INL/EXT-08-14799

Heat Exchanger Design

Options and Tritium
Transport Study for the
VHTR System

Chang H. Oh
Eung S. Kim

September 2008

The INL is a U.S. Department of Energy National Laboratory

operated by Battelle Energy Alliance
 INL/EXT-08-14799

Heat Exchanger Design Options and Tritium Transport

Study for the VHTR System

Chang H. Oh
Eung S. Kim

September 2008

Idaho National Laboratory

Next Generation Nuclear Plant Project
Idaho Falls, Idaho 83415

http://www.inl.gov

Prepared for the

U.S. Department of Energy
Office of Nuclear Energy
Under DOE Idaho Operations Office
Contract DE-AC07-05ID14517
 DISCLAIMER
This information was prepared as an account of work sponsored by an
agency of the U.S. Government. Neither the U.S. Government nor any
agency thereof, nor any of their employees, makes any warranty, expressed
or implied, or assumes any legal liability or responsibility for the accuracy,
completeness, or usefulness, of any information, apparatus, product, or
process disclosed, or represents that its use would not infringe privately
owned rights. References herein to any specific commercial product,
process, or service by trade name, trade mark, manufacturer, or otherwise,
does not necessarily constitute or imply its endorsement, recommendation,
or favoring by the U.S. Government or any agency thereof. The views and
opinions of authors expressed herein do not necessarily state or reflect
those of the U.S. Government or any agency thereof.
vi
 ABSTRACT
This report presents the results of a study conducted to consider heat
exchanger options and tritium transport in a very high temperature reactor
(VHTR) system for the Next Generation Nuclear Plant Project.
The heat exchanger options include types, arrangements, channel patterns in
printed circuit heat exchangers (PCHE), coolant flow direction, and pipe
configuration in shell-and-tube designs. Study considerations include: three types
of heat exchanger designs (PCHE, shell-and-tube, and helical coil); single- and
two-stage unit arrangements; counter-current and cross flow configurations; and
straight pipes and U-tube designs in shell-and-tube type heat exchangers.
Thermal designs and simple stress analyses were performed to estimate the heat
exchanger options, and the Finite Element Method was applied for more detailed
calculations, especially for PCHE designs. Results of the options study show that
the PCHE design has the smallest volume and heat transfer area, resulting in the
least tritium permeation and greatest cost savings. It is theoretically the most
reliable mechanically, leading to a longer lifetime. The two-stage heat exchanger
arrangement appears to be safer and more cost effective. The recommended
separation temperature between first and second stages in a serial configuration is
800oC, at which the high temperature unit is about one-half the size of the total
heat exchanger core volume. Based on simplified stress analyses, the high
temperature unit will need to be replaced two or three times during the plant’s
lifetime. Stress analysis results recommend the off-set channel pattern
configuration for the PCHE because stress reduction was estimated at up to 50%
in this configuration, resulting in a longer lifetime.
The tritium transport study resulted in the development of a tritium behavior
analysis code using the MATLAB Simulink code. In parallel, the THYTAN code,
previously performed by Ohashi and Sherman (2007) on the Peach Bottom data,
was revived and verified. The 600 MWt VHTR core input file developed in
preparation for the transient tritium analysis of VHTR systems was replaced with
the original steady-state inputs for future calculations. A Finite Element Method
analysis was performed using COMSOL Multiphysics software to accurately
predict tritium permeation through the PCHE type heat exchanger walls. This
effort was able to estimate the effective thickness for tritium permeations and
develop a correlation for general channel configurations, which found the
effective thickness to be much shorter than the average channel distance because
of dead spots on the channel side.

vii
viii
 ACKNOWLEDGMENT
This work was supported through the U.S. Department of Energy’s
NGNP-Engineering Program under DOE Idaho Operations Office Contract
DE-AC07-05ID14517.

ix
x
 CONTENTS
ABSTRACT................................................................................................................................................ vii

ACKNOWLEDGMENT.............................................................................................................................. ix

ACRONYMS............................................................................................................................................ xvii

1. HEAT EXCHANGER DESIGN OPTIONS FOR THE NEXT GENERATION NUCLEAR

PLANT ............................................................................................................................................... 1
1.1 Introduction.............................................................................................................................. 1
1.2 Heat Exchanger Thermal Design ............................................................................................. 2
1.2.1 Heat Exchanger Requirements and Reference Reactor Configurations...................... 2
1.2.2 Heat Exchanger Design Conditions ............................................................................ 5
1.2.3 IHX and SHX Thermal Design ................................................................................... 7
1.3 Stress Analysis of the VHTR Heat Exchangers ..................................................................... 28
1.3.1 Simplified Stress Analysis Method........................................................................... 28
1.3.2 Simplified Stress Evaluation for IHX and SHX ....................................................... 31
1.4 Computational Analysis on the Printed Circuit Heat Exchanger Channel
Configuration Options............................................................................................................ 38
1.4.1 COMSOL Modeling for Heat Transfer and Tritium Penetration in the PCHE......... 38
1.4.2 ABAQUS Modeling for Stress in the PCHE ............................................................ 51

2. TRITIUM BEHAVIOR STUDY IN THE VHTR SYSTEM ........................................................... 67

2.1 Introduction............................................................................................................................ 67
2.1.1 Basic Equations......................................................................................................... 69
2.1.2 Development of Dynamic Tritium Behavior Analysis Code by MATLAB
Simulink .................................................................................................................... 74
2.2 Follow-up of THYTAN Verification Study........................................................................... 88
2.2.1 Verification of Core Model....................................................................................... 89
2.2.2 Verification of Permeation Model ............................................................................ 95
2.2.3 Verification of Leak Model....................................................................................... 96
2.2.4 Verification of Purification System Model ............................................................... 97
2.3 Development of THYTAN Core Input for Transient Analysis of VHTR Tritium
Behaviors ............................................................................................................................... 98
2.3.1 Input Parameters for Tritium Source from Ternary Fission...................................... 99
2.3.2 Input Parameters for Tritium Source from 3He in the Coolant.................................. 99
2.3.3 Input Parameters for Tritium Source from Li in Graphite ...................................... 101
2.3.4 Input Parameters for Tritium Source from 10B in Graphite..................................... 102
2.4 Evaluation of the Effective Thickness for Tritium Permeation in PCHEs.......................... 102

3. SUMMARY ................................................................................................................................... 107

3.1 Design Options for NGNP Heat Exchanger......................................................................... 107
3.1.1 Heat Exchanger Thermal Design and Design Options............................................ 107
3.1.2 Stress Analysis of the VHTR Heat Exchangers ...................................................... 108
3.1.3 Computational Analysis of PCHE Channel Configuration Options ....................... 108
3.2 Tritium Behavior Study in the VHTR System ..................................................................... 109
3.2.1 Heat Exchanger Thermal Design and Design Options............................................ 109

xi
4. REFERENCES ............................................................................................................................... 110

Appendix A—High Temperature Heat Exchanger Selection and Design Guideline ............................... 114

Appendix B—Thermal Design Method of Helical Coil Heat Exchanger................................................. 130

FIGURES
Figure 1-1. Configuration 1—indirect parallel cycle.................................................................................... 3
Figure 1-2. Configuration 2—indirect serial cycle. ...................................................................................... 4
Figure 1-3. HTSE system.............................................................................................................................. 4
Figure 1-4 Design space with designed regions............................................................................................ 5
Figure 1-5. Potential optimal solutions. ........................................................................................................ 5
Figure 1-6. LMTD of IHX vs. system overall efficiency for the reference reactor
(Configuration 1). ......................................................................................................................... 6
Figure 1-7. Pressure drop (kPa) vs. system efficiency (%) in the reference reactors
(Configuration 1). ......................................................................................................................... 6
Figure 1-8. Operating condition of two-stage IHX..................................................................................... 13
Figure 1-9. Variations of Sm and St of alloy 617 with temperature and time (Natesan et al. 2006)............ 28
Figure 1-10. Comparisons of failure theories for a biaxial state of stress (Martin 1962). .......................... 30
Figure 1-11. Comparisons of biaxial strength data with failure theories (Collins 1981)............................ 30
In this report, the maximum shear stress theory was used for conservative failure estimations................. 31
Figure 1-12. Smt (MPa) vs. life time (h) for Alloy 617. .............................................................................. 31
Figure 1-13. Maximum stress as a function of plate thickness-to-diameter ratio (PCHE, 900oC). ............ 33
Figure 1-14. Maximum stress as a function of plate thickness-to-diameter ratio (PCHE, 800°C). ............ 33
Figure 1-15. Maximum stress as a function of plate thickness-to-diameter ratio (PCHE, 750°C). ............ 34
Figure 1-16. Maximum stress as a function of tube thickness-to-radius ratio (shell-and-tube,
900°C)......................................................................................................................................... 34
Figure 1-17. Maximum stress as a function of tube thickness-to-radius ratio (shell-and-tube,
800°C)......................................................................................................................................... 35
Figure 1-18. Maximum stress as a function of tube thickness-to-radius ratio (shell-and-tube,
750°C)......................................................................................................................................... 35
Figure 1-19. Plate thickness-to-diameter ratio vs. estimated heat exchanger lifetime (PCHE). ................. 37
Figure 1-20. Tube thickness-to-inner radius ratio vs. estimated heat exchanger lifetime (shell-
and-tube)..................................................................................................................................... 37
Figure 1-21. PCHE channel configurations. ............................................................................................... 39
Figure 1-22. Calculated temperature and tritium concentration profiles (for the standard
configuration). ............................................................................................................................ 46

xii
Figure 1-23. Calculated temperature and tritium concentration profiles (for the off-set
configuration). ............................................................................................................................ 47
Figure 1-24. Diffusion flux and path of tritium in the PCHE (for the standard configuration). ................. 48
Figure 1-25. Diffusion flux and path of tritium in the PCHE (for the off-set configuration). .................... 49
Figure 1-26. Tritium flux for different horizontal pitches. ......................................................................... 50
Figure 1-27. Tritium flux for different vertical pitch (plate thickness)....................................................... 51
Figure 1-28. Normal (left) and offset (right) channel configurations. ........................................................ 52
Figure 1-29. Stress versus time-to-1% strain and time to rupture from creep for several
temperatures (Schubert 1984, p. 328)......................................................................................... 53
Figure 1-30. ABAQUS models for different numbers of HTUs................................................................. 54
Figure 1-31. Close up view of a channel corner. ........................................................................................ 55
Figure 1-32. The 25 HTU model. ............................................................................................................... 56
Figure 1-33. Mesh grid around a half channel. ........................................................................................... 56
Figure 1-34. Maximum stress as function of heat transfer units used......................................................... 57
Figure 1-35. Stress distribution of 25 HTUs model, entire model view. .................................................... 58
Figure 1-36. Stress distribution of 25 HTUs model, top three rows, second and third columns. ............... 59
Figure 1-37. Stress distribution of 25 HTUs model, left corner of second row, second column
channel view. .............................................................................................................................. 59
Figure 1-38. Maximum stress as function of the outside pressure for the normal configuration. .............. 60
Figure 1-39. Maximum stress as function of the vertical pitch for normal and shifted
configurations. ............................................................................................................................ 61
Figure 1-40. Normal 0.8 mm vertical pitch. ............................................................................................... 62
Figure 1-41. Shifted 0.8 mm vertical pitch. ................................................................................................ 62
Figure 1-42. Normal 2.4 mm vertical pitch. ............................................................................................... 63
Figure 1-43. Shifted 2.4 mm vertical pitch. ................................................................................................ 63
Figure 1-44. Maximum stress as function of the horizontal pitch for normal and shifted
configurations. ............................................................................................................................ 64
Figure 1-45. Normal 1.264 mm horizontal pitch. ....................................................................................... 65
Figure 1-46. Shifted 1.264 mm horizontal pitch. ........................................................................................ 65
Figure 1-47. Normal 2.6 mm horizontal pitch. ........................................................................................... 66
Figure 1-48. Shifted 2.6 mm horizontal pitch............................................................................................. 66
Figure 2-1. Internal structure of the Mass Balance Model Block. .............................................................. 77
Figure 2-2. Internal block diagram of Core Model Block. ......................................................................... 78
Figure 2-3. Parameter setting window for the tritium source from tertiary fission..................................... 79
Figure 2-4. Submodel for tritium source from tertiary fission.................................................................... 79
Figure 2-5. Submodel for tritium source from Birth from 6Li. ................................................................... 80

xiii
Figure 2-6. Submodel for tritium source from Birth from 7Li. ................................................................... 80
Figure 2-7. Submodel for tritium source from Birth from 7He. .................................................................. 81
Figure 2-8. Submodel for tritium source from Birth from 10B.................................................................... 81
Figure 2-9. Internal block diagram of Reactor Block. ................................................................................ 84
Figure 2-10. Internal block diagram of the Heat Exchanger Block. ........................................................... 86
Figure 2-11. GUI and workspace of the current tritium behavior analysis code. ....................................... 88
Figure 2-12. Primary coolant system of the Peach Bottom reactor (Wichner and Dyer 1979). ................. 89
Figure 2-13. Comparison of tritium concentrations in the containment vessel of the Peach Bottom
reactor. ........................................................................................................................................ 96
Figure 2-14. Comparisons between analytical solution and THYTAN results for verification of
the purification system................................................................................................................ 98
Figure 2-15. PCHE channel configurations. ............................................................................................. 104
Figure 2-16. Effective thickness for various channel dimensions (diameter, pitch and thickness). ......... 105

