant pressure, (5) reactor power, (6) heat removal of auxiliary
heat exchanger. Estimation duration is for 10 hr from the beginning of the loss of off-site
electric power.
Figures 3.11 and 3.12 show transient behaviors of the temperatures of the hot plenum
block, the reactor inlet and outlet coolant during the loss of off-site electric power tests from
15 and 30 MW operations, respectively. The hot plenum block temperature decreased
continuously after the startup of the auxiliary cooling system. The reactor inlet coolant
164
FIG.3.9. Transient behaviors of rotation and flow rate of auxiliary helium circulators during loss of off-site electric power test from 15 MW
operation.
165
166
FIG. 3.10. Transient behaviors of rotation and flow rate of auxiliary helium circulators during loss of off-site electric power test from 30 MW
operation.
temperature reduced rapidly by the startup of the auxiliary cooling system in about 50 s after
the loss of off-site electric power. The reactor inlet coolant temperature decreased suddenly
by the stop of one of the two auxiliary helium circulators in 40 min after the startup of the
auxiliary cooling system, and since then it dropped gradually. The reactor outlet coolant
temperature reduced continuously due to the insertion of the control rods and the startup of
the auxiliary cooling system.
Figures 3.13 and 3.14 show transient behaviors of the primary coolant pressure and
the reactor power during the loss of off-site electric power tests from 15 and 30 MW
operations, respectively. The primary coolant pressure decreased gradually with reduction of
the coolant temperature. The reactor power dropped drastically by the insertion of the control
rods, and then the subcriticality of the reactor was maintained.
Figures 3.15a and 3.15b show transient behaviors of the heat removal of the auxiliary
heat exchanger during the loss of off-site electric power tests from 15 and 30 MW operations,
respectively. In about 50 s after the loss of off-site electric power the auxiliary cooling system
started up, and subsequently about 10 min later flow rates and temperatures of helium and
water became steady. Then the heat removal of the auxiliary heat exchanger was
approximately 1.8 MW at the test from 15 MW, while it was approximately 4.0 MW at the
test from 30 MW operation. The heat removal decreased suddenly by the stop of one of the
two auxiliary helium circulators in 40 min after the startup of the auxiliary cooling system,
and since then it dropped gradually.
FIG. 3.11. Transient behaviors of temperatures of hot plenum block, reactor inlet and outlet
coolant during loss of off-site electric power test from 15 MW operation.
167
FIG. 3.12. Transient behaviors of temperatures of hot plenum block, reactor inlet and outlet
coolant during loss of off-site electric power test from 30 MW operation.
FIG. 3.13. Transient behaviors of primary coolant pressure and reactor power during loss of
off-site electric power test from 15 MW operation.
168
FIG. 3.14. Transient behaviors of primary coolant pressure and reactor power during loss of
off-site electric power test from 30 MW operation.
FIG. 3.15a. Transient behavior of heat removal of auxiliary heat exchanger during loss of offsite electric power test from 15 MW operation.
169
FIG. 3.15b. Transient behavior of heat removal of auxiliary heat exchanger during loss of offsite electric power test from 30 MW operation.
The following major results were determined from the loss of off-site electric power tests.
(1) Because helium circulators and water pumps coasted down immediately after the loss
of off-site electric power, flow rates of helium and water decreased to the scram
points. In about 50 s after the loss of off-site electric power, the auxiliary cooling
system and the vessel cooling system started up by supplying electricity from two
emergency power feeders.
(2) At the test from 15 MW operation, all the control rods were inserted simultaneously
into the reactor core by gravity. At the test from 30 MW operation, the outer nine pairs
of the control rods were inserted into the replaceable reflector region of the core at
first, and 40 min later the other inner seven pairs fell into the fuel region. The reactor
power dropped drastically by the insertion of the control rods, and then the
subcriticality of the reactor was kept.
(3) In about 10 min after the startup of the auxiliary cooling system, flow rates and
temperatures of helium and water of the auxiliary heat exchanger became steady. Then
the heat removal of the auxiliary heat exchanger was approximately 1.8 MW at the test
from 15 MW operation, while it was approximately 4.0 MW at the test from 30 MW
operation. In 40 min after the startup of the auxiliary cooling system, one of two
auxiliary helium circulators stopped. Temperature of hot plenum block decreased
continuously after the startup of the auxiliary cooling system.
170
3.2.2. Russian Federation

3.2.2.1. HTTR Vessel Cooling (HTTR-VC) [3-6]
Description of the codes
SM-1, GTAS-M and DUPT codes were used for computation of power transferred
from the reactor vessel to the reactor cavity cooling system and temperature distribution on
the side panel. The SM-1 code is intended for computation of transient temperatures in
structures with arbitrary geometry and based on solution of the heat conduction equation by
heat balance method.
Thermal conductivity, gas convection and radiation are taken into account while
determining effective conductivity factor in gas gaps. The code doesnt include solution of
fluid motion and energy equations.
The GTAS-M code is intended to analyze transient temperature of HTGR cores and
structures. The code allows the capability to obtain 3-D temperature distribution for a given
geometry at designed regime parameters - time" relations. The code is based on solution of
the energy equation for solids in 3-D approximation. The three dimensions are restricted by
steady geometry in any section along the height. The fluid motion equation is solved in
stationary approximation. Cooling of a region to be calculated is allowed by several groups of
flow rates which are united by collectors. Distribution of the coolant between channels of
each group is performed an a condition of equal pressure difference between high and low
pressure collectors. The code doesn't allow calculation of connective coolant flow in 2-D and
3-D approximations.
DUPT code is intended for computation of coolant connective flow in 2-D geometry.
The code solves Navier-Stokes equations in Buossinesq approximation.
More detailed description of the codes is presented in [3-5].
Methodology
A simplified methodology was developed for the benchmark taking into account
complex conditions for heat transfer between the reactor vessel and the reactor cavity cooling
system and lack of a code to solve the problem as a whole. The space between the reactor
vessel and the cavity is divided on three sections:
-
upper (EL.27.22 m and EL/31/44.3 m);

side (EL.19.65 m and EL.27.22 m);
lower (EL.16.75 m and EL.19.65 m).
The schematic diagram for the benchmark problem is shown on Figure 3.16.
171
FIG. 3.16. Computing range diagram.
172
The GTAS-M code is used for estimation of heat removal from the reactor vessel only
for its cylindrical section. The collector model of the code does not allow taking into
consideration the real geometry of the upper and lower sections.
Heat transfer from the reactor vessel for the upper and lower sections is performed
with using SM1 and DUPT codes. The temperature of the in-vessel structures and water inlet
temperature of the reactor cavity cooling system was adopted as boundary conditions.
Preliminary calculations to estimate the temperature of the in-vessel structures and sensitivity
study showed that its value may be taken equal to 425 C.
Use of SM1 code for the upper and lower sections enables to account for the head and
bottom shape in calculations of heat transport by radiation. The heat transfer factor of the air
in the cavity was derived from velocity profiles in the upper that were calculated by DUPT
code. The air temperature was assumed to be equal to the air temperature at outlet of the
middle section. Heat removal by the standpipes was taken for increase of surface of heat
transport by convection. ISI plates were not taken into consideration in the analysis.
Results
The results of analyses are presented in Figures 3.17 and 3.18 and in Tables 3-28 and
3-29. Figure 3.17 presents temperature distribution over the reactor vessel height and side
panel surface of the vessel cooling system during the reactor operation at 9 MW power.
Figure 3.18 depicts the temperature distribution during reactor operation at 30 MW power.
Table 3-28. Computed values of heat removal during the reactor operation at 9 MW power
Computing range
Upper part
Middle part
Lower part
Total
Amount of removed heat, kW

3.19
119
11.3
133.5
The analysis results demonstrated that during the reactor operation at 9 MW power
133.5 kW of heat is removed to the reactor vessel cooling system panels, 77 % of them in the
middle part of the computing range. The portion of heat transferred by radiation amounts to
55 %.
Table 3-29. Computed values of heat removal during the reactor operation at 30 MW power
Computing range
Upper part
Middle part
Lower part
Total
Amount of removed heat, kW

16.5
419
58.7
494.1
In accordance with performed analyses, during the reactor operation at 30 MW power

about 494 kW of heat is removed to the reactor vessel cooling system panels. 74.2 % of this is
removed in the middle part of the computing range. This portion of heat being transferred by
radiation amounts to 64 %.
173
FIG. 3.17. Distribution of reactor vessel and side panel temperatures over height during the
reactor operation at 9 MW power.
3.2.2.2. HTTR Loss of Electric Power (HTTR-LP) [3-10]
Description of the VGM code
The VGM-code is intended for calculating normal and emergency transients in nuclear
power plants cooled by water or helium. Simulating regimes include:
- normal operation,
- urgent drop in power due to failure of equipment,
- emergency regimes (inadvertent withdrawal of control rods, loss of load, loss of
power, etc.).
Circuit and system simulation includes main equipment having an effect on the
transients. The mathematical simulation comprises neutronic process, thermal-hydraulic
process in the primary, secondary circuits and emergency decay heat removal system, heat
transfer in fuel elements, a model of reactivity control system.
174
FIG. 3.18. Distribution of reactor vessel and side panel temperatures over height during the
reactor operation at 30 MW power.
Reactor power is determined by the point kinetics method in approximation of six