TABLES
Table 1-1. Heat exchanger design conditions for Configuration 1 (Figure 1-1)........................................... 7
Table 1-2. Heat exchanger design conditions for Configuration 2 (Figure 1-2)........................................... 7
Table 1-3. Single-stage IHX thermal design (PCHE)................................................................................... 9
Table 1-4 Single stage IHX thermal design (shell-and-tube)...................................................................... 10
Table 1-5. Single stage IHX thermal design (helical coil).......................................................................... 11
Table 1-6. Two-stage IHX thermal design (high temperature section, PCHE). ......................................... 14
Table 1-7. Two-stage IHX thermal design (high temperature section, shell-and-tube).............................. 15
Table 1-8. Two-stage IHX thermal design (high temperature section, helical coil). .................................. 15
Table 1-9. Two-stage IHX thermal design (low temperature section, PCHE). .......................................... 16
Table 1-10. Two-stage IHX thermal design (low temperature section, shell and tube).............................. 17
Table 1-11. Two-stage IHX thermal design (low temperature section, helical coil). ................................. 17
Table 1-12. Summary of heat exchanger types in the two-stage IHX in series. ......................................... 18
Table 1-13. Volume of heat exchangers for two-stage IHX. ...................................................................... 18
Table 1-14. Heat transfer area of two-stage IHX........................................................................................ 19
Table 1-15. Two-stage IHX thermal design (high temperature section, PCHE, Tsep = 800oC). ................. 20
Table 1-16. Two-stage IHX thermal design (high temperature section, shell-and-tube, Tsep =
800°C)......................................................................................................................................... 21
Table 1-17. Two-stage IHX thermal design (high temperature section, helical coil, Tsep = 800°C)........... 21
Table 1-18. Two-stage IHX Thermal design (low temperature section, PCHE, Tsep = 800°C).................. 22
Table 1-19. Two-stage IHX thermal design (low temperature section, shell-and-tube, Tsep =
800°C)......................................................................................................................................... 23

xiv
Table 1-20. Two-stage IHX thermal design (low temperature section, helical coil, Tsep = 800°C)............ 23
Table 1-21. Volume of heat exchangers for Two-stage IHX (800°C separation)....................................... 24
Table 1-22. Heat transfer area of two-stage IHX (800°C separation)......................................................... 24
Table 1-23. SHX thermal design (PCHE)................................................................................................... 26
Table 1-24. SHX thermal design (shell and tube)....................................................................................... 27
Table 1-25. SHX thermal design (Helical). ................................................................................................ 27
Table 1-26. Assumed allowable stresses for Alloy 617.............................................................................. 31
Table 1-27. Estimated life time of reference heat exchangers. ................................................................... 32
Table 1-28. Required thickness for IHX (for 'Pmax = 1.0 MPa)................................................................. 36
Table 1-29. Required thickness for SHX (for 'Pmax = 2.0 MPa). ............................................................... 36
Table 1-30. Required thickness for SHX (for 'Pmax = 5.0 MPa). ............................................................... 36
Table 1-31. Calculated diffusion coefficients in the PCHE. ....................................................................... 41
Table 1-32. Calculated heat transfer coefficients in the PCHE. ................................................................. 41
Table 1-33. Default channel dimensions and operating conditions. ........................................................... 52
Table 1-34. Alloy 617 mechanical properties (Special Metals Corporation, 2006, p. 2-3). ....................... 53
Table 1-35. Meshing scheme for HTU study.............................................................................................. 55
Table 1-36. Meshing scheme for vertical pitch and horizontal pitch studies.............................................. 61
Table 2-1. Comparison of tritium activity by ternary fission...................................................................... 90
Table 2-2. Comparison of tritium activity from 6Li at the Core 2 operation of the Peach Bottom
reactor. ........................................................................................................................................ 91
Table 2-3. Parameters for the calculation of tritium birth from 6Li (Ohashi and Sherman 2007). ............. 91
Table 2-4. Input data of THYTAN for calculation of tritium birth from 6Li (Ohashi and Sherman
2007)........................................................................................................................................... 91
Table 2-5. Comparison of tritium activity from 3He at the Core 2 operation of the Peach Bottom
HTGR. ........................................................................................................................................ 93
Table 2-6. Input data of THYTAN for calculation of tritium birth from 3He (Ohashi and Sherman
2007)........................................................................................................................................... 93
Table 2-7. Input data for the subnode of THYTAN for the calculation of tritium birth from 3He
(Ohashi and Sherman 2007). ...................................................................................................... 94
Table 2-8. Comparison of tritium activity from 10B in the poisoned spine at the Core 2 operation
of Peach Bottom reactor. ............................................................................................................ 95
Table 2-9. The heat transfer tube dimensions of the steam generator in the Peach Bottom reactor
(Ohashi and Sherman 2007). ...................................................................................................... 95
Table 2-10. Comparison of permeation rate. .............................................................................................. 96
Table 2-11. Input data of THYTAN for verification of the leak model (Ohashi and Sherman
2007)........................................................................................................................................... 97

xv
Table 2-12. THYTAN boundary conditions for verification of the purification system model
(Ohashi and Sherman 2007). ...................................................................................................... 98
Table 2-13. Tritium production reaction and cross section (Wichner 1979). ............................................. 99
Table 2-14. Basic data for THYTAN evaluation of ternary fission............................................................ 99
Table 2-15. Basic data for THYTAN evaluation of 3He Source............................................................... 100
Table 2-16. Parameters used for estimating tritium production from 3He. ............................................... 100
Table 2-17. Parameters used for estimating tritium production from Li. ................................................. 101
Table 2-18. Parameters used for estimating tritium production from Li. ................................................. 102
Table 2-19. Effective Thickness over a Temperature Range. ................................................................... 104

xvi
 ACRONYMS
ASME American Society of Mechanical Engineers
CFD computational fluid dynamics
FEM Finite Element Method
GCI grid convergence index
GUI graphical user interface
HTE high-temperature electrolysis
HTR-10 High Temperature Reactor (10 MW)
HTGR high temperature gas-cooled reactor
HTSE high temperature steam electrolysis
HTTR high temperature test reactor
HTU heat transfer unit
IHX intermediate heat exchanger
LMTD log mean temperature difference
NGNP Next Generation Nuclear Plant
PCU power conversion unit
PCHE printed circuit heat exchanger
PHX process heat exchangers
SHX secondary heat exchanger
VHTR Very High Temperature Gas-Cooled Reactor

xvii
xviii
Heat Exchanger Design Options and Tritium Transport
Study for the VHTR System
1. HEAT EXCHANGER DESIGN OPTIONS FOR THE NEXT
GENERATION NUCLEAR PLANT

1.1 Introduction
One Next Generation Nuclear Plant (NGNP) reference concept consists of a very high temperature
gas-cooled Reactor (VHTR) that provides the first demonstration of a closed-loop Brayton cycle at a
commercial scale of a few hundred megawatts electric and hydrogen production. The power conversion
system for the NGNP will take advantage of the significantly higher reactor outlet temperatures to
provide higher efficiencies than can be achieved in the current generation of light water reactors. Besides
demonstrating a system design that can be used directly for subsequent commercial deployment, the
NGNP will demonstrate key technological elements that can be used in subsequent advanced power
conversion systems for other Generation IV reactors. In anticipation of the design, development, and
procurement of an advanced power conversion system for the NGNP, the system integration of NGNP
with a hydrogen plant was initiated to identify the important design and technology options that must be
considered in evaluating the performance of the proposed NGNP.
Integrating the VHTR and hydrogen production plant will require an intermediate heat exchanger
(IHX) to transfer the process heat from the VHTR to the hydrogen plant. The selection, design, and
configuration of this IHX is important because its effectiveness is directly related to the overall efficiency
of the system, the gas coolant has poor heat transfer capabilities, and it is operating under critical
conditions (900°C, 7 MPa).
The main purpose of this study was to investigate and estimate the various heat exchanger options for
application with VHTRs. Three different types of heat exchangers were considered: printed circuit heat
exchanger (PCHE), shell-and-tube heat exchanger, and helical coil heat exchanger. A PCHE is a compact
heat exchanger, and a reference design is manufactured by Heatric™ through a chemical etching and
diffusion bonding process. The PCHE provides large heat transfer surface area density (as high as 2,500
m2/m3) that allows operation at very high temperatures and pressures (up to 900°C and 50 MPa). This
high compactness implies an appreciable reduction in material, reducing cost. A shell-and-tube heat
exchanger is the most classical and common type of heat exchanger, generally used in oil refineries and
other large chemical processes. It consists of a shell with a bundle of tubes inside of it; one fluid runs
through the tubes and the other flows over the tubes. This heat exchanger is suitable for high pressure
applications. The helical coil heat exchanger is a variation of the shell-and-tube heat exchanger. It is
composed of helically coiled tubes and a shell that usually provide higher heat transfer enhancement and
reliability during thermal expansion than the classical shell and tube design.
This study investigated two IHX arrangements: single-stage and two-stage (Patterson 2007), the
single-stage being the most common. In the single-stage concept, one big IHX unit is connected to the
reactor’s primary side and secondary side to exchange heat between them. Sometimes, the size of the IHX
is almost the same scale as the reactor vessel or power conversion units. In the two-stage concept, we split
the IHX into high temperature and low temperature sections. The low temperature section, downstream of
the high temperature section connected to the VHTR outlet, is designed for the full lifetime, and the high
temperature section for replacement within plant lifetime. The maximum temperature of the low
temperature section is around 760°C, which allows use of American Society of Mechanical Engineers
(ASME) Section III, Subsection NH materials such as Alloy 800H. In the high temperature section,
potential candidates include alloys 617 and 230 or ceramics. The two-stage method is expected to have

1
some cost, safety, and maintenance advantages. For example, if the IHX is split into two parts, the
material cost can be saved because relatively inexpensive commercial materials like stainless steel or
Alloy 800H can be used for the low temperature section—only the high temperature heat exchanger
requires expensive alloy or ceramic materials. In addition, having two heat exchangers will reduce the
thermal stress problem, because the temperature differences in the two component concept will be
reduced. It is also easier to manage and reduce the risk, since the safety problems are concentrated mainly
on the high temperature heat exchanger side operating in more severe conditions.
The overall organization of this section is as follows:
x Section 1.2 presents the results of extensive comparisons in the relative size and thermal performance
of the three heat exchangers.
x Section 1.3 presents the results of simple stress analyses carried out order to estimate the lifetime and
mechanical performance of the heat exchanger options.
x Section 1.4 focuses on the heat performance and mechanical stress of the PCHE, which are calculated
in detail by Finite Element Method (FEM) using ABAQUS code.

1.2 Heat Exchanger Thermal Design

As a first step, IHX thermal design was performed for various heat exchanger types and
arrangements. The process consisted of determining the IHX requirements, determining the IHX design
conditions for some reference reactor configurations, and performing thermal design of the IHX based on
the design conditions. Various design options were considered. The design methods and guidelines used
in this report are summarized in Appendix A.

1.2.1 Heat Exchanger Requirements and Reference Reactor Configurations

In order to design an IHX for VHTR applications, the following reference design requirements and
conditions were determined:
x Maximum operating temperature: ~1000°C
x Pressure: 50–100 bar
x Duty: 600 MWt
x Effectiveness: 95%
x Helium environment
x Low pressure drop
x 30–60 year design life.
The design conditions of the IHX were obtained from the previous research report by Oh et al.
(2006). A number of VHTR system configurations are recommended in this report. Among them, two
configurations were selected: indirect parallel system, and indirect serial system. In the indirect parallel
system (see Figure 1-1), the flow in the secondary coolant system is divided, with most of the flow going
towards the power conversion unit (PCU) and the remainder going through a secondary heat exchanger
(SHX) that directs heat towards the high temperature steam electrolysis (HTSE) plant. The flow through
the hot side of the SHX is then mixed with the flow from the PCU to feed the cold side of the IHX.
However, some of the flow is diverted away from the PCU, which acts to decrease the efficiency of the
cycle.

2
 Compressor

Nuclear
Reactor

IHX

Mixer
Generator

Turbine
SHX

Compressor
Recuperator

Precooler
MPC

Circulator

LPC

Figure 1-1. Configuration 1—indirect parallel cycle.

There are three coolant loops. The primary coolant system contains the nuclear reactor, the hot side of
the IHX, and a compressor. The secondary coolant system contains the cold side of the IHX, the hot side
of the SHX, the PCU, and connecting piping, which is assumed to be short. The intermediate heat
transport loop connects the secondary coolant system to the HTSE plant through several process heat
exchangers (PHXs).
In the indirect serial configuration (see Figure 1-2), the SHX is located upstream of the IHX that is
linked to the PCU. Therefore, the heat from the VHTR is transferred to the HTSE system first, and then to
the PCU. This configuration is able to supply higher temperatures to the HTSE system, but decreases the
PCU maximum temperature, resulting in a decreased PCU efficiency. However, in this configuration, the
system is more controllable because of its decreased connectivity. The reduction in the number of
circulators can reduce the cost and increase the overall efficiency. The same HTSE system configuration
shown in Figure 1-3 is used here in terms of coupling the VHTR and the HTSE.
Figure 1-3 shows the High Temperature Steam Electrolysis (HTSE) system. For electrolysis, the
steam is heated up to 830°C by the SHX. The heated steam is converted into hydrogen and oxygen in the
electrolyzer and discharged through the fuel and oxidizer outlet, respectively, as shown in Figure 1-2. The
heat of the discharged gases is recovered through three recuperators. The product gas in the fuel side
contains hydrogen and steam, and the oxidizer outlet gas contains oxygen and steam. Hydrogen and steam
separated from the electrolyzer are recycled back to the separator where hydrogen is separated as a
product at the top of the separator and steam is fed back to the electrolyzer. A small fraction of hydrogen
is recycled back to the electrolyzer to minimize the nickel oxidation in the hydrogen side of the
electrolyzer. Stream heat is first recuperated in the oxidizer outlet and then run through an expander to
recover work. The oxygen and water components of the stream are then separated.

3
 Figure 1-2. Configuration 2—indirect serial cycle.

Figure 1-3. HTSE system.

4
1.2.2 Heat Exchanger Design Conditions
In this section, IHX design conditions were determined for the thermal design. Two parameters are
important in performing this: the log mean temperature difference (LMTD) and the maximum pressure
drop in the reactor channel. If we decrease the LMTD of the IHX, we can obtain higher thermal
efficiencies, but it requires a large heat exchanger. Similarly, if we increase the maximum pressure drop
in the IHX, we can make the IHX smaller, but this results in a significant efficiency drop. Therefore, to
adequately design the IHX, these two parameters need to be determined very carefully. The process we
used in determining these parameters is discussed below.
In order to determine the IHX design conditions, the IHX was considered in view of its overall
system performance. HYSYS process modeling software was used (ASPEN 2005) to optimize this
analysis. HYSYS has a built in optimization tool to integrate the simulation model of the system. The
optimization program searches for the maximum value of a given objective function subject to a number
of imposed constraints. Figure 1-4 shows a two-dimensional (2-D) design space with defined regions. The
goal is to maximize f (x1, x2, x3,...) where x1, x2, x3 are independent variables such as mass flow,
pressure, temperature, etc. For the optimization, the x variables are manipulated within a specified range
of a lower and upper bound. The regions in Figure 1-5 are defined by a feasible design space within
functional constraints and an infeasible design space outside of the constraint boundaries (Mckellar 1992).
Functional constraints are material and energy balances such as positive pressure drop in every stream in
the direction of flow, positive power in turbine and compressor, temperature requirements at the inlet and
outlet of the heat exchangers, etc. The function, f, is the objective function defined as the overall plant
efficiency above. Constant values of the objective function define contours on the design space.
Figure 1-5 shows the optimal design points of the 2-D design vector. First, the initial calculation should
be in the feasible design space and the search continues towards a direction in the design space until a
maximum is reached. A new direction is found and the search continues in that direction as long as the
objective function value increases. Once a maximum is reached, the search continues towards the optimal
design point. However, if the optimal design point is outside of the feasible design space, the closest
contour to the optimal solution that coincides with the constraint boundary can be the optimal solution
within the constraints specified.
X2 X2

Infeasible Design Space f = C4

Infeasible Design Point
f = C3
f = C2 Optimal
Design
Functional Points
Constraint f = C1

f = C1 Contour of
Objective f = C5
f = C2 Function
f = C3
f = C5
f = C4 X2
Feasible Design f = C5
Space S1
X1

f = C5

Feasible Design
Point f = C1 f = C1
f = C2 f = C2
f = C3 f = C3
Side Constraint
f = C4 f = C4
X1 X1
03-GA50385-04a 03-GA50385-04b

Figure 1-4 Design space with designed regions. Figure 1-5. Potential optimal solutions.

5
 In the electrolyzer model (Oh et al. 2006a), the oxygen stream produced at the anode is assumed to
mix with a sweep gas stream that is introduced at the anode. The combined stream then exits the
electrolyzer. The hydrogen stream produced at the cathode is assumed to mix with a feed stream that is
introduced at the cathode. The feed stream is composed of water vapor to be electrolyzed, hydrogen gas
for maintaining reducing environment, and possibly an inert gas, presently assumed to be nitrogen.
Figure 1-6 shows the LMTD vs. System overall efficiency for system Configuration 1 (see
Figure 1-1). As shown in this figure, the large LMTD decreases the efficiency. The decrease rate is
smooth before point 4 (LMTD = 66.5), but the efficiency is sharply decreased to lower than 40% after
this point. So, the operating condition was determined at Point 3 in this report, where both high efficiency
and high LMTD can be obtained.
Figure 1-7 shows the relationship of the IHX pressure drop and system efficiency. The square symbol
is the primary side, and the circle the secondary PCU side. This figure shows that the efficiency decreases
with pressure drop linearly. The decrease rates are 0.0167%/kPa in the Primary side, and 0.0092%/kPa in
the secondary side. The lower decrease rate in the PCU side is because of the lower mass flow rate than
the primary side. Lower mass flow rate generally requires less pumping power. In this report, the pressure
drops in both sides were constrained to 70 kPa for thermal design. In this constraint, system efficiency
higher than 44% can always be achieved.

46
1
45 2 3 Primary Side
45
4 Reference Point PCU
System Overall Efficiency (%)

44
40 5
System Efficiency (%)

43

35 42

41
30
40

6
25 39

38
20 40 60 80 100 120
0 100 200 300 400 500 600
LMTD (K)
Pressure Drop (kPa)

Figure 1-6. LMTD of IHX vs. system overall Figure 1-7. Pressure drop (kPa) vs. system
efficiency for the reference reactor efficiency (%) in the reference reactors
(Configuration 1). (Configuration 1).

From the above estimations, the IHX design conditions were obtained for both Configuration 1 and
Configuration 2. They are summarized in Table 1-1 (Configuration 1) and Table 1-2 (Configuration 2). In
addition to the IHX design, this work considered SHX design, because SHXs usually experience much
more serious operating conditions. The SHX usually requires a 900¶C operating temperature and
2–5 MPa in the maximum pressure differences between the hot and cold side channels.

6
Table 1-1. Heat exchanger design conditions for Configuration 1 (Figure 1-1).
Conditions IHX SHX
Duty [MWt] 611 53
LMTD [°C] 45.37 50
Tube Side Coolant He He
Shell Side Coolant He He
Tube Inlet Temperature [°C] 900 885
Tube Outlet Temperature [°C] 594.5 480.9
Tube Side Pressure [MPa] 7 7.535
Mass Flow in Tube Side [kg/s] 385.3 25.3
Shell Inlet Temperature [°C] 492.5 337.2
Shell Outlet Temperature [°C] 884.8 875.1
Shell Side Pressure [MPa] 7.584 1.95
Mass Flow in Shell Side [kg/s] 300 19

Table 1-2. Heat exchanger design conditions for Configuration 2 (Figure 1-2).
Conditions IHX SHX
Duty [MWt] 565 55
LMTD [°C] 34 50
Tube Side Coolant He He
Shell Side Coolant He He
Tube Inlet Temperature [°C] 877.9 900
Tube Outlet Temperature [°C] 649.6 877.9
Tube Side Pressure [MPa] 6.97 7
Mass Flow in Tube Side [kg/s] 476.2 476.2
Shell Inlet Temperature [°C] 610.4 321
Shell Outlet Temperature [°C] 850 875.1
Shell Side Pressure [MPa] 7.323 1.92
Mass Flow in Shell Side [kg/s] 454 19

1.2.3 IHX and SHX Thermal Design

This section describes thermal designs of the IHX and SHX, which are two major heat exchangers in
the VHTR/HTSE system. The IHX transfers heat from primary side to gas turbine system for electric
generation. Generally, almost 90–100% of the heat (550–600 MWt) generated in the reactor core is
exchanged by this heat exchanger. On the other hand, the SHX transfers heat from the primary side or
PCU side to the hydrogen production system, depending on the system configurations. About 8–10% of
heat is transferred here. Although SHX transfers only a small portion of heat, it suffers the most serious
operating conditions in the system. In this section, we carried out parametric studies of the preliminary
thermal design of those two heat exchangers.

7
 The following factors were taken into consideration in our thermal design:
x System configuration
- Configuration 1ˁIndirect Parallel System (Figure 1-1)
- Configuration 2ˁIndirect Serial System (Figure 1-2)
x Heat Exchanger type
- PCHE
- Counter Current Flow
- Cross Flow
- Shell-and-Tube
- Straight Pipe Design
- U-Tube Design
- Helical Coil
x Heat exchanger arrangement
- Single Stage
- Two Stage.

We used separate design methods and guidelines for three different heat exchanger types. For thermal
design of PCHEs, we followed the general compact heat exchanger design procedure with the equations
and parameters summarized by Hesselgreaves (2001). In the thermal design, the printed circuit heat
exchangers are generally subjected to few constraints compared to the shell-and-tube heat exchanger.
Fluids may be liquid, gas, or two-phase. Multistream and multipass configurations can be assembled and
flow arrangements can be truly counter-current, co-current, cross-flow, or a combination of these, at any
required pressure drop. The guidelines and constraint used for design of this heat exchanger are well
summarized in Appendix A.
The shell-and-tube heat exchanger is the most common heat exchanger type. So the design methods
are very well established. In designing this heat exchanger, we followed the general design guidelines
provided by the heat transfer and heat exchanger handbooks (Kakac 2002, Hewitt 2002, Kern 1991).
Basically, we followed Kern’s Method (1991). The design guidelines and criteria for this heat exchanger
are summarized in Appendix A.
The helical coil heat exchanger is not a common heat exchanger type. So, there is no good established
design method or guideline to follow yet. However, this heat exchanger is a variation of the shell-and-
tube type. So, we developed our design method for this heat exchanger based on the shell-and-tube heat
exchanger design method. The heat transfer and friction loss correlations for helical coils were obtained
from the general heat transfer reference books (Kakac 2001, Bejan 2003). The design method of helical
coil heat exchangers is described in detail in Appendix B.
Before we carried out the thermal design, some basic design parameters were predetermined in
advance. The parameters include channel diameters and pitches. In our PCHE design, the channel
diameter was 1.2 mm and the channel horizontal pitch was 1.46 mm. The channel shape was semicircular.
In the shell-and-tube design, the tube inner diameter was 18 mm and the outer diameter was 20 mm. The
tube horizontal pitch was 30 mm and the pitch arrangement was triangular. In the helical coil design, the
diameters, thickness and pitches of the tubes were determined to be the same as the shell-and-tube design.
However, in the practical applications, this exchanger type usually requires larger tube size than general
shell-and-tube heat exchangers because of manufacturing difficulties. Therefore, the calculated design
variables could be different from the real designs. In the helical coil design, the tube arrangement was
parallel in line.