groups of delayed neutron sources. Reactivity dependence on coolant temperature and
density, fuel and moderator temperatures is taken into account at calculation of feedback
reactivity effects. The feedback effects are described taking into consideration their nonlinearity and non-uniformity of multiplying characteristics in the core space.
The thermal-hydraulic process is simulated in the frame of 1-D model with account of
natural convection of the coolant and possibility to change a direction of the circulation.
Mixing of flow rates with different temperatures is assumed to be ideal. The core thermal
model allows computation of coolant and moderator temperatures in axial and radial
directions versus core inlet temperature, flow rate distribution and power.
175
FIG. 3.19. HTTR plant calculation diagram.
The mathematical model of reactor control and protection system simulates operation
of automatic regulator on relay principle, account of possible time delay during formation of
shutdown signals, acceleration time of absorber rods and balls.
The VGM code provides the capability to calculate core power density, temperature
distribution in fuel elements, coolant temperatures and flow-rates in primary, secondary and
emergency cooling circuits.
The text of the code is written on the following:
FORTRAN-77
Operational system-MS DOS 6.22
Type of language translator-MICROSOFT FORTRAN VERSION 5.00
A more detailed description of the code is presented in [3-7].
176
16
Reactor power (MW)
12
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
FIG. 3.20. Transient behavior of reactor power during loss of off-site electric power test
(15 MW).
600
Temperature (
400
200
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
Temperatures:
core inlet (VGM)
core outlet (VGM)
core inlet (experiment)
core outlet (experiment)
FIG. 3.21. Transient behavior of temperatures during loss of off-site electric power test
(15 MW)).
177
Pressure (MPa)
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
VGM
experiment
FIG. 3.22. Transient behavior of pressure during loss of off-site electric power test
(15 MW)).
3
Heat removal (MW)
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
VGM
experiment
FIG. 3.23. Transient behavior of heat removal during loss of off-site electric power test
(15 MW).
178
Reactor power (MW)
30
20
10
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
FIG. 3.24. Transient behavior of reactor power during loss of off-site electric power test
(30 MW).
1000
800
Temperature (
600
400
200
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
Temperatures:
core inlet (VGM)
core outlet (VGM)
core inlet (experiment)
core outlet (experiment)
FIG. 3.25. Transient behavior of temperatures during loss of off-site electric power test
(30 MW).
179
Pressure (MPa)
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
VGM
experiment
FIG. 3.26. Transient behavior of pressure during loss of off-site electric power test
(30 MW).
4
Heat removal (MW)
0
0
3600
7200 10800 14400 18000 21600 25200 28800 32400 36000
Time (sec)
VGM
experiment
FIG. 3.27. Transient behavior of heat removal during loss of off-site electric power test
(30 MW).
180
Methodology
The HTTR calculational model was developed (Figure 3.19) on the basis of data
presented in [3-8, 3-9]. The number of breakdown sections that were adopted included the
following:
- core height
- IHX and PPWC length
- heat exchange tube thickness
- gas spaces
42
10
4
2.
For a description of the transient heat transfer in the core, two types of cells with
average and peak power are chosen. Every cell includes a fuel rod, a gap with cooling helium
and equivalent mass of graphite in one block per one cell. The heat transfer in the axial
direction by conductivity is not taken into account.
The heat exchange equipment is simulated as a single equivalent tube. Axial heat
transfer is ignored.
Results
Results of the calculation of HTTR behavior during the loss of power are shown on
Figures 3.20, 3.21, 3.22 and 3.23 for 15 MW thermal power and on Figures 3.24, 3.25, 3.26
and 3.27 for 30 MW thermal power.
Flow rates of coolant through the primary gas circulators are reduced to zero in 10 s
after the loss of off-site electric power signal.
Emergency protection actuates in 1.6 s after loss of off-site electric power. Neutron
power decreases down to close to decay heat for 7 s.
The comparison of the calculated and experimental data shows the following.
In the beginning of the transient (0-2 hr) one can see a good agreement between
calculated and experimental values of coolant temperatures. Ratios of experimental and
calculated heating of the coolant are more than 1.1 and 1.3 for 15 and 30 MW tests,
respectively. Further, the discrepancy increases up to 2 for 15 MW and 1.5 for 30 MW at the
transient end-point (10 hrs).
The calculated values of heat removal for 15 MW are in a good agreement with the
experimental data during first five hours of the transient. Further, the experimental values are
well higher than calculated ones. Their ratio is 1.75 for the transient end-point (10 hrs).
For 30 MW an acceptable agreement between calculated and experimental results is
only during first hour of the transient. The further transient behavior is also characterized by
excess of the experimental heat removal as against the experimental one which reaches ~ 1.7
for the end-point.
The discrepancies can be explained by the following reasons:
1) The decay heat error has a strong effect on the calculated results of temperatures
and heat removal, since the decay heat several times more than accumulated in the core and
181
reflectors. This specially affects on the discrepancies in the interval of 5-10 hrs. Increase of
decay heat leads to reduction of discrepancies. For example, decay heat 1.5 as large gives full
agreement between calculated and experimental results.
2) Another reason may be connected with influence of the finite break-down of the
core in axial and radial directions on the results. Further analysis is needed to be carried out
for this case.
182
3.2.3 South Africa [3-11]

This report details the benchmarking of the Flownet Nuclear simulation package as used
by Pebble Bed Modular Reactor in South Africa. The benchmark is of a loss of off-site power
event that was tested at the HTTR facility. The event involves the electric power supply to
the circulators being cut off, reactor SCRAM and the activation of the Auxiliary Cooling
system to remove decay heat.
3.2.3.1 System Overview
The HTTR system was modelled and analysed with Flownet Nuclear and is comprised
of a graphite-moderated prismatic block reactor, a primary cooling circuit, a secondary
cooling circuit and an auxiliary cooling circuit. Helium is used as the working fluid. During
normal operation the helium is heated in the reactor and is then circulated via concentric ducts
to two parallel heat exchangers. The first is the Primary Pressurised Water Cooler (PPWC)
that is a shell-and-tube water-cooled exchanger. The second is the Intermediate Heat
Exchanger (IHX) that is a spirally wound shell and tube type. Its function is to transfer heat
to the secondary helium circuit. Both heat exchangers exits are connected to helium
circulators providing the mass flow rate needed in the circuit. The HTTR system can either
be operated using only the PPWC as cooler or both the PPWC and the secondary circuit (the
latter is called parallel cooling operation). A schematic of the HTTR cycle is shown in Figure
3.28.
Secondary cooling circuit
Reactor
IHX
SPWC
AHX
Auxiliary cooling circuit
Primary cooling circuit
PPWC
FIG. 3.28. Schematic diagram of the High Temperature engineering Test Reactor.
183
In the secondary helium circuit a helium circulator is used to circulate flow between the
IHX and the Secondary Pressurised Water Cooler (SPWC), the latter being of similar but
smaller design to the PPWC. Both water coolers watersides are on water circuits that
exchange heat with the atmosphere via water-to-air heat exchangers.
During an accident condition such as the loss of off-site electrical power the Auxiliary
cooling circuit is used. This circuit uses two gas circulators to force a low mass flow rate of
helium between the reactor and the Auxiliary Heat Exchanger (AHX). The AHX is a water
cooler of similar but smaller design to the PPWC. The function of this system is to remove
decay heat from the reactor core after a reactor shutdown or SCRAM has been initiated.
3.2.3.2 Flownet Nuclear
Flownet Nuclear is a pipe-network simulation code used for solving thermodynamic and
fluidic problems in pipe networks and energy systems. M-Tech Industrial in collaboration
with the School for Mechanical and Materials Engineering at Potchefstroom University in
South Africa has developed the code over the past ten years. Initially the code was used for
large gas pipe networks but has since been extended to handle most types of thermalhydraulic systems. The following points highlight the capabilities of the program:
The ability to deal with compressible as well as incompressible fluids.

Full heat transfer by means of conduction, convection and radiation.
High solution speed. The typical solution time of a 200-element network is less than
30 seconds on a standard PC.
Various types of network components are available in the program
Constant and variable area pipes and ducts

Restrictors with loss or discharge coefficients.
Cooling orifices for which NASA empirical correlations are used to calculate the
discharge coefficient.
Non-circular ducts with area changes. With this element the user has the option of
taking the effect of the injected momentum and variations in total temperature on
the change in total pressure into account.
Various types of heat exchangers.
Compressible flow through long pipes.
Compressors and turbines
Pebble-bed and solid fuel HTGR reactors as special models.
Heat transfer paths
The ability to specify a fixed mass source or a sink in junctions and volumes
between flow elements
Variable fluid properties.
Ability to automatically subdivide long pipes into smaller increments.
Gravitational effects in compressible flows
Flownet Nuclear has also been extensively benchmarked and it was found that results
are in close agreement with experimental data and analytical calculations. An example is a
test involving compressible gas flow through a long pipe. At time zero a valve is suddenly
closed at the end of the pipe. The inertia of the fluid causes auto-compression at the end of
the pipe that in turn results in a pressure wave travelling back through the pipe. Flownet
Nuclear data is benchmarked against data obtained using the Lax-Wendroff method. Figure
3.29 shows a plot of total pressure in the middle of the pipe verses time.
184
10000
9000
Total Pressure [kPa]
8000
7000
6000
5000
4000
3000
2000
1000
0
0
10
15
20
25
30
35
40
45
50
55
60
Time [s]
Bench
Flownet
FIG. 3.29. Total pressure versus time for the sudden closure of a valve at the end of the pipe.
Applications of the code include:
Ventilation systems.
Compressed air systems of mines and factories.
Air distribution system of a conveyor belt system that uses an air cushion to suspend
the belt.
Water reticulation systems.
Heat recovery systems.
Flow through the combustion chamber of an aircraft engine.
Flow through the re-heater of an aircraft engine.
City gas pipe networks.
Brayton power cycle as in the PBMR.
Complete aircraft air-conditioning system.
The detail concerning the theory behind Flownet Nuclear falls beyond the scope of this
document.
3.2.3.3 Modeling of the HTTR in Flownet
Cycle description
Flownet uses elements and nodes to represent system components. An element refers to
the actual components such as pipes, turbo-machines or heat exchangers. Nodes are junction
points for elements but can also represent regions of large volume and low flow velocity
tanks and reservoirs.
185
Figure 3.30 is a diagram of the HTTR system as implemented in terms of Flownet elements
and nodes. Elements are shown as numbered circles while the numbered squares are nodes.
The diagram shows the use of pipe elements (DW and DG), compressors (CM), heat
exchangers (HX and RX) and conductive heat transfer paths (CHT). Only the main, auxiliary
and secondary cooling systems are modelled for simplicity. Cooling water flows are
modelled as constant-mass flow cold-water streams. The vessel cooling system is not
modelled in this investigation.
53
20
48
22
14
11
9
33
13
20
21
31
15
47
23
18
Reactor
30
27
51
28
29
22
PH
46
11
24
26
15
36
17
SHC
34
24
33
38
35
39
SPWC
14
36
27
40
24
12
37
19
41
38
21
HX
50
HX
25
10
PPWC
16
6
24
45
48
28
44
42
17
16
25
47
46
18
12
43
41
49
AHX
34
37
AHC
45
19
7
PHC
24
49
32
31
13
52
32
35
29
30
IHX
42
44
40
43
39
FIG. 3.30. HTTR Flownet Network
The descriptions of the elements used in the network are shown in Table 3-30.
Table 3-30. Elements used in Flownet network
Element
no.
Description
Type
1
2
3
4
Annular duct to top of reactor