8
1.2.3.1 Single-Stage IHX Thermal Design
This subsection summarizes the estimated IHX design for the single-stage heat exchanger
arrangement. In this layout, the IHX consists of one heat exchanger unit or separate modules aligned in
parallel. There are no serial combinations of IHX. Three different heat exchanger types were designed:
PCHE, shell-and-tube, and helical coil.
Table 1-3 summarizes the design specifications of PCHEs. Two simple flow configurations—counter
current flow and cross flow—were considered for the PCHE design. According to this result, the total
core volume of the heat exchanger ranges between 4.78 and 6.028 m3, respectively, and the heat transfer
area ranges between 4,645 and 6,466 m2, respectively, for given pressure drop requirements (dP <70 kPa).
These values are based on preliminary calculations that may change in the final design after mechanical
or thermal stress is considered. Thermal stress results can limit the channel length because there are large
temperature variations in the flow direction.
Table 1-3. Single-stage IHX thermal design (PCHE).
Configuration 1 Configuration 2
Counter Counter
 Current Cross Flow Current Cross Flow
Duty [MWt] 612 612 565 565
LMTD [°C] 45.57 45.54 33.3 33.3
U [W/m2K] 2,313 2,564 2,565 2,625
A [m2] 5,805 5,241 4,645 6,466
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
Surface Area Density [m2] 2,195 2,195 2,195 2,195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.55 0.6 0.42
Stack Length [m] 0.43 0.38 0.41 0.41
Stack Height [m] 0.6 0.6 0.6 0.6
# of Stacks 34 39 41 57.02
Total Core Volume [m3] 5.29 4.78 6.028 5.891
HTC—Primary [W/m2K] 5,514 5,428 5,562 5,657
HTC—Secondary [W/m2K] 4,116 5,058 4,949 5,103
Pressure Drop (Primary) [kPa] 66.83 56.34 67.75 67.59
Pressure Drop (Secondary) [kPa] 31.76 70.02 51.57 57.53

9
 Table 1-4 summarizes the design specifications of shell-and-tube-type heat exchangers. The design
constraints and guidelines for this design are described in Appendix A. Two different tube configurations
were considered for this design: straight pipe and U-tube. The total estimated core volumes of the heat
exchangers range between 379 and 483 m3, and the heat transfer areas range between 28,300 and
42,300 m2. The U-tube design is smaller and much more resistant to the thermal expansion than the
straight pipe design, making it the generally preferred design for the high temperature application.
Compared to the PCHE design, the volume is about 60 times larger and the heat transfer area is about
seven times larger than the shell-and-tube design. The difference is PCHE’s large surface area density and
large heat transfer coefficient caused by a small channel size.
Table 1-4 Single stage IHX thermal design (shell-and-tube).
Configuration 1 Configuration 2
Conditions Straight Pipe U Tube Straight Pipe U Tube
Duty [MWt] 612 612 565 565
LMTD [°C] 45.57 45.54 33.3 33.3
2
U [W/m K] 380.5 474 405.7 496.7
2
A [m ] 3.53E+04 2.83E+04 4.13E+04 3.42E+04
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 19.56 15.42 19.6 14.69
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 2.87E+04 2.924 2.36E+04 3.70E+04
Shell Diameter [m] 5,550 5,600 6,000 6,300
Baffle Spacing [m] 4,000 4,000 5,200 5,000
Aspect Ratio 0.28 0.3631 0.3061 0.4287
3
Total Core Volume [m ] 473 379.9 554.2 458.1
Pressure Drop (Tube Side) [kPa] 13 73.11 14.6 69.36
Pressure Drop (Shell Side) [kPa] 69.83 57.15 69.88 59.56

Table 1-5 summarizes the design specifications of the helical coil heat exchanger. A detailed design
method is described in Appendix B. Diameters, thickness, and pitches of the tubes were determined to be
the same as the standard shell-and-tube heat exchangers. The total estimated core volumes of the heat
exchangers range between 163 and 192 m3, and the heat transfer area ranges between 13,200 and
13,500 m2, which is about one-half the size of the standard shell-and-tube design. The size reduction is
due to the heat transfer enhancement in the helical coil. However, its size is still much larger than the
PCHE type. In addition, helical coil heat exchangers usually require larger sized tubes in practice because
of the manufacturing problems. Therefore, the actual size of the helical coil heat exchanger would be a
little larger than the dimensions given in Table 1-5.

10
Table 1-5. Single stage IHX thermal design (helical coil).
Conditions Configuration 1 Configuration 2
Duty [MWt] 612 565
LMTD [°C] 45.58 33.31
2
U [W/m K] 1,189 1,284
2
A [m ] 1.354e4 1.32e4
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 5,025 5,899
Number of Coiled Columns 3.33 3.8
Shell Inner Diameter [m] 0.49 0.57
Shell Outer Diameter [m] 4.6 5.4
Shell Length [m] 9.86 8.37
3
Volume [m ] 163.8 191.6
Aspect Ratio (Ds/L) 0.47 0.65
dP (tube) [kPa] 66.57 70.27
dP (shell) [kPa] 62.51 72.45

An extensive comparison of these three heat changer types was conducted to determine their
advantages and disadvantages with VHTR IHX applications. The results are summarized below.
Using the PCHE type heat exchanger as an IHX would have the following advantages in the thermal
design:
x Smaller Size. A heat exchanger’s size can be represented by its total volume. A smaller volume is
generally preferred in high temperature applications because high-alloy metals or ceramics are very
costly and a smaller heat exchanger reduces the total system size. As shown in Tables 1-3, 1- 4, and
1-5 the PCHE with cross flow is the smallest, the helical tube next, and shell-and-tube the largest.
x Smaller Heat Transfer Area. Tritium permeation into a hydrogen plant is a serious problem that can
occur when the hydrogen production system is integrated with a VHTR. Since tritium permeation is
proportional to the heat transfer surface area, a smaller heat transfer area is preferred. The PCHE has
the smallest heat transfer area, the helical coil type is next, and shell-and-tube type has the largest heat
transfer area.
Compared with PCHEs, the tubular heat exchanger design (shell-and-tube and helical coil) has the
following advantages as an IHX:
x Greater Simplicity. The number of modules is related to the system complexity, fewer modules are
preferred to simplify the design. The PCHE design requires rather complex manifold and flow
connections between modules, which is not the case in the tubular design. A shell-and-tube design
would make the system simpler.
x Lower Thermal Stress. In high temperature applications, thermal stress is an important issue. The
compactness of the PCHE provides excellent heat performance, but the fast temperature variations in
the flow direction seems to create extensive thermal stress.

11
x System Readiness. Tubular type heat exchangers are a well proven technology in the industry. Helical
coil heat exchangers have an especially good operating history such as its application in the Japanese
High Temperature Test Reactor (HTTR). Technically, the PCHE type appears to have great potential
for use high temperature applications, but they are not proven for large-scale, long-lifetime
applications.
x Easier Maintenance. Tubular type heat exchangers are a proven technology with well known
problems and limitations throughout the industry. Generally, tubular heat exchangers allow for in-
service inspection and have well-established maintenance methods. Conversely, finding defects and
failures in PCHE operations appears difficult and it will not be easy to perform maintenance
compared to the tubular design.
In summary, the PCHE has a definite size and heat transfer advantage, which would lead to enormous
cost savings and enhanced safety. But the tubular-type heat exchanger is a ready-to-use technology,
especially the helical coil heat exchanger, which has good operating records such as the HTTR in Japan.
Technically, the PCHE type is strongly recommended for the VHTR-hydrogen production application
because of its significantly reduced size, cost efficiency, and safety. However, the tubular-type heat
exchanger can be a good near-term option that avoids some long-term operation and maintenance risks
associated with the unproven PCHE. The heat exchanger design specifications shown in Tables 1-3, 1-4,
and 1-5 above were calculated based on design conditions listed in Tables 1-1 and 1-2.

1.2.3.2 Two-Stage IHX Thermal Design

Another option for the IHX layout is the two-stage arrangement where the IHX is split into two units:
High Temperature Unit and Low Temperature Unit. This option was recommended to reduce associated
risk and cost. In this concept, two heat exchangers are arranged and connected in serial. The low
temperature unit is designed to last the full plant lifetime where the maximum temperature is proposed to
be around 760°C, allowed in ASME Section III, Subsection NH materials. The high temperature unit is
designed for replacement within the plant lifetime. Alloys 617 and 230 are considered potential
candidates for this unit. Figure 1-8 shows the flow schematics and design conditions for each unit
estimated by HYSYS code (ASPEN 2005). The overall design conditions are the same as the single-stage
IHX in Tables 1-1 and 1-2, but is split into two units, which affects the LMTD required for both. For
example: the LMTD in the high temperature unit decreases to 31.69°C as seen in Configuration 1, while
the LMDT in the low temperature unit increases to 77.74°C compared to the value in the single-stage
IHX of 45.37°C.
As in the single-stage layout, each unit was designed for three different heat exchanger types: PCHE,
shell-and-tube, and helical coil. Even though Alloy 617 was the proposed construction material, the effect
of materials on the heat exchanger has not been considered in the current thermal analysis because the
type of material has only a slight effect on the heat exchanger thermal design according to Natesan et al.
(2006)’s research. The reason is that the thermal resistance in the heat exchanger wall is much smaller
than in the gas fluids. Even for the ceramic materials, Natesan’s report (2006) only showed a size
reduction of about 13%.

12
 (a) Configuration 1

(b) Configuration 2

Figure 1-8. Operating condition of two-stage IHX.

Tables 1-6 through 1-8 summarize the design specifications of high temperature unit and Tables 1-9
through 1-11 summarize the design specifications of low temperature unit. PCHE, shell-and-tube, and
helical coil types of heat exchangers were designed for both units. The pressure drop for each unit was
also determined within the total pressure drop (in the high temperature section + low temperature
section), 70 kPa to match the single-stage IHX design.

13
Table 1-6. Two-stage IHX thermal design (high temperature section, PCHE).
Configuration 1 Configuration 2
Counter Counter
Conditions Current Cross Flow Current Cross Flow
Duty [MWt] 300 300 317 317
LMTD [°C] 31.74 31.73 30.97 30.97
2
U [W/m K] 2410 2443 2615 2440
2
A [m ] 3928 3877 3911 4191
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
2
Surface Area Density [m ] 2195 2195 2195 2195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.34 0.6 0.23
Stack Length [m] 0.29 0.25 0.24 0.23
Stack Height [m] 0.6 0.6 0.6 0.6
# of Stacks 34 70 41 120
3
Total Core Volume [m ] 3.569 3.5 3.564 3.819
2
HTC—Primary [W/m K] 5505 5060 5561 5182
2
HTC—Secondary [W/m K] 4440 4908 5138 4788
Pressure Drop (Primary) [kPa] 49.2 33.67 42.36 1.82
Pressure Drop (Secondary) [kPa] 26.89 40.94 34.11 26.92

14
Table 1-7. Two-stage IHX thermal design (high temperature section, shell-and-tube).
Configuration 1 Configuration 2
Conditions Straight Pipe U Tube Straight Pipe U Tube
Duty [MWt] 300 300 316 316
LMTD [°C] 31.74 31.74 31.77 31.77
2
U [W/m K] 3.96E+02 4.90E+02 4.13E+02 5.08E+02
2
A [m ] 2.39E+04 1.93E+04 2.41E+04 2.01E+04
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 1.04E+01 10.16 1.14E+01 8.65E+01
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 2.87E+04 3.03E+04 3.357E+04 3.701E+04
Shell Diameter [m] 5550 5700 6000 6300
Baffle Spacing [m] 4000 4000 5200 5000
Aspect Ratio 0.72 0.56 0.52 0.72
3
Total Core Volume [m ] 320 259 323 269.8
Pressure Drop (Tube Side) [kPa] 10 53.88 9.54 48.65
Pressure Drop (Shell Side) [kPa] 51.16 41.07 46.89 41.30

Table 1-8. Two-stage IHX thermal design (high temperature section, helical coil).
Configuration 1 Configuration 2
Duty [MWt] 300 317
LMTD [°C] 31.74 30.97
2
U [W/m K] 1249 1314
2
A [m ] 7580 7780
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 5025 5899
Number of Coiled Columns 3.007 2.239
Shell Inner Diameter [m] 0.4883 0.57
Shell Outer Diameter [m] 4.6 5.4
Shell Length [m] 6.614 4.9
3
Volume [m ] 109.9 112.8
Aspect Ratio (Ds/L) 0.7 1.1
dP (tube) [kPa] 44.67 41.37
dP (shell) [kPa] 52.87 47.73

15
Table 1-9. Two-stage IHX thermal design (low temperature section, PCHE).
Configuration 1 Configuration 2
 Counter Counter
Current Cross Flow Current Cross Flow
Duty [MWt] 312 312 249 249
LMTD [°C] 77.83 77.82 36.69 36.69
2
U [W/m K] 2259 2735 2518 2603
2
A [m ] 1772 1464 2691 2604
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
2
Surface Area Density [m ] 2195 2195 2195 2195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.16 0.6 0.17
Stack Length [m] 0.13 0.14 0.16 0.17
Stack Height [m] 0.6 0.6 0.6 0.6
# of Stacks 34 100 41 136.8
3
Total Core Volume [m ] 1.606 1.334 2.452 2.373
2
HTC—Primary [W/m K] 5259 8125 5348 5531
2
HTC—Secondary [W/m K] 4110 5146 4948 5117
Pressure Drop (Primary) [kPa] 19.02 30.45 26.11 27.38
Pressure Drop (Secondary) [kPa] 9.6 21.74 20.97 23.48

16
Table 1-10. Two-stage IHX thermal design (low temperature section, shell and tube).
Configuration 1 Configuration 2
Straight Straight

Pipe U Tube Pipe U Tube
Duty [MWt] 312 312 249 249
LMTD [°C] 77.83 77.83 37.09 36.69
2
U [W/m K] 3.74E+02 4.61E+02 4.00E+02 4.91E+02
2
A [m ] 1.07E+04 8.69E+03 1.677E+04 1.38E+04
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 5.97E+00 4.564 7.95E+00 5.933
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 2.87E+04 3.03E+04 3.357E+04 3.701E+04
Shell Diameter [m] 5550 5700 6000 6300
Baffle Spacing [m] 4000 4000 5200 5000
Aspect Ratio 0.9394 1.249 0.75 0.79
Total Core Volume [m3] 143.4 116.5 224 185.0
Pressure Drop (Tube Side) [kPa] 4.793 26.95 6.463 33.61
Pressure Drop (Shell Side) [kPa] 29.42 24.84 37.06 33.07

Table 1-11. Two-stage IHX thermal design (low temperature section, helical coil).
Configuration 1 Configuration 2
Duty [MWt] 312 249
LMTD [°C] 77.83 36.69
U [W/m2K] 1176 1271
2
A [m ] 3405 5332
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 5025 5899
Number of Coiled Columns 1.35 1.54
Shell Inner Diameter [m] 0.49 0.57
Shell Outer Diameter [m] 4.6 5.4
Shell Length [m] 2.97 3.376
3
Volume [m ] 49.38 77.32
Aspect Ratio (Ds/L) 1.5 1.6
dP (tube) [kPa] 17.51 25.21
dP (shell) [kPa] 18.84 29.23

17
 In the two-stage IHX layout, each unit can have three different heat exchanger types independently in
series such that nine serial combinations are available as shown in Table 1-12.
Table 1-12. Summary of heat exchanger types in the two-stage IHX in series.
Case HT Unit LT Unit
C1 PCHE PCHE
C2 PCHE Shell and Tube
C3 PCHE Helical Coil
C4 Shell and Tube PCHE
C5 Shell and Tube Shell and Tube
C6 Shell and Tube Helical Coil
C7 Helical Coil PCHE
C8 Helical Coil Shell and Tube
C9 Helical Coil Helical Coil

Table 1-13 summarizes the heat exchanger core volumes for each option in Table 1-12. The volumes
specified in this table were taken from Tables 1-6 though 1-11. Table 1-13 apparently shows that C1
(PCHE-PCHE) has the smallest total volume thanks to the highest compactness (surface area density =
2,195 m2). The smaller size of high temperature unit is highly recommended for cost saving and safety
enhancement. Because of the material problems from severe high temperature conditions, careful
considerations are necessary for selection and design of the high-temperature-related components. If the
high temperature unit becomes smaller, the less effort and cost will be required for manufacturing and
maintenance. In addition, the size reduction will enhance the component safety by reducing the
probability of an incident.
Table 1-13. Volume of heat exchangers for two-stage IHX.
Conditions Configuration 1 Configuration 2
Table 1-12 HT Unit LT Unit Total HT Unit LT Unit Total
C1 3.5 1.334 4.834 3.6 2.37 5.97
C2 3.5 116.5 120 3.6 185 188.6
C3 3.5 49.38 52.88 3.6 77.32 80.92
C4 259 1.334 260.334 269 2.37 271.37
C5 259 116.5 375.5 269 185 454
C6 259 49.38 308.38 269 77.32 346.32
C7 109.9 1.334 111.23 112.8 2.37 115.17
C8 109.9 116.5 225 112.8 185 297.8
C9 109.9 49.38 159.28 112.8 77.32 189.32

Table 1-14 shows the estimated heat transfer area for the heat exchangers in the serial configurations.
As predicted, C1 (PCHE-PCHE) shows the smallest heat transfer area because of the decreased thermal
boundary layer with reduced channel diameters. The heat transfer surface areas of other combinations are
at least three times larger than that of C1.

18
Table 1-14. Heat transfer area of two-stage IHX.
Conditions Configuration 1 Configuration 2
Table 1-12 HT Unit LT Unit Total HT Unit LT Unit Total
C1 3877 1464 5341 3911 2604 6515
C2 3877 8690 12567 3911 13800 17711
C3 3877 3405 7282 3911 5332 9241
C4 19300 1464 20764 20100 2604 22704
C5 19300 8690 27990 20100 13800 33900
C6 19300 3405 22705 20100 5332 25432
C7 7580 1464 9044 7780 2604 10384
C8 7580 8690 16270 7780 13800 21580
C9 7580 3405 10985 7780 5332 13112

The total duty in the two-stage IHX design was split in half (300 MWt/300 MWt). However, because
of the reduced LMTD in the high temperature unit, it requires more than twice of the volume and surface
area of the low temperature unit. The total average LMTD of the Configuration 1 system is 45–50°C, but
it reduces to about 30°C in the high temperature unit as it is increased up to 70¶C in the low temperature
unit. Since the decrease of the LMTD requires larger heat transfer surface area for the same duty, most
parts of the IHX should be operated in the high temperature region.
To reduce the size of the high temperature unit, the separation temperature (Tsep) shown in Figure 1-8
has been adjusted to 800¶C. The increase of the separation temperature leads to the decrease of duty in
the high temperature unit. It also leads to a decrease in heat exchanger size. However, the increase of the
separation temperature makes the low temperature section more vulnerable to the stress. The details about
the stress analysis are described in Tables 1-15, 1-16, 1-17, and 1-18.

19
Table 1-15. Two-stage IHX thermal design (high temperature section, PCHE, Tsep = 800oC).
Configuration 1 Configuration 2
Counter Counter
Conditions Current Cross Flow Current Cross Flow
Duty [MWt] 200 200 193 193
LMTD [°C] 26.81 27.36 29.75 29.75
2
U [W/m K] 2438 2617 2638 2844
2
A [m ] 3064 2792 2458 2279
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
2
Surface Area Density [m ] 2195 2195 2195 2195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.27 0.6 0.16
Stack Length [m] 0.23 0.2 0.15 0.16
Stack Height [m] 0.6 0.6 0.6 0.6
# of Stacks 34 79 41 135.2
3
Total Core Volume [m ] 2.784 2.549 2.239 2.077
2
HTC—Primary [W/m K] 5505 5384 5561 6007
2
HTC—Secondary [W/m K] 4536 5311 5228 5647
Pressure Drop (Primary) [kPa] 39.35 32.59 27.29 33.65
Pressure Drop (Secondary) [kPa] 22.37 40.4 22.57 29.24

20
Table 1-16. Two-stage IHX thermal design (high temperature section, shell-and-tube, Tsep = 800°C).
Configuration 1 Configuration 2
Conditions Straight Pipe U Tube Straight Pipe U Tube
Duty [MWt] 200 200 193 193
LMTD [°C] 26.82 26.82 29.75 29.75
2
U [W/m K] 4.00E+02 4.96E+02 4.17E+02 5.13E+02
2
A [m ] 1.87E+04 1.50E+04 1.57E+04 1.26E+04
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 1.04E+01 7.911 7.376E+01 5.43
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 2.87E+04 3.03E+04 3.357E+04 3.70E+04
Shell Diameter [m] 5550 5700 6000 6300
Baffle Spacing [m] 4000 4000 5200 5000
Aspect Ratio 0.53 0.72 0.813 1.16
3
Total Core Volume [m ] 250 201.9 208.6 169.3
Pressure Drop (Tube Side) [kPa] 13 45.98 7.004 36.35
Pressure Drop (Shell Side) [kPa] 69.83 34.55 35.44 31.54

Table 1-17. Two-stage IHX thermal design (high temperature section, helical coil, Tsep = 800°C).
Conditions Configuration 1 Configuration 2
Duty [MWt] 200 193
LMTD [°C] 26.81 29.75
2
U [W/m K] 1266 1329
2
A [m ] 5901 4880
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 5025 5899
Number of Coiled Columns 2.34 1.4
Shell Inner Diameter [m] 0.488 0.57
Shell Outer Diameter [m] 4.6 5.4
Shell Length [m] 5.15 3.09
3
Volume [m ] 85.58 70.76
Aspect Ratio (Ds/L) 0.89 1.743
dP (tube) [kPa] 34.78 25.95
dP (shell) [kPa] 43.89 31.53

21
Table 1-18. Two-stage IHX Thermal design (low temperature section, PCHE, Tsep = 800°C).
Configuration 1 Configuration 2
Counter Counter
Conditions Current Cross Flow Current Cross Flow
Duty [MWt] 400 400 3.72E+05 3.72E+05
LMTD [°C] 66.72 66.71 35.4 35.4
2
U [W/m K] 2281 2704 2537 2452
2
A [m ] 2627 2216 4147 4292
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
2
Surface Area Density [m ] 2195 2195 2195 2195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.24 0.6 0.25
Stack Length [m] 0.2 0.2 0.26 0.25
Stack Height [m] 0.6 0.6 0.6 0.6
# of Stacks 34 70 41 104.3
3
Total Core Volume [m ] 2.387 2.019 3.779 3.91
2
HTC—Primary [W/m K] 5344 6009 5433 5248
2
HTC—Secondary [W/m K] 4110 5117 4948 4778
Pressure Drop (Primary) [kPa] 28.89 40.49 41.15 34.86
Pressure Drop (Secondary) [kPa] 14.27 32.13 32.32 28.77

22
Table 1-19. Two-stage IHX thermal design (low temperature section, shell-and-tube, Tsep = 800°C).
Configuration 1 Configuration 2
Conditions Straight Pipe U Tube Straight Pipe U Tube
Duty [MWt] 400 400 372 372
LMTD [°C] 68.72 68.72 35.80 35.80
2
U [W/m K] 3.76E+02 4.63E+02 4.02E+02 4.93E+02
2
A [m ] 1.59E+04 1.30E+04 2.59E+04 2.13E+04
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 8.82E+00 6.803 12.27 9.17
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 2.87E+04 3.03E+04 2.87E+04 3.701E+04
Shell Diameter [m] 5550 5700 6000 6300
Baffle Spacing [m] 4000 4000 5200 5000
Aspect Ratio 0.6293 0.8379 0.4891 0.69
3
Total Core Volume [m ] 213.4 173.6 346.8 285.9
Pressure Drop (Tube Side) [kPa] 6.513 35.45 9.244 46.50
Pressure Drop (Shell Side) [kPa] 38 31.33 49.22 42.86

Table 1-20. Two-stage IHX thermal design (low temperature section, helical coil, Tsep = 800°C).
Conditions Configuration 1 Configuration 2
Duty [MWt] 415 372
LMTD [°C] 68.72 35.40
2
U [W/m K] 1180 1276
2
A [m ] 5077 8244
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 5025 5899
Number of Coiled Columns 2.0 2.4
Shell Inner Diameter [m] 0.488 0.57
Shell Outer Diameter [m] 4.6 5.4
Shell Length [m] 4.43 5.22
3
Volume [m ] 73.63 119
Aspect Ratio (Ds/L) 1.038 1.034
dP (tube) [kPa] 27.37 40.88
dP (shell) [kPa] 28.09 45.20

23
 Tables 1-21 and 1-22 summarize the volumes and heat transfer surface areas for two-stage IHXs at
the increased separation temperature (800°C). The volumes and surface areas were estimated by the same
method used for the design at the original separation temperature (750°C). When we increase the
separation temperature to 800°C, the size of the high temperature unit is reduced to about a half of the
total volume as shown in Table 1-21. Since the high temperature unit is a replaceable during the plant
lifetime, the smaller PCHEs will be more beneficial, leading to lower cost and easier maintenance. In
addition, the smaller surface area of the PCHE in high temperature operations will greatly reduce the
tritium penetration rate from the primary to secondary side. On the other hands, the low temperature unit
is a nonreplicable component and unit reliability over the plant’s lifetime is important. Also, tritium
penetration in the low temperature unit is much lower than in the high temperature unit, since the
diffusion coefficient is exponential to the material temperature.

Table 1-21. Volume of heat exchangers for Two-stage IHX (800¶C separation).
Configuration 1 Configuration 2
HT Unit LT Unit Total HT Unit LT Unit Total
C1 2.549 2.019 4.568 2.08 3.78 5.86
C2 2.549 173.6 176.15 2.08 285.9 288.0
C3 2.549 73.63 76.179 2.08 119 121.08
C4 201.9 2.019 203.92 169 3.78 172.8
C5 201.9 173.6 375.5 169 285.9 454.9
C6 201.9 73.63 275.53 169 119 288
C7 85.58 2.019 87.599 70.76 3.78 74.54
C8 85.58 173.6 259.18 70.76 285.9 356.7
C9 85.58 73.63 159.21 70.76 119 189.76

Table 1-22. Heat transfer area of two-stage IHX (800¶C separation).