Reactor to T-junction
Contraction to gas duct
Contraction from T-junction
Gas duct
Gas duct
Gas duct
Gas duct
Element
no.
Description
Type
5
6
7
8
9
10
Intermediate heat exchanger

IHX to PHC
SPWC circulator
PPWC circulator B
PPWC to PHC
Primary Pressurized Water Cooler
Heat exchanger
Gas duct
Helium circulator
Helium circulator
Gas duct
Heat exchanger
186
11
12
13
14
15
16
17
18
19
20
22
23
24
25
26
27
28
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
PPWC circulator C
IHX secondary side
PPWC to T-junction
T-junction to reactor
PPWC circulator A
9th reflector layer
Hot plenum block
T-junction to PPWC
Secondary IHX to T-junction A
Annulus around PPWC
Upper reflector and shield layers
Fuel assembly (Reactor core)
Lower reflector layer
PPWC water side
HTTR
Secondary IHX to T-junction B
Distribution plate
PHC to PPWC
Contraction to IHX
PHC to IHX
Annulus around IHX
IHX to T-junction
Distribution plate
Hot plenum outlet channel
Secondary Pressurized Water Cooler
SPWC water side
SPWC to SHC
Secondary Helium Circulator
SHC to SPWC
Annulus around SPWC
SPWC to T-junction
T-junction to tube header feed pipes
Header feed pipes
Reactor to AHX
Auxiliary Heat Exchanger
AHX water side
AHX to AHC
Auxiliary Helium Circulator A
Auxiliary Helium Circulator B
AHC to AHX
Annulus around AHX
AHX to Reactor
Helium circulator
Heat exchanger
Gas duct
Gas duct
Helium circulator
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
HTTR
Gas duct
Heat exchanger
HTTR
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Heat exchanger
Heat exchanger
Gas duct
Helium circulator
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Gas duct
Heat exchanger
Heat exchanger
Gas duct
Helium circulator
Helium circulator
Gas duct
Gas duct
Gas duct
The detailed modelling falls beyond the scope of this document, however some
comments can be made on the modelling of generic components
Piping and duct elements
Most of the piping in the HTTR plant consists of double-walled pipes with hot flow
from the reactor in the inner pipe and cooled return gas in the outer annulus of the pipe.
Figure 3.31 shows a lateral cross-section through the primary cooling circuit piping. The
187
diagram clearly shows the double-walled piping and the lagging material surrounding the pipe
for insulation. Flow from the reactor goes via the inner pipe to a T-junction where the flow is
split up to a ratio of 3:1 between the PPWC and the IHX. Upon returning to the reactor the
helium flows through the outer annular pipe.
The pipes are modelled as Darcy-Weisbach (DW) elements in Flownet. These pipe
elements use the following formula to calculate the pressure drop through a pipe for
compressible flow.
fL T0 p 0 M 2
p0 =
+
T0 2
D
(1)
where f friction factor

L Pipe length
[m]
D Pipe diameter
[m]
T0 Total temperature
[K]
p0 Total pressure
[Pa]
ratio of specific heats of the gas

M Mach number
The friction factor is calculated using a correlation fitted to the Moody diagram with the
transition between laminar and turbulent flow beginning at Reynolds numbers greater than
2300.
FIG. 3.31. Primary circuit piping.
188
In the case of the annular flows the cross-sectional flow area and wetted perimeter are
inputs to Flownet that will then calculate a hydraulic diameter for use in (1). The flow area is
used along with mass flow rate to calculate the velocity component of (1). All the other terms
are inputs in the pipe element model.
Heat exchanger elements
Shell-and-tube heat exchangers
The PPWC, SPWC and AHX are three shell-and-tube heat exchangers of similar
design. Figure 3.32 shows a birds eye view of the inside of the HTTRs PPWC. The tubes
used are un-finned. There are 5 gas passes and 2 tube passes in the heat exchanger. A
distribution plate is placed at the inlet to spread the gas flow evenly through the first stage.
Flownet Nuclear has a Shell-and-Tube Heat exchanger or STX type element specifically
catering for this type of heat exchanger. In all Flownets heat exchanger models provision is
made for both steady-state heat transfer parameters and parameters describing the heat
exchangers thermal inertia. Thermal inertia is very important when one wants simulating
transient changes in the heat exchangers inlet and outlet conditions.
In Flownets STX model the tube and baffle layout and dimensions are required as
inputs along with parameters such as total heat transfer area, primary/secondary area ratio and
tube conduction coefficients. For transient simulations the mass capacitance and gas volume
are inputs. The heat transfer coefficient and friction factor on the tube side are calculated
using the Dittus-Boelter equation and the standard Darcy-Weisbach pipe model, respectively.
On the shell side the Reynolds number is calculated and a corresponding Colburn j factor and
friction factor is read off an input graph. These two graphs are specifically set-up for the tube
layout used in the heat exchanger. The two values are then used to calculate the heat transfer
and pressure drop through the heat exchanger using standard correlations.
FIG. 3.32. Birds eye view of the PPWC.
189
Intermediate Heat Exchanger

This heat exchanger functions as an energy interface between the primary and
secondary cooling circuits. The IHX is a helically coiled shell-and-tube gas-to-gas heat
exchanger. Figure 3.33 shows a diagram of the actual IHX heat exchanger. Hot gas from the
reactor enters from below and flows up over the helical tube bundles to a manifold volume at
the top of the heat exchanger. From there it goes to a circulator and back again through the
outer annulus of the heat exchanger pressure vessel.
The tubes contain cool gas from the secondary cooling circuit. This gas flows down
from the top of the heat exchanger making the configuration counter flow.
The gas-to-gas type heat exchanger is modelled as a recuperator heat exchanger (RX) in
Flownet. This type of element requires the hydraulic diameter, total free flow area and flow
length of both the hot and cold sides. It calculates the heat transfer and friction losses by
means of Nusselt number and friction factor correlations respectively. The thermal inertia is
accounted for by means of a mass-capacitance or MC value.
FIG. 3.33. Birds eye view and sections of the Intermediate Heat Exchanger.
190
The RX model is mostly suited to simple plate-fin recuperators where the hot and cold
side are of similar length and cross-section. As such the helical-coil configuration of the IHX,
that has a much longer flow path on the secondary side, had to be tuned to get correct heat
transfer in steady-state operation. This was done by altering the Nusselt and friction drop
factors that are inputs in the RX model.
Helium circulators
The helium circulators used in the gas circuits are single-stage centrifugal compressors.
Helium from the associated cooler enters the circulator inlet and goes through a filter before
being compressed. A variable speed motor is used to control the helium flow rate delivered by
the circulator.
In Flownet turbines, blowers and compressors are modelled by means of characteristic

curve data. The pressure ratio verses corrected mass flow (m T /P ) and efficiency verses
corrected mass flow characteristic curves for different corrected shaft speeds (N / T ) are used.
An input shaft speed is used along with the inlet gas conditions to find the operating point on
the curves.
0
The graphs of pressure ratio versus invariant mass flow rate and invariant power versus
invariant mass flow rate were supplied by JAERI. Graphs were supplied for the primary,
secondary and auxiliary circuit machines. These graphs had to be converted into the Flownet
compressor curve file format before it could be used in the circuit.
Conversion of invariant mass flow to corrected mass flow:
mcor = minv / 3600
(2)
mcor p01
(3)
m=
T01
Since Flownet requires a graph of efficiency versus corrected mass flow rate the
invariant power data had to be modified to obtain efficiency. This was done as follows:
(4)
P = PInv p01 T01

Pfluidic = ( pratio
( 1) /
1)cPT01m / 1000
= Pfluidic / P
(5)
(6)
Where P Power input to the compressor

Pfluidic Power delivered to fluid
pratio pressure ratio (JAERI)
- compressor efficiency
T01 Inlet temperature
p01 Inlet pressure
191
The characteristic maps for the primary and secondary circulators are the same and are
shown in Figure 3.34 and 3.35. The primary circulator runs at a speed of 9400 rpm while the
secondary machine is slightly faster at 9900rpm at 30MW. For 15MW they operate at 8900
rpm and 9500 rpm respectively.
FIG. 3.34. Primary and Secondary circuit Pressure ratio map.
FIG. 3.35. Primary and Secondary circuit Efficiency map.
192
HTTR Reactor
The reactor is modeled by means of a specially developed HTTR element in Flownet

Nuclear. This model encompasses all the core internals starting from the channel entry at the
upper reflector through to the hot plenum beneath the core. The core nuclear dynamics are
modeled with a point kinetics model that uses the core dynamics parameters as prescribed by
JAERI. During the transient event the decay heat production of the core is input by means of
the prescribed decay heat equation given by JAERI.
3.2.3.4 Steady State Results
Steady-state results comparison 15MW operation
This section discusses the results for the 15MW loss of power benchmark. The results
tables contain three columns: the first column tabulates the predicted HTTR data as sent out in
JAERIs first document on the benchmark [3-12], the second column contains the test data as
obtained in the 15MW loss of power experiment while the last column gives the Flownet
Nuclear data.
The data shows that the Flownet Nuclear results are closer to the experiment than to the
predicted values. This is possibly because, for the experiment, the speeds of the circulators
were greatly reduced from predicted values. It was found that if the predicted speeds are used
in the simulation the mass flow becomes much higher than what is needed for 15MW power
delivery in the reactor. It is assumed that JAERI was conservative in their calculation of the
system resistance and hence the circulators were over sized.
Table 3-31. Reactor core results comparison

Parameter
HTTR Estimate
HTTR
Experimental
Flownet
Nuclear
Thermal power
[MW]
15
15
15
Helium inlet temp.

[C]
241 (About)
242
240.5
Helium outlet
temp. [C]
470 (About)
468
472.7
Helium pressure
[kPa]
3000 (About)
2900
2795
Helium mass flow

rate [kg/s]
12.4
12.4
12.44
Hot plenum block

temp. [C]
490 (About)
391
193
Table 3-32. Intermediate Heat Exchanger results comparison

Parameter
HTTR
Estimate
HTTR
Experimental
Flownet
Nuclear
Primary helium inlet

temp. [C]
468 (About)
468
472.7
Primary helium
outlet temp. [C]
238 (About)
235
240.77
Primary helium
pressure [kPa]
3000 (About)
2800
2784
Primary helium mass

flow rate [kg/s]
4.1
4.1
4.078
Secondary helium
inlet temp. [C]
154 (About)
158
160.6
Secondary helium
outlet temp. [C]
431 (About)
420
423.7
Circulator speed
[rpm]
9500
8770
8770
Secondary helium
pressure [kPa]
3100 (About)
2900
2900
Secondary helium
mass flow rate [kg/s]
3.6
3.6
3.59
Otherwise, there are small differences in pressure and temperatures but until more
detailed data becomes available on pressure drops through the HTTR these differences are
considered negligible. Tables 3-31, 3-32, 3-33 and 3-34 provide a comparison of the Flownet
results against the computed results supplied by JAERI for the various parts of the HTTR
system.
Steady-state results comparison 30MW operation
This section gives the results for the 30MW simulations. Tables 3-35, 3-36, 3-37 and 338 each contain the following: HTTR predicted results, Flownet Nuclear results using the
prescribed shaft speeds and lastly Flownet results based on the same shaft speeds as used in
the 15MW test. As can be seen the first column of Flownet results contains large differences
to the predicted HTTR results. After careful analysis of the 15MW experimental data it was
found that the system resistance was overestimated and so the 15MW shaft speeds were
reduced.
194
Table 3-33. Primary Pressurized Water Cooler results comparison

Parameter
HTTR
Estimate
HTTR
Experimental
Flownet
Nuclear
Helium inlet temp.

[C]
468 (About)
468
472.7
Helium outlet temp.

[C]
242 (About)
235
240.77
Helium pressure [kPa]
3000 (About)
2800
2784
Helium mass flow rate

[kg/s]
8.3
8.2
8.308
Circulator Speed
[rpm]
8900
8000
8000
Water inlet temp. [C]
89 (About)
79
79
Parameter
HTTR
Estimate
HTTR
Experimental
Flownet
Nuclear
Water outlet temp.

[C]
110 (About)
100
101.29
Water pressure [kPa]
2600 (About)
2300
2300
Water mass flow rate

[kg/s]
115
111
111
For the 30MW case the mass flow required is the same as for 15MW, the difference in
power is due to differences in inlet and outlet temperatures of the core. However, the mass
flow requirement is dominant in sizing the circulators rotational speeds and since the mass
flow is the same as for 15MW the rotational speeds should be roughly the same. Column
three clearly shows much better correlation between the Flownet simulation and the HTTR
predicted data. These conclusions will only be clarified once experimental data is received
for the 30MW case.
3.2.3.5 Transient Results
This section presents the transient results of the loss of power benchmark for both the
30MW and 15MW cases.
Simulation setup
This transient involves a coasting down of the main loop circulators of the HTTR. A
reactor SCRAM and loss of heat exchanger cooling mass flow are also included in the
transient. Later in the analysis the auxiliary circulators are started and the cooling of the
reactors decay heat is commenced.
195
Table 3-34. Secondary Pressurized Water Cooler results comparison

Parameter
HTTR Estimate
HTTR
Experimental
Flownet
Nuclear
Helium inlet temp.

[C]
430 (About)
420
423.6
Helium outlet
temp. [C]
154 (About)
152
155.86
Helium pressure
[kPa]
3100 (About)
2900
2861
Helium mass flow

rate [kg/s]
3.6
3.6
3.59
Circulator speed
[rpm]
8100
8130
8130
Water inlet temp.

[C]
89 (About)
79
79
Water outlet temp.

[C]
110 (About)
100
99.9
Water pressure
[kPa]
2500 (About)
2300
2300
Water mass flow

rate [kg/s]
60
57
57
Table 3-35. Reactor core results comparison
196
Parameter
HTTR
Estimate
Flownet
Nuclear
Pre-scribed
Shaft speeds
Flownet
Nuclear 15
MW Shaft
speeds
Thermal power [MW]
30
30
30
Helium inlet temp. [C]
395
412
402.7
Helium outlet temp. [C]
850
789.2
858.9
4000
3993
3995.7

[kg/s]
12.4
15.31
12.66
Hot plenum block temp.

[C]
890
789.2
858.9
Table 3-36. Intermediate Heat Exchanger results comparison
Parameter
HTTR
Estimate
Flownet
Nuclear
Pre-scribed
Shaft speeds
Flownet
Nuclear 15
MW Shaft
speeds
Circulator speed [rpm]
9900
9900
8770
Primary helium inlet temp.

[C]
850
789.2
858.9
Primary helium outlet

temp. [C]
395
419.53
421
Parameter
HTTR
Estimate
Flownet
Nuclear
Pre-scribed
Shaft speeds
Flownet
Nuclear 15
MW Shaft
speeds

[kPa]
4000
3977.4
3983.67
Primary helium mass flow

rate [kg/s]
4.1
4.8
4.08
Secondary helium inlet

temp. [C]
241
260.7
262.74
Secondary helium outlet

temp. [C]
783
724.4
721.1

[kPa]
4100
4092.5
4092
Secondary helium mass

flow rate [kg/s]
3.6
3.87
3.86
The loss of power benchmark is modelled as a transient in Flownet. The code requires a
transient events file that contains the user-defined transient events that happen at specified
times in the transient. The transient event editor allows the user to change time step at any
time during the simulation. For the first part of the transient (to ~15s) a time step of 0.1
seconds is used from then to where the auxiliary cooling system starts a time step of 1 second
is used and then 0.1s is again used for the auxiliary cooling. Table 3-39 includes the transient
events as modelled in Flownet Nuclear.
197
Table 3-37. Primary Pressurized Water Cooler results comparison
Parameter
HTTR
Estimate
Flownet
Nuclear
Pre-scribed
Shaft speeds
Flownet
Nuclear 15
MW Shaft
speeds
9400
9400
8000
850
789.2
856.8
395
401
388.4
4000
3977
3984

[kg/s]
8.3
10.45
8.54
135
135
135
Water outlet temp. [C]
175
177
177
3500
3500
3500
Water mass flow rate [kg/s]
115
115
115
Table 3-38. Secondary Pressurized Water Cooler results comparison
198
Parameter
HTTR
Estimate
Flownet
Nuclear
Pre-scribed
Shaft speeds
Flownet
Nuclear 15
MW Shaft
speeds
8100
8100
8130
782
719.9
723.4
240
256
256
4100
4054.7
4054

[kg/s]
3.6
3.87
3.88
135
135
135
Water outlet temp. [C]
175
171
171
3400
3400
3400
Water mass flow rate [kg/s]
60
60
60
Table 3-39. Transient events as modeled in Flownet Nuclear

Time [s]
Simulated event
Coast down of the five main loop circulators commences. The

speed of the circulators is hard-coded to be the same as the speed
decrease prescribed by JAERI.
Water flow in the PPWC and SPWC is stepped down to 5 kg/s
The circulator models become unstable in Flownet due to
operation far out of the circulator maps range. The circulators
are replaced by mass flow sources that decrease at the same rate
as the circulator mass flows did before 3.4 seconds
Reactor SCRAM is initiated. The decay heat curve is read as
input into the reactor model.
The IHX circulator flow has reached 0 kg/s
The PPWC circulators have reached 0 kg/s
The SPWC mass flow has reached 0 kg/s
The auxiliary cooling system is activated
15 MW
30 MW
3.4
3.4
8.2
8.2
12
10.8
13.4
14.5
15
18
52
52
15 MW Loss of power transient results
Flownet Nuclear models circulators and other turbo-machinery by means of their