Configuration 1 Configuration 2
HT Unit LT Unit Total HT Unit LT Unit Total
C1 2792 2216 5008 2279 4147 6426
C2 2792 13000 15792 2279 21300 23579
C3 2792 5077 7869 2279 8244 10523
C4 19300 2216 17216 12600 4147 16747
C5 19300 13000 28000 12600 21300 33900
C6 19300 5077 24377 12600 8244 20844
C7 5901 2216 8117 4880 4147 9027
C8 5901 13000 18901 4880 21300 26180
C9 5901 5077 10978 4880 8244 13124

24
1.2.3.3 Comparisons of Two-Stage IHX vs. Single-Stage IHX
The results of a comparison of the single-stage IHX and two-stage IHX, as presented in Tables 1-3
through 1-22, is summarized as follows:
x Total IHX volume. In high temperature applications, the smaller IHX is generally preferred. However,
the total volume of the IHX is not affected by IHX layout, meaning that splitting the IHX does not
lead to any volume increases or decreases. Practically, a two-stage IHX will require more space for
manifold of flow distribution, but this is also negligible.
x Total Heat Transfer Area. The total heat transfer area is another important parameter in VHTR
applications. In the hydrogen production system integrated with the VHTRs, tritium permeation into a
hydrogen plant will be affected by the surface area. A smaller heat transfer area can reduce the tritium
permeation rate, but the total heat transfer area of the IHX is not affected by its layout. It means that
splitting the IHX does not lead to any size increases or decrease of the IHX surface area.
x Size of high temperature section. The smaller size of the high temperature unit is preferred because it
can reduce the maintenance cost and enhance component safety. The two-stage IHX unit with an
800°C separation temperature is the smallest. A single-stage IHX unit is the largest.
x System complexity. The number of modules relates to the system complexity. Fewer modules are
preferred for simplicity. The single-stage IHX has the simplest design because it requires less
modules.
According to the above comparisons, the two-stage PCHEs are conceptually expected to offer more
advantages in system cost, safety, and maintenance compared to the single-stage IHXs. For example:
material cost can be saved in the two-stage IHXs because some cheap commercial materials like stainless
steel or Alloy 800H can replace high-alloy materials in the low temperature units. It is also easier to
manage and control for risks by concentrating safety issues on the high temperature units. The thermal
stress can be relieved in the two-stage IHXs by reducing temperature differences in the two components.
On the other hand, a single stage PCHE provides a quite small size and simplicity.

1.2.3.4 SHX Thermal Design

A SHX transfers heat from the primary or secondary side to the hydrogen production system. In
Configuration 1 of Figure 1-1, the SHX is placed in the bypass stream in the PCU cycle. In
Configuration 2 of Figure 1-2, the SHX is placed in the primary system between the reactor core and an
IHX. The two-stage concept was not considered for SHX design in this study. First, the SHX is much
smaller than the IHX because it exchanges only 1/11 of the total power. Therefore, the benefit of
replacing an SHX is very slight. On the other hand, this concept makes the system more complicated by
increasing the number of modules. In addition, the total volume and heat transfer area of the SHX are not
significantly affected by the IHX arrangement (e.g., one-stage, two-stage). Therefore, most of the
valuable specifications (total volume, heat transfer area, LMTD, heat transfer coefficient) can be obtained
for the SHX using a single-stage design. It is also obvious that the PCHE type is more advantageous in
SHX applications than the other types because for the same reasons as the IHX. Tables 1-24, 1-25 and
1-26 summarize the SHX thermal design specifications for PCHE, shell-and-tube, and helical coil types,
respectively.

25
Table 1-23. SHX thermal design (PCHE).
Configuration 1 Configuration 2
Counter Counter
Current Cross Flow Current Cross Flow
Duty [MWt] 53 53 54.7 54.7
LMTD [°C] 50.67 50.67 171 171
U [W/m2K] 1547 1446 1018 1188
A [m2] 678 746.9 314.2 269.3
Channel Diameter [m] 1.20E-03 1.20E-03 1.20E-03 1.20E-03
Channel Pitch [m] 1.46E-03 1.46E-03 1.46E-03 1.46E-03
Plate Thickness [m] 9.60E-04 9.60E-04 9.60E-04 9.60E-04
Surface Area Density [m2] 2195 2195 2195 2195
Ratio of free flow and frontal area 0.2148 0.2148 0.4024 0.4024
Effective Diameter [m] 7.33E-04 7.33E-04 7.33E-04 7.33E-04
Stack Width [m] 0.6 0.38 0.6 0.06
Stack Length [m] 0.6 0.5 0.053 0.05
Stack Height [m] 0.42 0.6 0.6 0.6
# of Stacks 4 4.675 15 136.3
Total Core Volume [m3] 0.6162 0.6805 0.29 0.25
HTC—Primary [W/m2K] 3863 4003 1.23E+04 1.84E+04
HTC—Secondary [W/m2K] 2635 2307 1141 1348
Pressure Drop (Primary) [kPa] 22.05 28.32 49.65 52.16
Pressure Drop (Secondary) [kPa] 3.73 22.62 0.53 0.942

26
Table 1-24. SHX thermal design (shell and tube).
Configuration 1 Configuration 2
Straight Pipe U Tube Straight Pipe U Tube
Duty [MWt] 53 53 55 55
LMTD [°C] 49.25 49.25 171.1 171.1
U [W/m2K] 226.9 342.3 238.5 200.2
A [m2] 4.76E+03 3.15E+03 1.34E+03 1.60E+03
Tube Inner Diameter [mm] 1.80E+01 1.80E+01 1.80E+01 1.80E+01
Tube Outer Diameter [mm] 2.00E+01 2.00E+01 2.00E+01 2.00E+01
Tube Pitch [mm] 3.00E+01 3.00E+01 3.00E+01 3.00E+01
Pitch-to-diameter ratio 1.5 1.5 1.5 1.5
Tube Length [m] 20.3 20.3 1.869 1.092
Tube Thickness [mm] 1.00E+00 1.00E+00 1.00E+00 1.00E+00
# of Tubes 3.74E+03 3021 1.14E+04 2.33E+04
Shell Diameter [m] 2000 1800 3500 5000
Baffle Spacing [m] 1500 1200 2000 2000
Aspect Ratio 0.099 0.1083 1.873 4.581
Total Core Volume [m3] 63.78 42.28 17.98 46.6
Pressure Drop (Tube Side) [kPa] 3.297 28.85 28.29 28.29
Pressure Drop (Shell Side) [kPa] 17.75 26.24 0.7915 0.47

Table 1-25. SHX thermal design (Helical).

Configuration 1 Configuration 2
Duty [MWt] 53 55
LMTD [°C] 50 171
U [W/m2K] 701 238
A [m2] 1519 1337
Tube Inner Diameter [mm] 18 18
Tube Outer Diameter [mm] 2 2
Tube Pitch [mm] 30 30
Pitch-to-diameter Ratio 1.5 1.5
Number of Tubes 795 3724
Number of Coiled Columns 10.95 0.66
Shell Inner Diameter [m] 0.17 0.53
Shell Outer Diameter [m] 1.6 5.0
Shell Length [m] 10.95 0.98
Volume [m3] 22.03 19.36
Aspect Ratio (Ds/L) 0.14 5.1
dP (tube) [kPa] 12.77 26.47
dP (shell) [kPa] 17.28 0.025

27
 1.3 Stress Analysis of the VHTR Heat Exchangers
1.3.1 Simplified Stress Analysis Method

1.3.1.1 Maximum Allowable Stress

This section presents the results of a simplified stress analysis performed for the VHTR heat
exchangers. The heat exchanger lifetime and required thickness is estimated so that the circumferential
stress was less than or equal to an assumed allowable value. The use of consistent stresses allowed us to
identify limiting components and make a fair comparison between different configurations.
Since the VHTR heat exchangers (IHX and SHX) are operating in the high temperature environment,
the creep deformation is important; Subsection NH of ASME Code, Section III applies. However, there
are several problems with applying ASME Section III code rules at this time because the use of the
primary candidate structural Alloy 617 is currently not approved in Subsection NH. For this reason, we
used the draft Code Case for Alloy 617 in this report. The reported allowable stress data in the draft Code
for Alloy 617 are shown in Figure 1-9. The primary stress limit of Alloy 617 draft code case is
determined as follows (Natesan et al. 2006):
A basic high temperature primary stress limit is Smt, which is the lesser of Sm and St and a function of
both time and temperature. Sm is basically defined as follows:

1 2
Sm min( S u , S y ) (1-1)
3 3
where Su is the lesser of ultimate tensile strength at temperature and the minimum ultimate tensile
strength at room temperature and Sy is the lesser of yield strength at temperature and the minimum yield
strength at room temperature. For each specific time t and temperature T, St is defined as the least of the
following three stresses:
x 100% of the average stress required to obtain a total strain of 1%
x 80% of the minimum stress to cause initiation of tertiary creep
x 67% of the minimum stress to cause ruptures.
In the draft code case for Alloy 617, bullet 2 above is dropped because nickel alloy does not exhibit
classical creep behavior. Figure 1-9 plots the reported
Sm and St values for Alloy 617. For this report, the
allowable stress is assumed to be the minimum value
of Sm and St.

1.3.1.2 Calculation of Stress and Failure

Theory
The simple stress and failure theories used to
determine the internal pressure required to yield the
inner surface of the tubular-type heat exchangers are
presented below. The method used for the PCHE type
heat exchangers is first. Generally, the design of the
heat exchanger channels is defined by the channel
diameter (d), pitch (p), and plate thickness, (tp).
According to the method used by Dostal et al. (2004), Figure 1-9. Variations of Sm and St of alloy 617
the minimum wall thickness between channels, tf, can with temperature and time (Natesan et al. 2006).

28
be approximated as

p
tf t , (1-2)
VD
1
'P
where V D is the allowable stress, p is the channel horizontal pitch, and 'P is the differential pressure
between the hot and cold streams. Expressing Equation (1-5) in terms of 'P yields

§ p ·
'P d ¨  1¸V D . (1-3)
¨t ¸
© f ¹
The required plate thickness can be calculated based on the method of Dostal et al. (2004). The plate
is assumed to be a thick-walled cylinder, with an inner radius of d/2 and an outer radius of tp. For thick
walled cylinders, the tangential stress, V t , is calculated as (Crandall et al. 1972)

Vt
> 2
@ >
Pi r0 / ri   1  P0 r0 / ri   r0 / r 
2 2
@ (1-3)
r0 / ri 
2
1

where r is the radius, P is the pressure, and the subscript i and o refer to the inner and outer surface,
respectively. The stress is negative if the external pressure exceeds the internal pressure, but the
maximum magnitude always occurs at the inner surface. The radius ratio that causes the maximum stress
to be less than or equal to the allowable stress, V D , can be calculated from Equation (1-3). For cases
where the internal pressure exceeds the external pressure, the limiting ratio is

ro V D  Pi
t . (1-4)
ri V D  2 P0  Pi

For cases where the external pressure exceeds the internal pressure, the maximum, absolute value of
the stress will be less than or equal to the allowable stress when the radius ratio is

ro V D  Pi
t . (1-5)
ri V D  2 P0  Pi

Engineering problems concerned with the design and development of structural or machine parts are
generally involving biaxial (occasionally triaxial) stresses. However, available strength data usually
pertain to uniaxial stress, and often only to uniaxial tension. To resolve this problem, a failure theory is
used in the engineering practice. The failure theories are generally based on the assumption that tensile
yielding occurs as a result of exceeding the capacity of the materials in one or more respects, such as:
1. Capacity to withstand normal stress (Maximum Normal Stress Theory)
2. Capacity to withstand shear stress (Maximum Shear Stress Theory)
3. Capacity to withstand normal strain (Maximum Normal Strain Theory)
4. Capacity to withstand shear strain (Maximum Shear Strain Theory)
5. Capacity to absorb strain energy (Total Strain Energy Theory)
6. Capacity to absorb distortion energy (Maximum Distortion Energy Theory).

29
 Hence, in the simple classical theories of failure, it is assumed that the same amount of whatever
caused the selected tensile specimen to fail will also cause any part made of the materials to fail
regardless of the state of stress involved. The model details are well described in Collins (1981). The
failure theories are compared graphically for a biaxial state of stress in Figures 1-10 and 1-11 (Wolf et al.
2004).

Figure 1-10. Comparisons of failure theories for a biaxial state of stress (Martin
1962).

Figure 1-11. Comparisons of biaxial strength data with failure theories (Collins 1981).

From Figure 1-10 it can be seen that:

1. The distortion energy and maximum shear stress theories predict similar results with the shear stress
theory being more conservative.