characteristic curves or maps. Hence, in a situation such as the loss of power benchmark
where the performance of the circulators determines the behavior of the rest of the system it is
very important that the maps used are correct. In this case the speeds of the circulators are
specified as input to the circulator models and an appropriate operating point is found on the
maps. Figure 3.36 shows the speed of three of the circuits circulators over the first part of
the transient. Figure 3.37 shows a locus plot of the movement of the operating point of the
PPWC circulator on its map. It is clear that during normal operation at 15MW power the
operating point is already off the map. As speed is reduced the operating point moves toward
the origin. It is for this reason that the circulator models become unstable after 3.4 seconds
the models cannot extrapolate far enough to simulate the performance of the machine.
199
160
140
Shaft speed [rps]
120
100
80
60
40
20
0
0
10
12
Time [s]
PPWC (A)
SPWC
IHX
FIG. 3.36. Circulator shaft rotational speed JAERI result used as input to Flownet Nuclear.
FIG. 3.37. Locus plot of circulator coast down on a Pressure ratio vs. corrected mass flow
rate map.
200
The fact that the circulator is operating of the map has an effect on the mass flow that is
produced. Figure 3.38 shows the simulation mass flows plotted against the test results. The
figure shows that the simulation predicts a much faster decline in mass flow rate than what
happens in reality. It is assumed that the form of the map in the inner quadrant is different to
the actual map and as such actual coasting down cannot be accurately simulated.
4.5
4.0
3.5
Mass flow rate [kg/s]
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0
10
15
20
25
30
Time [s]
FN IHX
FN SPWC
FN PPWC tot
JAERI PPWC
JAERI SPWC
JAERI IHX
FIG. 3.38. Circulator mass flow rates of Flownet Nuclear compared to JAERI test results
The transient performance of the reactor is shown in Figure 3.39. Note the sudden drop
in reactor power at 8 seconds due to the SCRAM. The remaining decay heat curve is as
prescribed by JAERI. Figure 3.40 shows the transient fluctuation in system pressure. It is
expected that once the system is not running the pressure will be close to the operating
pressure due to the lack of a high system pressure ratio and large volumes of helium.
500
16000
495
14000
490
485
10000
480
475
8000
470
6000
465
4000
Reactor Outlet Temperature [C]
Reactor Thermal Power [kW]
12000
460
2000
455
0
0
10
20
30
40
50
60
70
80
90
450
100
Time [s]
Thermal power
Reactor outlet T
FIG. 3.39. HTTR Reactor thermal power and outlet temperature.
201
2850
2840
2830
System pressure [kPa]
2820
2810
2800
2790
2780
2770
2760
2750
0
50
100
150
200
250
300
Time [s]
FIG. 3.40. HTTR system pressure (at PPWC circulator outlet).
In Figure 3.41, the mass flow rates are plotted for the whole transient from coast down
to when the auxiliary cooling system is activated. Figure 3.42 provides the HTTR inlet and
outlet temperatures while Figure 3.43 is a plot of the heat removed by the auxiliary heat
exchanger throughout the loss of power transient from 15MW.
14
12
10
Massflow [kg/s]
0
0
10
20
30
40
50
60
-2
Time [s]
Reactor
Auxilliary circuit
FIG. 3.41. Reactor and auxiliary mass flow rates.
202
70
80
600
Temperature [C]
500
400
300
200
100
0
0
5000
10000
15000
20000
25000
30000
35000
Time [s]
Outlet
Inlet
FIG. 3.42. Inlet and outlet temperatures of the HTTR core for the 15 MW loss of power
transient.
2500
Heat removed [kW]
2000
1500
1000
500
0
0
5000
10000
15000
20000
25000
30000
35000
Time [s]
FIG. 3.43. Heat removed by Auxiliary heat exchanger for the 15 MW loss of power transient.
203
The assumptions described in this section are based on simulation judgment and as such
the actual cause of discrepancies may be due to totally different reasons. The only way to
really verify these assumptions is to get more experimental data from JAERI, specifically
temperatures and pressures through the system as measured during the experiment.
30 MW Loss of power transient results
At the time of documenting these results no 30MW experimental data was available
from JAERI. Therefore the 30MW simulation results are unverified but are still presented in
this section. Figure 3.44 and 3.45 show the system mass flows during the coast-down and
entire periods of the transient respectively. As with 15MW it is expected that the actual mass
flow rate decrease will slower than in the simulation.
4.5
Mass flow rate [kg/s]
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0
10
12
14
16
18
20
Time [s]
IHX
SPWC
PPWC
FIG. 3.44. Circulator mass flow rates in coast-down phase.
14
12
Massflow [kg/s]
10
8
6
4
2
0
-2
10
20
30
40
50
60
70
80
Time [s]
IHX
PPWC
Reactor
Auxilliary circuit
FIG. 3.45. Reactor and auxiliary mass flow rates.
204
90
100
The reactor performance is shown in Figure 3.46 while the change in system pressure is
given in Figure 3.47.
900
35000
890
880
25000
870
20000
860
850
15000
840
10000
830
Temperature [C]
Thermal Power [kW]
30000
820
5000
810
0
0
20
40
60
80
800
120
100
Time [s]
Thermal power
Reactor outlet T
FIG. 3.46. HTTR Reactor thermal power and outlet temperature.

4250
4200
4150
Pressure [kPa]
4100
4050
4000
3950
3900
3850
3800
3750
0
10
20
30
40
50
60
70
80
90
100
Time [s]
FIG. 3.47. HTTR system pressure (at PPWC circulator outlet).

Figures 3.48 and 3.49 provide HTTR inlet and outlet temperatures and the heat removed
by the auxiliary heat exchanger throughout the loss of power transient from 30MW,
respectively. Figure 3.50 shows a locus plot of the PPWC circulators operating point
movement on the map. Due to the same steady-state speed being used as for 15MW the
starting point is off the map. Hence the same conclusions can be drawn as for 15MW.
205
1000
900
800
Temperature [C]
700
600
500
400
300
200
100
0
0
5000
10000
15000
20000
25000
30000
35000
Time [s]
Outlet
Inlet
FIG. 3.48. Inlet and outlet temperatures of the HTTR core for the 30 MW loss of power
transient.
5000
4500
4000
Heat removed [kW]
3500
3000
2500
2000
1500
1000
500
0
0
5000
10000
15000
20000
25000
30000
35000
Time [s]
FIG.3.49. Heat removed by Auxiliary heat exchanger for the 30 MW loss of power transient.
206
FIG. 3.50. Locus plot of circulator coast down on a Pressure ratio vs. corrected mass flow
rate map.
As for 15MW, the assumptions made in the 30MW transient can only really be verified
with more detailed transient data from JAERI.
3.2.3.6 Conclusions
The steady-state results obtained from Flownet Nuclear for the 15MW case showed
good correlation to the experimental data when the circulator speeds were lowered to the
actual values. It appears that original speed estimates provided by JAERI were based on
over-conservative system resistance predictions. This resulted in the circulators being
oversized for the required operating point. The working points are just off the circulator map
ranges and as such the accuracy of the analyzed working point is dependent on the
extrapolation routines in the simulation code. This accuracy becomes more evident in the loss
of power event where the working point moves further away from the map. The effect can be
seen in the mass flow plots where the experimental flow rates take longer to decrease than the
simulated results.
However, other factors cannot be ruled out as temperature and pressure effects can also
play a role in the discrepancies. The only way to verify that they are not the cause is to get
more detailed pressure and temperature data of the HTTR system for the loss of off-site power
experiments.
In general, the steady state simulation results and those obtained from the actual plant
are in fair agreement. The transient results show greater deviations but the trends simulated
are still the same as the actual performance.
207
3.2.4. United States of America [3-14]