30
2. The maximum normal stress and maximum shear stress theories agree in the first and third quadrants
where the signs of the principal stresses are the same but not in the second and fourth quadrants.
Biaxial strength data for a variety of ductile and brittle materials are shown in Figure 1-11 with
several failure theory limits. From this figure it can be seen that experimental data support:
1. Maximum normal stress theory is appropriate for brittle behavior.
2. Distortion energy or Maximum shear stress theories is appropriate for ductile failure.
In this report, the maximum shear stress theory was used for conservative failure estimations.

1.3.2 Simplified Stress Evaluation for IHX and SHX

This section describes the results of the simplified stress evaluation for the VHTR heat exchangers.
The construction material used in this estimate was assumed to be Alloy 617, which is the potential
candidate for the VHTR IHXs. Figure 1-12 shows the relations between the lifetime and the stress limit
(Smt) of this material for three different temperatures; 750, 800, and 900°C. Each data of the graph
represents the maximum operational lifetime under the given stress load and temperature. According to
this graph, the lifetime of this material exponentially decreases as the stress increases. Because of the
limited reported lifetime data over 1 × 105 hrs (about 10 years), a simple extrapolation method was used
for estimations over this range. Table 1-26 summarizes the allowable stresses of Alloy 617 assumed in
this work based on the data in Figure 1-12. To obtain more accurate analyses, available experimental data
or proven extrapolation methods will need to be measured or developed.
7
10
Alloy 617 900 C
o

6 o
10 800 C
o
750 C
5
10
Life Time (h)

4
10

3
10

2
10

1
10
0 5 10 15 20 25 30 35 40 45 50 55 60 65
Smt (MPa)

Figure 1-12. Smt (MPa) vs. life time (h) for Alloy 617.

Table 1-26. Assumed allowable stresses for Alloy 617.

Allowable Stress (MPa)
5
10 hours 5x105 hours 106 hours
Temperature (11 years) (57 years) (114 years)
750°C 33.1 25 22.5
800°C 23.2 15 12.3
900°C 10.2 5 3.5

31
 Table 1-27 summarizes the estimated lifetime of the referenced heat exchangers designed in the
previous section. In those heat exchangers, the thickness-to-inner radius ratios (t/ri) are 0.57 for the PCHE
and 0.1 for the tubular-types (See Section 1.2). The maximum pressure differences ('Pmax) between hot
and cold channels was assumed to be 1.0 MPa for the IHX, and 0.2 and 0.5 MPa for the SHX. The
maximum shear stress theory was applied to obtain the most conservative estimation of the lifetime.
Theoretically, PCHE has larger t/ri values than tubular-type heat exchangers because it shows a larger
lifetime. However, as the thickness-to-inner radius ratio in the tubular type heat exchanger is increased,
the lifetime of the tubular heat exchanger gets closer to the PCHE’s. However, the common t/ri values in
commercial tubing range from 0.1 to 0.5, and most of them are within 0.3; the designed tube thickness
should not be larger than this.

Table 1-27 shows that the PCHE can be operated in 900¶C for about 35 years as an IHX and a
SHX (Pmax = 2.0 MPa) with the reference design. However, the tubular heat exchangers (shell-and-tube
or helical coil) can only be operated for about 13 years (1.16x105 hours) as an IHX and about 0.5 year
(4,040 hours) as a SHX (Pmax = 2.0 MPa) with the reference design. With a differential pressure of
5.0 MPa between the hot and cold channels, the PCHE heat exchangers cannot be operated for more than
1.6 years (1.44x104 hours), whereas the tubular type heat exchangers cannot be operated at all for this
condition.
Table 1-27. Estimated life time of reference heat exchangers.
Life Time (hrs)
IHX (1 MPa) SHX (2.0 MPa/5.0 MPa)
Temperature PCHE Shell-and-Tube PCHE Shell-and-Tube
7
1.34 × 10 hrs/ 1.16 × 106 hrs/
750°C 1.34 × 107 hrs 7.42 × 106 hrs
2.62 × 106 hrs 5290 hrs
6
2.62 × 10 hrs/ 1.98 × 105 hrs/
800°C 2.62 × 106 hrs 1.32 × 106 hrs 5
4.08 × 10 hrs 155 hrs
5
3.12 × 10 hrs/ 4040 hrs/
900°C 3.12 × 105 hrs 1.16 × 105 hrs
1.44 × 104 hrs NONE

Figures 1-13 through 1-15 show the maximum effective stress of the PCHEs as a function of pate
thickness-to-diameter ratio (tp/d) for three different applications; (1) IHX ('Pmax = 1 MPa), (2) SHX
('Pmax = 2 MPa) and (3) SH ('Pmax = 5 MPa). The blue lines in this graph show the rupture stress for the
given lifetime (see Figure 1-12). As described, in this report, the effective stresses were calculated based
on the maximum shear stress theory, the most conservative failure theory. Therefore, for reliable
operation, the effective maximum stress should always be designed to be lower than the rupture stress.
According to these graphs, as the plate thickness-to-diameter ratio increases, the effective stress is
exponentially decreased. In addition, lower pressure differences between hot and cold channels led to
much less effective stress, meaning that the thicker the plate, the stronger the mechanical integrity gets.
Figures 1-16 through 1-18 show the maximum effective stress for the tubular heat exchangers as a
function of tube thickness-to-inner radius ratio, which is basically the same trend as the PCHEs. The
required thicknesses for the IHXs and SHXs were estimated based on these results.

32
 80
PCHE 900 oC
IHX ('P=1.0MPa)
70 SHX ('P=2.0MPa)
SHX ('P=5.0MPa)
60

Maximum Stress (MPa)

50

40 Design is not
allowed above
30 the blue line.

20
5
Allowable Stress at Life time = 10 hours (~11 yrs)
10
Design is allowed.
Allowable Stress at Life time = 5X105 hours (~57 yrs)
0
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

plate thickness-to-diameter ratio (tp/d)

Figure 1-13. Maximum stress as a function of plate thickness-to-diameter ratio

(PCHE, 900oC).

80
PCHE 800 oC
IHX ('P=1.0MPa)
70 SHX ('P=2.0MPa)
SHX ('P=5.0MPa)
60
Maximum Stress (MPa)

Design is not
50
allowed above
the blue line.
40

30
Allowable Stress at Life time = 105 hours (~11 yrs)

20
Allowable Stress at Life time = 5X105 hours (~57 yrs)

10 Design is allowed.
0
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

plate thickness-to-diameter ratio (tp/d)

Figure 1-14. Maximum stress as a function of plate thickness-to-diameter ratio

(PCHE, 800°C).

33
 80
PCHE 750 oC
IHX ('P=1.0MPa)
70 SHX ('P=2.0MPa)
SHX ('P=5.0MPa)
Maximum Stress (MPa) 60 Design is not
allowed above
50 the blue line.

40 5
Allowable Stress at Life time = 10 hours (~11 yrs)

30 5
Allowable Stress at Life time = 5X10 hours (~57 yrs)

20
Design is allowed.
10

0
0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

plate thickness-to-diameter ratio (tp/d)

Figure 1-15. Maximum stress as a function of plate thickness-to-diameter ratio

(PCHE, 750°C).

60
Shell-and-Tube 900oC IHX ('P=1.0MPa)
SHX ('P=2.0MPa)
50 SHX ('P=5.0MPa)
Maximum Stress (MPa)

40 Design is not
allowed above
30
the blue line.

20

Allowable Stress at Life time = 1X105 hours (~11 yrs)

10
Design is allowed.
Allowable Stress at Life time = 5X105 hours (~57 yrs)
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

tube thickness-to-inner ratius ratio (t/ri)

Figure 1-16. Maximum stress as a function of tube thickness-to-radius ratio
(shell-and-tube, 900°C).

34
 60
o
Shell-and-Tube 800 C IHX ('P=1.0MPa)
SHX ('P=2.0MPa)
50 SHX ('P=5.0MPa)
Design is not
Maximum Stress (MPa) allowed above
40
the blue line.

30
Allowable Stress at Life time = 1X105 hours (~11 yrs)

20
Allowable Stress at Life time = 5X105 hours (~57 yrs)

Design is allowed.
10

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

tube thickness-to-inner ratius ratio (t/ri)

Figure 1-17. Maximum stress as a function of tube thickness-to-radius ratio
(shell-and-tube, 800°C).

60
Shell-and-Tube 750oC IHX ('P=1.0MPa)
SHX ('P=2.0MPa)
50 SHX ('P=5.0MPa) Design is not
allowed above
Maximum Stress (MPa)

the blue line.

40
5
Allowable Stress at Life time = 1X10 hours (~11 yrs)

30
Allowable Stress at Life time = 5X105 hours (~57 yrs)

20
Design is allowed.
10

0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

tube thickness-to-inner ratius ratio (t/ri)

Figure 1-18. Maximum stress as a function of tube thickness-to-radius ratio

(shell-and-tube, 750°C).

35
 The above estimates are summarized in Tables 1-28, 1-29, and 1-30. Table 1-28 summarizes the
required thickness estimated for an IHX, which shows that the PCHE can be operated for most of the
whole lifetime (~35 years) with the reference design (tp/d = 0.58). The tubular-type heat exchangers show
similar performance, but in a larger thickness-to-inner radius ratio (t/ri=0.22) than the reference design
(t/ri = 0.1). The lifetime t/ri for the reference tubular heat exchanger design (t/ri = 0.1) is estimated to be
1.0 × 105 hrs (about 11 years) as presented in Table 1-27. Below 800°C, PCHE and tubular-type heat
exchangers do not show any problems for the 50-year operational design.
Tables 1-29 and 1-30 present the required thickness-to-inner radius ratios for SHXs. Two maximum
pressure differences ('Pmax) were considered between the hot and cold channels: 2.0 MPa and 5.0 MPa.
When the case result is 'Pmax = 2.0 MPa, the estimated results were similar to the IHX case, but in the
5.0 MPa range, neither the PCHE nor tubular type heat exchanger can be operated in 900°C range for the
reasonable lifetime, since the required thicknesses are too large. However, below 800 ×, the required
thicknesses are within the reasonable design range.

Table 1-28. Required thickness for IHX (for 'Pmax = 1.0 MPa).
PCHE Shell-and-Tube
(Plate Thickness/Diameter ) (Tube Thickness/Inner Radius)
Temperature 105 hours 5 × 105 hours 105 hours 5 × 105 hours
750°C 0.51 0.51 0.006 0.002
800°C 0.51 0.53 0.02 0.025
900°C 0.56 ’ 0.1 ’

Table 1-29. Required thickness for SHX (for 'Pmax = 2.0 MPa).
PCHE Shell-and-Tube
(Plate Thickness/Diameter ) (Tube Thickness/Inner Radius)
Temperature 105 hours 5 × 105 hours 105 hours 5 × 105 hours
750°C 0.51 0.52 0.08 0.05
800°C 0.51 0.59 0.2 0.1
900°C 0.64 ’ 0.23 ’

Table 1-30. Required thickness for SHX (for 'Pmax = 5.0 MPa).
PCHE Shell-and-Tube|
(Plate Thickness/Diameter ) (Tube Thickness/Inner Radius)
Temperature 105 hours 5 × 105 hours 105 hours 5 × 105 hours
750°C 0.6 0.62 0.2 0.31
800°C 0.65 0.93 0.3 0.8
900°C 1.65 ’ ’ ’

Figures 1-19 and 1-20 show the relationships between the heat exchanger wall thickness and the
estimated lifetime for PCHE and tubular-types, respectively. As shown this figures, the lifetime of the
heat exchangers sharply increases in the small thickness. However, the lifetime is suddenly saturated for a
certain thickness level. The reason is because of the normal stress on the inner wall. If the wall is getting
thicker, the tangential stresses on the tubes are sharply reduced. However, since the normal stress is not
affected by the tube thickness, the total effective stress is limited by this value. We used the maximum
shear stress theory for failure analysis, so the lifetime is finally determined by the normal stress of the
tube inside.

36
 3.5e+5

3.0e+5

2.5e+5 IHX ('P = 1.0 MPa)

SHX ('P = 2.0 MPa)
Life Time (hrs) SHX ('P = 5.0 MPa)
2.0e+5

1.5e+5

1.0e+5

5.0e+4

0.0

0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8

plate thickness-to-diameter ratio (tp/d)

Figure 1-19. Plate thickness-to-diameter ratio vs. estimated heat exchanger

lifetime (PCHE).

3.5e+5

3.0e+5

2.5e+5
Life Time (hours)

IHX ('P = 1.0 MPa)

SHX ('P = 2.0 MPa)
2.0e+5
SHX ('P = 5.0 MPa)

1.5e+5

1.0e+5

5.0e+4

0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2

tube thickness-to-inner radius ratio (t/ri)

Figure 1-20. Tube thickness-to-inner radius ratio vs. estimated heat

exchanger lifetime (shell-and-tube).