3.2.4.1. Analysis methodology and code description
The ORNL Graphite Reactor Severe Accident Code (GRSAC) was used for the IAEA
CRP-5 HTTR-VC and HTTR-LP benchmark problems, utilizing both steady state and
dynamic code features. These calculations relate to the HTTR initial rise to power sequence
and safety demonstration tests. An existing HTTR model in GRSAC [3-15] was upgraded to
provide more detail in certain critical areas.
The GRSAC (Graphite Reactor Severe Accident Code) software is a general-purpose
program developed at Oak Ridge National Laboratory (ORNL). It is based on the ORNL
MORECA code for simulating accident scenarios for selected gas-cooled reactor (GCR)
design types [3-13]. The MORECA code and its predecessors were originally developed at
ORNL under the sponsorship of the U.S. Nuclear Regulatory Commission (NRC) to perform
confirmatory licensing-related studies of a variety of High-Temperature Gas-Cooled Reactor
(HTGR) designs, including the Fort St. Vrain HTGR and subsequently the 350-MW(t) steamcycle Modular HTGR (MHTGR). MORECA was later developed - under U.S. Department of
Energy (DOE) sponsorship - to simulate the MHTGR design for the 600 MW(t) direct cycle
gas turbine modular helium reactor (GT-MHR).
Since MORECA is a "hard-wired" code, configured only for a particular reactor
design, the conversion of MORECA to GRSAC was motivated by the need to generate the
connectivities necessary to assemble, verify, and run simulations for a wide variety of
graphite-moderated GCR designs.
Since GRSAC was developed to study a wide variety of core transient and heatup
accident scenarios, it includes a detailed 3-D thermal-hydraulic model for the core, plus
models for the reactor vessel, shutdown cooling system (SCS), and shield or reactor cavity
cooling systems. There is an option to include neutronics (point kinetics) with xenon and
samarium poisoning to study Anticipated Transients Without Scram (ATWS) transients.
The 3-D, hexagonal geometry core model allows for detailed investigations of
azimuthal temperature asymmetries in addition to axial and radial profiles. Variable core
thermal properties are computed functions of temperature and are dependent on orientation
and radiation damage. An annealing model for graphite accounts for the increase in thermal
conductivity that occurs during heatup accidents.
The primary coolant flow models cover the full ranges expected in both normal
operation and accidents, including pressurized and depressurized accidents (and in between),
for forced and natural circulation, for up-flow and down-flow, and for turbulent, laminar, and
transition flow regimes. The primary loop pressure calculation can consider variable
inventory (due to depressurization actions) and loop temperature changes and may use a
simplified model for balance-of-plant temperatures. The models for the reactor pressure
vessel and the shield or cavity cooling system are different for each of the various basic
reactor models. Models are also included for oxidation of graphite (including carbon
deposits), cladding, and fuel. Time-at-temperature failure models for TRISO fuel, fission
product release models for metal fuel, and Wigner stored energy release models are also
available.
208
Other GRSAC features of interest are: fast-running (typically >2000 times faster than
real time on a SUN SparcStation-20 workstation), interactive user interface with on-line and
off-line plotting options, automated sensitivity study capabilities, and on-line documentation
and help screens. The basic designs that can be simulated using GRSAC, which the user may
modify via the interface to a large (but limited) extent, include the British and French Magnox
types (including the Calder Hall, G-2/3, and Bugey-2), Windscale (U.K.), G1 (France), and
the HTTR (Japan). Adaptations and analyses are planned for pebble-bed reactor designs, and
the GT-MHR Plutonium burner (U.S.-Russia). More details on GRSAC are summarized
below, and [3-15] is a complete users manual.
GRSAC code features
Reactor Design Setup
Specific design features for a chosen reactor type can be input by the user via design
screen selections in the following categories: fuel element, nuclear parameters, core layout
design, primary cooling system, vessel design, reactor cavity, fission product release, and
oxidation parameters. Program setup screens allow the user to activate or deactivate
oxidation, Wigner energy, or ATWS features, and to select the coolant gas, core flow
direction and computation time parameters. In some cases, such as for the radial and axial
power peaking factor inputs and flow coast-down curves, graphical displays and automated
consistency check features are included. Pop-up HELP windows and a choice of metric or
English unit entries are available for all user input screens in GRSAC. The user can also select
a "run with validation" option, which is a smart front-end check of the entire set of inputs for
data inconsistencies.
Initial Condition Runs
GRSAC accident sequence analyses require a large set of initial condition values
which are created automatically via the Initial Condition (IC) mode. The user can change
operational inputs such as power level, flow, pressure, etc., and observe the resulting detailed
temperature and flow distributions attain steady state conditions. At any point in the run, one
can store initial condition values in a RUN file.
Interactive and Programmed Inputs
The interactive input screen for accident simulations allows for user inputs (scram,
depressurization, changes in emergency and/or cavity cooling, etc.) at any time during a run.
Such inputs can also be pre-programmed, however, via a programmed input screen that is
available to the user during the run setup procedure.
Accident Sequence Runs
Long-term Loss Of Forced Convection (LOFC) accidents begin with a programmed
flow coastdown transient. LOFC transients in GCRs are generally characterized by slow
heatups due to low power densities and large heat capacities associated with the core. They
may be simulated both with and without total or partial depressurization of the primary
coolant and with or without scram. Optionally, both the active or passive shutdown cooling
systems can be made to be either unavailable or available only intermittently in degraded
states. For helium or CO2-cooled cores, there is an option to allow air ingress following a
209
depressurization, and subsequently to initiate oxidation models for graphite (and clad and
metal fuel, if applicable).
Sensitivity Study Option
Many variations of transient and LOFC accident scenarios have been studied to
observe the sensitivities of the predictions to parametric model and operational assumptions.
These provide guidance in design studies for determining plant operating parameters
(including design power level) and in identifying which physical properties and correlations
are most crucial to the outcome of postulated accidents.
In the GRSAC automated sensitivity study feature, the rationale is to seek out a set of
parameters within user-specified uncertainty bands that result in the worst (or best) case
accident consequences using a gradient search algorithm. Sets of 13 model or design
parameters (such as heat transfer correlations, etc.) and 12 operational/run parameters (such as
time of scram) have been set up to be available for automatic variation (from run to run). The
program allows the user to select up to 10 from this set for any given study. To study the
effects of a single parameter variation in more detail, a single-parameter option can be used.
That parameter is varied uniformly within the uncertainty band (reference run plus 4 others).
A report generator creates a summary of the sensitivity run results.
GRSAC model of the HTTR
The ORNL GRSAC code has a number of general-purpose features that allow
modification of design attributes and operational sequence options via a variety of user
interface screens. However, to accommodate HTTR calculations for previous IAEA CRP-3
benchmark calculations and for activities during a JAERI Foreign Researcher Inviting
Program [3-16], special models were added for HTTR-specific features not readily changed
via the input screens. These included a special model for the vessel cooling system (VCS or
Reactor Cavity Cooling System - RCCS), modeling of the core inlet flow path up the annulus
between the core and reactor vessel, accommodation of the smaller core/larger side reflector
design than is standard in GRSAC, and inclusion of an annealing model to account for
thermal conductivity changes in the higher grade graphite (GRSAC reference case is for
Magnox reactor type cores).
The RCCS (VCS) model used by ORNL previously for the HTTR-RCCS experiment
IAEA benchmark calculations (with success) was adapted to the actual HTTR VCS.
Preliminary calculations showed somewhat higher than expected heat removal rates.
Sensitivity studies (also preliminary) showed the predicted rates to be very sensitive to the
assumed emissivity values for the radiation shields, as would be expected. Also, it was not
clear how to estimate convection heat transfer (if any) between the radiation shields and the
cooling panel itself. The GRSAC model used for convection heat transfer (Keyhani)
predicted an increase in effective h due to the addition of separation baffles in the vessel-to
VCS annulus, which were installed to break up the convective flow path and reduce the heat
transfer coefficient value.
A significant improvement in the GRSAC model for the HTTR VCS was made for the
CRP-5 benchmarks on the basis of a study done by Akio Saikusa while at ORNL (1998-9) as
a visiting JAERI Research Fellow. Using the PATRAN P/Thermal code system, he created a
very detailed model of the upper vessel region, including the individual control rod drive
standpipes, and the upper portion of the VCS. The objective was to derive an accurate
210
effective view factor for this complex geometry that could be used in the GRSAC VCSvessel model. This effort was successful, and the results were incorporated.
3.2.4.2. Results of HTTR-VC
Previous V&V calculations of HTTR vessel cooling (VC) performance
The ORNL GRSAC code was used previously to predict steady state core and VC heat
removal conditions for initial power levels of 10%, 30%, 50%, 75%, and 100%. The GRSAC
calculations were benchmarked against the JAERI code ACCORD [3-16] with the results
generally in very good agreement. For this exercise, the axial peaking factor curve was
modified from the IAEA case to correspond to the no-burnup profile that would apply to the
initial rise to power. The 100% reference case core flow distribution was also modified so
that 9.8% of the total flow of 12.51 kg/s would bypass the active core.
The model of the fuel pin accounts for the outer sleeve, gap, and hollow-center fuel
geometry. The gap model was revised to give a reference case 0.25 mm gap with helium,
where the helium conductivity is included as a function of temperature, along with radiant
heat transfer also across the gap.
The GRSAC special model of the HTTR core consists of 61 radial regions
(representing the HTTRs 954 channels). The balance of GRSACs 205 radial regions
represent the removable and permanent side reflectors. There are 14 axial core nodes, where
the middle 10 represent the active core. The 10 axial peaking factors (inlet to outlet) are:
.5373 .6545 .8043 .9313 1.241 1.286 1.471 1.293 1.136 .6447
These values are very close to those derived from the CRP-5 VC and LP benchmark
descriptions, and are used in those calculations as well. The core model was also modified so
that the control rod channels are accounted for explicitly. Hence the axial flow modeling in
the 61 representative (GRSAC) active core model regions are no longer assumed to have
equal orificing characteristics. The 61 regions represent 37 assemblies (30 fuel elements and
7 control rod drive [CRD] elements), so some of the 61 regions partially represent the CRDs.
The center radial node in GRSAC is 100% CRD channel, but the next 3 rings of radial
nodes contain a mixture of CRD and fuel channel elements. The fourth ring consists of only
fuel element channels. To estimate the maximum temperatures in rings 2-4, the hot-channel
option in GRSAC is used, which allows one node in each of these rings to have an increased
peaking factor to represent a typical region in this ring where there is no CRD element.
Reference case V&V results - calculations
In Table 3-40 below, for each power level case, the predicted core mean coolant outlet
temperatures (one for the active core only, and the other with side reflector coolant mixed in
w/SR) are given, along with the maximum gas outlet (T-o max) and fuel temperatures (T-f
max). The maximum predicted fuel temperatures do not occur in the center radial region, as
they would in a homogeneous core representation. The peak fuel temperature for the 100%
power case appears in the outlet axial region for the active core (AR#12); for all other cases, it
is in AR#8 (just below mid-core). The maximum vessel temperature (T-v max) and rate of
heat removal from the vessel cooling system (P-vcs, in kW) are also shown (all temperatures
in degree C).
211
Table 3-40: VC Reference Case Calculations as a Function of % Power

%P
10
30
50
75
100
T-o active, w/SR

184-180
328-315
488-468
687-658
884-848
T-o max.
186
334
498
701
900
T-f max [J]

207 [179]
387 [375]
570 [560]
785 [788]
993 [1017]
T-v max.
120
159
210
273
336
P-vcs
100
180
280
460
650
The corresponding JAERI values (using the ACCORD code) for maximum fuel
temperatures for these cases are shown in the T-f max column in brackets [ ]. Considering the
differences in codes and many modeling assumptions, the result are in very good agreement,
the GRSAC temperature estimates being slightly lower for the full power case.
Sensitivity study
Several variations in parameters and assumptions were made to determine the
sensitivity of the predicted results to uncertainties. Since the effects would be most
pronounced at 100% power, they were run for that case. Two factors that could affect the
flow redistribution in the active core are the friction factor and pressure drop of the
orificing. A 50% increase in the assumed friction factor had no effect on flow
redistribution or resulting predicted temperatures. A 100% increase in the assumed core
pressure drop (from 0.12 kg/cm2 to 0.24 kg/cm2) also had a negligible effect on gas and fuel
temperatures.
An assumed 25% increase in the effective core heat transfer coefficient had some
effect on maximum fuel temperatures (the lower boundary of the given h-correlation curve
was used in the reference case). For this case, the maximum fuel temperature in axial region
12 was ~ 6C lower.
The nominal value of gap resistance was increased by 50%, increasing the maximum
fuel temperature in axial region 12 by ~ 25C.
A significant effect in predicted active core temperatures was seen for changes in the
assumed core bypass flow percentage (vs. the reference 9.8%). For a bypass of 5.3%, the
mean outlet gas temperature was 864 with a maximum of 873, and the maximum fuel
temperature (region 12) was ~ 27C less. For a bypass of 12.3%, the mean active core outlet
gas temperature was 897C with a maximum of 914C, and the maximum fuel was ~ 15C
more than for the reference case.
HTTR-VC V&V results experimental
Table 3-41 provides a comparison of steady state VCS measurements during the 1997
heat-up tests (circulator power was the heat source) with the reference model predictions:
Table 3-41: Measured vs. Predicted VC Results in kW
Run Time
1997/9/10 9:00
1997/9/11 13:00
1997/9/15 12:00
212
Measured kW
99
85
130
Predicted kW (GRSAC)
93
98
98
In the second case, the vessel temperatures are higher than those in the first, so the
model predicts a higher heat transfer rate. In the third case, the adjustment panel flow is on
and additional heat is removed. The model prediction does not include adjustment panel
cooling (39 kW measured).
There was a large uncertainty in the active core bypass flow in Fort St. Vrain, with
estimates up to about 15%. One reason for the large uncertainty was the variability due to the
positioning of the orifices controlling flow in each of the refueling regions (not a
consideration for HTTR). However, some uncertainty should be accounted for in the HTTR.
The cooler bypass gas exiting the side reflector and CRD channels mixes with the hot
gas from the active core fuel regions - and the differences in these are larger for the higher
bypass flow assumption cases. It may be possible to infer bypass flow percentages from outlet
temperature readings, and perhaps detect hot plenum mixing characteristics from fluctuations
in mixed (average) gas outlet temperature readings.
Steady-state benchmark calculations (HTTR-VC) for evaluation of vessel system
performance
The GRSAC code was run in the initial condition (steady-state) mode to determine
VCS heat removal rates and resulting vessel temperatures. For the 100% (30 MW) power
case, the predicted VCS heat removal rate was 0.670 MW. In this case the maximum vessel
temperature was 333C, and the average temperature of the vessel top head was estimated at
203C.
GRSAC was also run at 50% power (15 MW), giving a VCS power of 0.285 MW,
maximum vessel temperature of 209C, and a peak top head temperature of 128C. For this
case, the reactor inlet temperature was assumed to be 241.4C.
3.2.4.3. Results of HTTR-LP
Transient benchmark calculations for loss of off-site power (HTTR-LP) with scram and
post-trip cooling via the auxiliary cooling system
The GRSAC HTTR model was run in the accident mode for the LP benchmark,
representing a loss of off-site power (LOSP) event. The LP benchmark problem involves a
main circulator trip and flow coastdown, with a reactor trip (scram) and startup of the
Auxiliary Cooling System (ACS) shortly thereafter. The performance of the ACS for the LP
benchmark is approximated by running GRSACs simplified Shutdown Cooling System
(SCS) model with the prescribed helium cooling flow-rates (1.2 kg/s for the first 40 minutes
and 0.8 kg/s thereafter). An SCS cooling water flow of 18.3 kg/s is maintained throughout.
Primary pressure calculations are very approximate, since the HTTR balance of plant (BOP)
is not modeled explicitly. However, the results of the temperature predictions are not very
sensitive to pressure, and the GRSAC estimates of pressure changes were fairly close to those
shown in the JAERI results.
The results of the LOSP from 100% power are shown in Figures 3.51 and 3.52. They
show a gradual decline in core temperatures, with the rate of decline slowing somewhat after
the decrease in coolant flow after 40 minutes, as expected. In the GRSAC figures, the
Auxiliary Cooler heat removal rate (requested in the benchmark) is approximated by the core
heat removal rate (curve labeled Primary) in the figure.
213
FIG. 3.51. HTTR-LC Benchmark: LOSP from 100% Power (1 hr).
214
215
FIG. 3.55. HTTR LOSP with no auxiliary cooling flow (pressurized).
FIG. 3.56. HTTR LOSP with no auxiliary cooling flow (depressurized).

216
The results of the LOSP from 50% power are shown in Figure 3.53 and 3.54. They
also show a gradual decline in core temperatures, at a slower rate than the first case since the
Auxiliary Cooler is removing less heat. Both sets of results appear to be in good agreement
with the JAERI results published in the benchmark problem descriptions.
Figures 3.55 and 3.56 provide the results of HTTR loss of electric power even with no
auxiliary cooling flow in the pressurized and depressurized condition, respectively.
It can be seen that no temperature excursions are expected, and the cool-downs
proceed in an orderly fashion according to both GRSAC and JAERI (ACCORD) predictions.
217
3.2.5. France
A model of the High Temperature Engineering Test Reactor was developed for the
benchmark concerning the evaluation of the Performance of Vessel Cooling System during
normal operation. The CAST3M code [3-17] has been used to model the HTTR. CAST3M is
a multi-purpose finite element code developed at CEA, which allows 3D, 2D and R-Z
axisymmetric calculations. Domains of applications are structural mechanics, fluids
mechanics heat transfer and hydro-magnetics. As far as heat transfer is concern, the code is
able to take into account radiation phenomenon.
3.2.5.1. Relevant Hypothesis and Limits of the Model
The geometry
Considering on the one hand the design geometry and its symmetry and on the other
hand, the boundary conditions, a R-Z model has been chosen to describe the problem as
shown on Figure 3.57. The heat transfers in the core and the coolant channels are supposed to
be approximately 2D axisymmetric.
FIG. 3.57. 2D-reactor mesh.

Main correlations used for the reactor modelling
Helium hydraulics and heat transfer in the reactor core as well as in the reflectors, the
bottom replaceable reflector block and the hot plenum, are modelled using an equivalent
porous media formulation comprising a solid phase and a gaseous phase (Figure 3.58).
Indeed, the number of helium channels per blocks as well as the number of fuel holes disables
to use a complete description of the geometry that would be required for solving the
convection equation in the coolant holes and the conduction and radiation equations in the
graphite and in the gas.
218
In the porous medium, the relation between the pressure drop and the mass flow is
computed from the Blasius equation used for turbulent fully developed flow in a smooth pipe.
The heat exchange coefficient between the helium in convection inside the fuel element is
computed from the correlation given in [3-18]. It is nevertheless weighted by the channels
related specific surface. Moreover, the Dittus-Boelter equation is used for the heat transfer in
the reflector and in the plenum.
FIG. 3.58. Solid, helium and air related meshes.

As far as the bottom replaceable reflector is concern, the characteristics (hydraulic
diameter, porosity, specific surface) correspond to the average characteristics of the hot
plenum. It is assumed that this assumption wont influence strongly the benchmark results
(external temperature of the vessel, panel temperatures and power evacuated by the Vessel
Cooling System).
Concerning the solid phase of the porous medium in the core, the equivalent
horizontal thermal conductivities have been determined, as function of the temperature. In
order to evaluate these equivalent conductivities, a simplified model of the fuel block has
been used taken into account the radiation heat transfer across the channels.
Finally, neither the standpipe nor the thermal shield of the upper cavity, have been
considered in the model. The helium is modelling as an incompressible fluid. Although the
mass density variation is important, the Boussinesqs approximation is used. Besides, the
thermal conductivity as well as the viscosity is function of the temperature. In all the pressure
vessel cavities the radiation heat transfer is taken into account, the helium being transparent
for this kind of transfer.
The reactor cavity
Four ring plates are located in the reactor cavity (50 mm from the pressure vessel and
100 mm from the external panels). There is also a ring plate in the lower part of the reactor
(under the pressure vessel). As depicted in Figure 3.59, additional plates have been added at
the extremity of the ring plates [3-21]. Therefore, it is assumed that these ring plates share the
lower part of the reactor cavity in two independent hydraulic zones and the side reactor cavity
219
in five others hydraulic zones. In these zones, the air is considered as an incompressible fluid
and transparent for the radiative heat transfer. The Boussinesqs approximation is also used.
The thermal conductivity and the viscosity have been assumed constant and considered at
60C.
Additional
l
ISI plate
RPV
VCS
FIG. 3.59. Ring plate between the reactor and the panel.
The side and bottom panels
For the reactor cavity cooling system (VCS), the water temperature distribution is
supposed constant along the tubes. Calculations show that the axial flow is fully turbulent in
the tubes of the side panel (Re number about 104). Then, the Mc Adams correlation allows
evaluating the Nusselt number from which a local exchange coefficient has been deduced.
Considering the R-Z model an average exchange coefficient has been used for the whole
panel.
Thermal Reflector
Water Cooling Panel
gap 10 mm
VCS panel
surface
gap 10 mm
FIG. 3.60. Thermal reflector plates along the VCS.

In front of the tubes, there are some vertical thermal reflector plates (2 or 5). These
plates have gaps to manage the axial thermal expansion of the whole reflector. Nevertheless,
they enhance air circulation between the panels. As it is difficult to take into account the
presence of these gaps in the whole reactor model, some preliminary calculations have been
carried out on the thermal reflector zone.
A simplified model has been developed where all the reflector plates have been
described. In order to simulate the radiative heat transfer between the pressure vessel and the
220
thermal reflector, a part of the pressure vessel at constant temperature (300C) radiating to the
reflector is assumed as a boundary condition. In this simplified model, the emissivities
corresponding to those of the global model. On the side panel, the heat exchange is
represented by similar parameters than those used in the global reactor model. Three
calculations have been performed: the first taking into account radiation between the reflector
plates but without air circulation, the two last considering air circulation with and without the
gaps. The results are presented in Table 3-42.
Table 3-42. Refined Calculations on the Thermal Reflector
Configurations
Radiative Power
Without air circulation
116 kW
With air circulation and without gaps
124 kW
With air circulation and with gaps
131 kW
The air circulation observed inside the thermal reflector corresponds to an upward
flow in the front cavities (reactor side) and a downward flow in the cavity along the VCS
panel (cooling tubes side). The calculations show a more equilibrated flow in the central
cavities. By taking into account the gaps, the air velocities are enhanced and this leads to
lower temperature on the reactor side part of the thermal reflector. The extracted power is
therefore higher in this case. Finally, according to these preliminary results, the air circulation
has been considered in the final model without describing the gaps.
3.2.5.2. Results for the global model
Algorithm and finite element
The finite elements are linear quadrangle for the heat transfer, quadrangles isoQ2 and
isoP1 for respectively velocity and discontinuous pressure [3-19, 3-20]. A finite element
isoQ2 comprises 4 linear elements. The global model has 11580 elements for the solid parts,
2808 elements in the helium and 1513 in the air (Figure 3.58). Hydraulics and energy
transport equation are treated by an implicit scheme and simultaneously solved in order to get
a stable solution for the fluid-structure heat transfer. Two calculations have been performed
corresponding respectively to core power 9 MW and 30 MW (Table 3-43).
Table 3-43. Reactor Characteristics Used in the Calculations
Reactor Power
Helium Inlet Temperature
Flow Rate
Cooling Water Temperature of VCS
Helium Outlet Temperature
9
175
12.4
28.5
312
MW
C
kg/s
C
C
30
395
12.4
25
850
MW
C
kg/s
C
C
Benchmark Results
The surface temperature of the thermal reflector (reactor cavity side) is also given for
both envisaged power level. The outlet temperature of the helium ranges from 321 C at
9 MW to 863 C at nominal power. Moreover, the temperature distributions in the helium, the
air and the solid are provided in Figure 3.61 for the whole reactor. Finally, the Table 3-44
gathered the main results of the total power extract by the Vessel Cooling System in steady
state.
221
FIG. 3.61. Solid and gas temperature distributions for 9 MW.
Table 3-44. VCS Performance in Steady State

Power level
Upper panel
Side panel
Lower panel
Total extracted power
PRV Temp. (19-27 m)
9 MW
20
150
8
178
168
kW
kW
kW
kW
C
30 MW
60
480
15
555
340-353
kW
kW
kW
kW
C
3.2.5.3. Concluding remarks

The results obtained in the upper part of the cavity could be better. Indeed the:
222
Thermal reflector located to the top head dome (reactor pressure vessel) has not been
taken into account and lead to an overestimation of vessel temperature.
Standpipes were also not considered. They certainly slow down the air circulation and
modify the radiative exchanges between the divers walls.
Heat exchange in the upper part of the VCS tubes is assumed to be uniform
The additional plates to the ring plates constitute a thermal continuity between the
reactor pressure vessel and the thermal reflector of the VCS. The impact of this assumption on
the temperatures remains to be tackled.
Finally, the concrete, beyond the VCS tubes, has not been considered. Therefore, a part
of radiation heat coming from the vessel (directly or through the support panel of the VCS
tubes) must induce a temperature gradient in the concrete. These thermal leakages are not
taken into account in the present model.
223
REFERENCES TO CHAPTER 3
[3.1.] SAIKUSA, A., TACHIBANA, Y., and KUNITOMI, K., Benchmark Problems for Riseto-power Test of High Temperature Engineering Test Reactor in IAEA Coordinated
Research Program, (Part II. Evaluation of Performance of Vessel Cooling System),
JAERI, Oarai, Japan, 2000
[3.2.] IOKA, I. et al.: JAERI-M88-032 (1998) (in Japanese)
[3.3.] TAKEDA, T., NAKAGAWA, S., TACHIBANA, Y., TAKADA, E., and KUNITOMI,
K., Analytical Evaluation on Loss of Off-site Electric Power Simulation of the High
Temperature Engineering Test Reactor, JAERI, (Proc. 3rd RCM of CRP on Evaluation of
HTGR Performance, held at Oarai, March 2001), IAEA, Vienna, 2001, (CD-ROM)
[3.4.] TAKEDA, T., NAKAGAWA, S., FUJIMOTO, N., TACHIBANA, Y. and IYOKU, T.,
Data on loss of off-site electric power simulation tests of the High Temperature
Engineering Test Reactor, JAERI-Data/Code 2002-015, July 2002.
[3.5.] INTERNATIONAL ATOMIC ENERGY AGENCY, Heat transport and after heat
removal for gas cooled reactors under accident conditions, IAEA - TECDOC - 1163,
Vienna, Jan. 2001.
[3.6.] KUZAVKOV, N., SHEPELYEV, S, AFANASYEV, V. and NIKANOROV,
O.,"Benchmark Problem on the VCS Heat Removal (HTTRVC)", (Presented at Third
RCM of CRP on Evaluation of HTGR Performance, held at Oarai, March 2001). OKBM,
Russia, 2001
[3.7.] INTERNATIONAL ATOMIC ENERGY AGENCY, Heat Transport and After Heat
Removal for Gas Cooled Reactors under Accident Conditions, IAEA-TECDOC-1163,
Vienna, January 2001.
[3.8.] JAPAN ATOMIC ENERGY RESEARCH INSTITUTE, Analytical evaluations on loss
of off-site electric power simulation of high temperature engineering test reactor, JAERI
2000-16, March 2000
[3.9.] JAPAN ATOMIC ENERGY RESEARCH INSTITUTE, Information on the benchmark
problem from JAERI, Japan, April 20, 2000 and July 25, 2000.
[3.10.] GORBUNOV, V., GORBOV, A., KUZAVKOV, N., "Loss of Off-site Electric Power of
the HTTR", OKBM, (Proc. 3rd RCM of CRP on Evaluation of HTGR Performance at
Oarai Japan, March 2001) IAEA, Vienna, 2001 (CD-ROM).
[3.11.] EMSLIE, F., Flownet Nuclear Analysis of HTTR Thermo-hydraulics Benchmarks,
PBMR (Pty.) Ltd., South Africa, 2002
[3.12.] TAKEDA, T., et al., Analytical Evaluation on loss of off-site electric power simulation
of the High Temperature Engineering Test Reactor, JAERI 2000-016, Oarai, Japan,
2000.
[3.13.] BALL, S.J., MORECA: A Computer Code for Simulating Modular HTGR Core Heatup
Accidents, NUREG/CR-5712, ORNL/TM-11823, ORNL, USA, October 1991.
[3.14.] BALL, S.J., HTTR Vessel Cooling (VC) and Loss-Of-Offsite-Power (LP) Benchmark
Solutions, ORNL, (Proc. Third RCM of CRP on Evaluation of HTGR Performance, held
at Oarai, March 2001), IAEA, Vienna, 2001, (CD-ROM)
[3.15.] BALL, S. J. and NYPAVER, D.J., GRSAC Users Manual, Oak Ridge National
Laboratory, ORNL/TM-13697, Feb. 1999
224
[3.16.] BALL, S. J., NAKAGAWA, S., SAIKUSA, A., TACHIBANA, Y., TAKEDA, T.,
HONTANI, K., OUCHI, K., and KUNITOMI, K., Independent Analysis: HTTR Riseto-Power and Test Analysis, Japan Atomic Energy Research Institute, JAERI-memo-09258, Dec. 1997
[3.17.] PAILLERE, H. and MAGNAUD, J.P., A finite element elliptic flow solver for low
Mach number compressible flows, (Proc. Tenth International Conference on Finite
Elements in Fluids, held in Tuscon, Arizona, January, 1998), M. Hafez/J.C. Heinrich,
1998.
[3.18.] SAIKUSA, A., TACHIBANA, Y. and KUNITOMI, K., Benchmark Problem for Riseto-Power Test of HTTR, Evaluation of the Performance of VCS, JAERI, Oarai, Japan,
1990
[3.19.] BOLAND, J. and NICOLAIDES, R.A., Stability of finite elements under divergence
constraint, SIAM Jour. Num. Ana. 20, pp.722-730, 1983
[3.20.] GUNZBURGER, D., Finite elements methods for viscous incompressible flows: a guide
to theory, practice and algorithms, Academic Press, pp. 28-31, 1989
[3.21.] SAIKUSA, A., Personal communication, JAERI, Japan, Feb.13, 2002.
225

TE 1382 Part1

Uploaded by

Copyright:

Available Formats

TE 1382 Part1

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

TE 1382 Part1

Uploaded by

Copyright:

Available Formats

IAEA-TECDOC-1382

Evaluation of high temperature

The originating Section of this publication in the IAEA was:

EVALUATION OF HIGH TEMPERATURE GAS COOLED REACTOR PERFORMANCE:

Institute of Energy Technology (INET), Beijing, China

4.2.10 South Africa ....................................................................................................... 314

the neutronic physics behaviour of the HTGR core,

Reactor physics benchmark analysis of the HTR-10 including initial criticality,

Establish and upgrade the technology base of the HTGR

FIG. 2.1. HTTR plant arrangement.

Contain fission products (FPs) as one of the multiple barriers against FP

The reactor pressure vessel (RPV) is formed as a vertical cylinder, with a

FIG. 2.2. HTTR reactor building.

FIG.2.3. HTTR pressure vessel and internals.

FIG. 2.4. HTTR cooling system.

Table 2-1. Specification of the HTTR

FIG.2.5. Vertical view of the HTTR.

FIG. 2.7. HTTR fuel column name and zone number.

FIG. 2.8: Structure of fuel assembly.

Fuel loading scheme

FIG. 2.9. Fuel assembly arrangement in the HTTR core.

FIG. 2.11. Fuel rod.

FIG. 2.12. Fuel block for 33 pin fuel assembly.

FIG. 2.13. Fuel graphite block for 31.pin fuel assembly.

FIG. 2.14. Control rod guide block in layers 1 through 5.

FIG. 2.15. Permanent reflector region.

FIG. 2.16. Schematic of control rod.

FIG. 2.17. Axial control rod position.

FIG. 2.19. Vertical location of temporary neutron detector system.

2.1.2. Benchmark Problem Descriptions

2.1.2.2. Initial Criticality [HTTR-FC] (Phase 2)

2.1.2.6. Isothermal Temperature Coefficient (HTTR-TC)

2.2. REACTOR PHYSICS BENCHMARK ANALYSIS AND RESULTS

Nucl. Data file

Cut off energy

Core cal. Code

No. of groups, (Fast +Thermal)

FIG. 2.21. Fuel cell model for DELIGHT code.

The group constants for BP and CR were calculated by TWOTRAN-II code.

FIG. 2.22. BP cell model by TWOTRAN-II

FIG. 2.23. Comparsion of CR model concrpt.

FIG. 2.24. Configuration of regions in block for core calculation mode

FIG. 2.25. Horizontal cross-section of 18 column core model.

FIG. 2.26. Horizontal cross-section of 24 column core model.

FIG.2.27. Horizontal cross-section of 30 column core model (fully-loaded core).

FIG: 2.28. Magnification of analytical model.

2.2.1.2. Results of HTTR-FC calculation

FIG. 2.30. Change in keff at fuel loading.

HTTR-FC by Monte Carlo

Table 2-6. Heterogeneity effects of CFP

FIG.2.31. Change in excess reactivity with fuel loading.

18 columns (thin annular core)

Control rod position at

HTTR-CR by Monte Carlo calculation

18 (Corrected homo. model)

24 (Corrected homo model)

HTTR-EX by Monte Carlo calculation

Model and column number

Excess reactivity (%k/k)