37
 1.4 Computational Analysis on the Printed Circuit Heat Exchanger
Channel Configuration Options
In the previous sections, we found from some simple thermal designs and stress analyses that the
PCHE type is conceptually very adequate for VHTR heat exchanger applications compared to other types.
In this section, some computational analyses have been performed for the PCHEs in order to investigate
the heat exchanger performance. The two cross-sectional channel configuration options considered were
standard and off-set. And the performances, including heat transfer, tritium diffusion, and maximum
stress, have been compared together for each option. Finally, this work provides and recommends a better
option for PCHE channel configuration, one that can maximize heat exchanger performance. COMSOL
Multiphysics 3.4 software was used to analyze the heat transfer and tritium diffusion. The details of this
software and the analysis results are summarized in the following Section 4.1. The stress analysis has
been performed by ABAQUS code, and the results are described in Section 4.2.

1.4.1 COMSOL Modeling for Heat Transfer and Tritium Penetration in the PCHE
COMSOL Multiphysics 3.4 was used to perform this analysis in order to compare the performance of
heat transfer and tritium penetration in the PCHEs (COMSOL 2008). The COMSOL Multiphysics
(formerly FEMLAB) software is a finite element analysis and solver software package for various physics
and engineering applications, especially coupled phenomena and multiphysics. This software offers
several application-specific modules including electromagnetic, acoustic, chemical engineering, earth
science, heat transfer, fluid dynamics, structure mechanics models, etc. It also allows the user to enter
coupled systems of partial differential equations, which can be entered directly or using weak forms. In
this work, simple heat transfer and diffusion modules were used for PCHE analysis.

1.4.1.1 Geometry and Dimensions

Figure 1-21 shows the PCHE channel configurations used: standard in-line and off-set. A PCHE
normally has a standard in-line configuration, but for this analysis, the off-set configuration was also
considered as an alternative. The referenced PCHE configuration had a plate thickness of 0.96 mm
(tp=0.96 mm) and a pitch of 1.464 mm (p=1.464 mm). The diameter of the semicircular ports was 1.2 mm
(dt=1.2 mm). Calculations were performed for five horizontal pitches (1.332 mm, 1.464 mm, 1.728 mm,
1.992 mm, and 2.5 mm) and five plate thicknesses (0.69 mm, 0.96 mm, 1.32 mm, 2.54 mm, and
3.48 mm).The horizontal pitch, plate thickness, and diameter are shown on Figure 1-21. The offset model
was made by shifting every other plate by half the model’s pitch.

38
(a) Standard in-line configuration

(b) Off-set configuration

Figure 1-21. PCHE channel configurations.

39
1.4.1.2 Properties and Boundary Conditions
The heat transfer and mass diffusion equations for tritium transport were solved in the PCHE solid
region, but they were not solved in the fluid region where simple convective heat and mass transfer
correlations were used to determine the boundary temperatures and tritium concentrations on the solid
surface. Tritium adsorptions or chemisorptions were not considered in this analysis for modeling, so the
diffusion in the heat exchanger solid is the rate controlling step. The data for Alloy 800 was therefore
implemented into the model because of the lack of Alloy 617 data for the tritium diffusion. Because the
purpose of this study is to compare the performance of different channel configurations, the results from
Alloy 800 are still valid for other metals such as Alloy 617 and 230. The following diffusion coefficient
was used for tritium permeation calculation (Richard 2006):

P 6250
Dtritium 61.02 ˜ ˜ exp( ) (1-7)
CH2 T

where,

PCi
Dtritium = diffusion coefficient of tritium in Alloy 800 ( )
m 2 hr
P = pressure (atm)
CH 2 = hydrogen concentration (ppmv)
T = absolute solid temperature (K)
As mentioned above, the boundary conditions for the heat transfer and mass transfer were determined
by convective heat/mass transfer correlations. In order to consider this, Schmidt number (Sc), kdiffusion, and
the diffusion coefficient were calculated based on the bulk conditions of the fluid assumed by the VHTR
IHX conditions in Tables 1-31 and 32. The assumed Reynolds numbers (Re) in the hot channel and cold
channel are 2,800 and 2,200, respectively. The following equations were selected for determinations of
heat transfer and diffusion boundary conditions (Perry’s 7th Ed.).
(a) Heat Transfer

hconvective ˜ d h
0.023 ˜ Re 0.83 Sc 0.44 (1-8)
k therm
(b) Tritium Diffusion
K diffusion ˜ d h
0.023 ˜ Re 0.83 Sc 0.44 (1-9)
Dtritium

Tables 1-31 and 1-32 summarizes the calculated heat transfer and diffusion coefficients in the PCHE
channels. All the fluid properties were obtained from National Institute of Standards and Technology
chemistry Webbook (2008).

40
Table 1-31. Calculated diffusion coefficients in the PCHE.
Diffusion kdiffusion kdiffusion Sc
Coefficient (hot) (Cold) (hot) Sc (cold)
m2/s m/s m/s
Primary Coolant Inlet 1.799E-05 0.385 0.866
Primary Coolant Outlet 1.048E-05 0.222 0.852
Secondary Coolant Inlet 1.020E-05 0.177 0.851
Secondary Coolant Outlet 1.782E-05 0.311 0.862
500°C 1.050E-05 0.213 0.174 0.765 0.765
600°C 1.294E-05 0.264 0.216 0.781 0.781
700°C 1.561E-05 0.318 0.261 0.777 0.777
800°C 1.782E-05 0.371 0.304 0.796 0.796
900°C 2.129E-05 0.436 0.357 0.785 0.785

Table 1-32. Calculated heat transfer coefficients in the PCHE.

W/m2K W/m2K
hconvection hconvection Pr Pr
(hot) (Cold) (hot) (cold)
Primary Coolant Inlet 7252.610843 0.659735352
Primary Coolant Outlet 5749.518929 0.656930324
Secondary Coolant Inlet 4667.650654 0.65679
Secondary Coolant Outlet 5883.805147 0.65969

The analysis was performed with a number of different boundary conditions. The following were the
conditions used (slightly modified from Ohashi and Sherman [2007]). As shown below, the temperature
in the fluid channels for tritium permeation analysis were matched to be the same for eliminating the
effect of temperature gradient.
Tritium Concentration in the Bulk Fluids
x Primary Coolant Tritium Concentration: 115 μCi/m3
x Secondary Coolant Tritium Concentration: 113 μCi/m3
x Primary Coolant Hydrogen Concentration: 200 ppm
Temperatures in the Bulk Fluids
For heat transfer analysis:
- Primary Coolant Inlet T: 900°C
- Primary Coolant Inlet P: 69.08 atm (7 MPa)
- Secondary Coolant Outlet T: 886.3°C
- Secondary Coolant Outlet P: 68.59 atm (6.95 MPa)
- Primary Coolant Outlet T: 495.5°C
- Primary Coolant Outlet P: 68.59 atm (6.95 MPa)
- Secondary Coolant Inlet T: 486.3°C

41
 - Secondary Coolant Inlet P: 69.08 atm (7 MPa).
For tritium permeation analysis:
- Primary Coolant T: 500°C
- Primary Coolant P: 69.08 atm (7 MPa)
- Secondary Coolant T: 500°C
- Secondary Coolant P: 69.08 atm (7 MPa)

- Primary Coolant T: 600°C

- Primary Coolant P: 69.08 atm (7 MPa)
- Secondary Coolant T: 600°C
- Secondary Coolant P: 69.08 atm (7 MPa)

- Primary Coolant T: 700°C

- Primary Coolant P: 69.08 atm (7 MPa)
- Secondary Coolant T: 700°C
- Secondary Coolant P: 69.08 atm (7 MPa)

- Primary Coolant T: 800°C

- Primary Coolant P: 69.08 atm (7 MPa)
- Secondary Coolant T: 800°C
- Secondary Coolant P: 69.08 atm (7 MPa)

- Primary Coolant T: 900°C

- Primary Coolant P: 69.08 atm (7 MPa)
- Secondary Coolant T: 900°C
- Secondary Coolant P: 69.08 atm (7 MPa).
The outer boundary conditions of the COMSOL models were set for two hot channels and two cold
channels. In order to accomplish the modeling, a periodic boundary condition was imposed on the wall of
the channels described above so that flux calculations assume an infinite array of hot and cold ports and
do not take into account the difference in flux at the sides of the heat exchanger. There are approximately
200,000 ports in the PCHE being considered for the NGNP, so the error associated with this assumption
is considered to be minimal.

1.4.1.3 Method of Grid Sensitivity Study

The mesh and grid size are very important issues in the computational fluid dynamics (CFD) or FEM
code analysis because they highly affect the quantitative values of the simulations. Generally, as the grid
and time step are refined, the spatial and temporal discretization errors should asymptotically approach
zero, excluding computer round-off error.
In this report, we followed the grid sensitivity analysis procedure proposed by the AIAA (1998) for
validation and verification of CFD codes. Since the basic concept of this grid sensitivity study is simply
based on the extrapolation methods, the guideline used in CFD codes are still conceptually valid for our
case. The following discussion summarizes the grid sensitivity study provided by the National
Aeronautics and Space Administration on the “NPARC Alliance CFD Verification and Validation”
Website.

42
Grid Consideration for a Grid Convergence Study
In generating a fine grid, one can build in the n levels of coarser grids by making sure that the number
of grid points in each coordinate direction satisfies the relation

N 2n m  1 (1-10)

where N is the total number of grid points in each coordinate direction, n is the level of grids, and m is an
arbitrary number defined for each coordinate direction.
For example: if two levels of coarser grids are desired (fine, medium, and coarse grids), the number
of grid points in each coordinate direction must equal 4m+1. The m may be different for each coordinate
direction. It is not necessary to halve the number of grid points in each coordinate direction to obtain the
coarse grid, but it is important to maintain the same grid generation parameters as the original grid.
Order of Grid Convergence
The order of grid convergence involves the behavior of the solution error defined as the difference
between the discrete solution and the exact solution as

E f (h)  f exact Ch P  H .O.T (1-11)

where C is a constant, h is some measure of grid spacing, p is the order of convergence, and H.O.T is
higher order terms. A “second-order” solution would have p=2.
Neglecting higher-order terms and taking the logarithm of both sides of the above equation results in

log( E ) log(C )  p log(h) (1-12)

The order of convergence p can be obtained from the slope of the curve of log(E) versus log(h).
A more direct evaluation of p can be obtained from three solutions using a constant grid refinement
ratio r,

§ f  f2 ·
p ln¨¨ 3 ¸¸ / ln(r ) (1-13)
© f 2  f1 ¹

where f1, f2, and f3 are results from fine, medium, and coarse grids, respectively.
The order of accuracy is determined by the order of the leading term of the truncation error and is
represented with respect to the scale of the discretization, h. The local order of accuracy is the order for
the stencil representing the discretization of the equation at one location in the grid. The global order of
accuracy considers the propagation and accumulation of errors outside the stencil. This propagation
causes the global order of accuracy to be, in general, one degree less than the local order of accuracy. The
order of accuracy of the boundary conditions can be one order of accuracy lower than the interior order of
accuracy without degrading the overall global accuracy.
Asymptotic Range of Convergence
Assessing the accuracy of code and calculations requires that the grid be sufficiently refined such that
the solution is in the asymptotic range of convergence. The asymptotic range of convergence is obtained
when the grid spacing is such that the spacing h and errors E result in the constancy of C as in

C E/hp (1-14)

43
Richardson Extrapolation
Richardson extrapolation is a method for obtaining a higher-order estimate of the continuum value
from a series of lower-order discrete values. A simulation will yield a quantity f that can be expressed in a
general form by the series expansion

f f h 0  g1 h  g 2 h 2  g 3 h 3  ... (1-15)

where h is the grid spacing. The quantity f is considered “second-order” if g1 = 0.0. fh=0 is the continuum
value when the mesh is infinitely small.
If one assumes a second-order solution and has computed f on a two-grid spacing of h1 and h2 with
h1 being the finer spacing, then one can write two equations for the above expansion, neglect third-order
and higher terms, and solve for fh=0 to estimate the continuum value,

f1  f 2
fh 0 # f1  (1-16)
r 2 1
where the grid refinement ratio is

f h2 / h1 (1-17)

The Richardson extrapolation can be generated for a p-th order methods and r-value of grid ration as

f1  f 2
fh 0 # f1  (1-18)
r p 1
Traditionally, Richardson extrapolation has been used with grid refinement ratios of r = 2. Thus, the
above equation simplifies to

4 1
fh 0 # f1  f 2 (1-19)
3 3
If a larger number of CFD computations are to be performed, one may wish to use the coarser grid
with h2. We will then want to estimate the error on the coarser grid. The Richardson extrapolation can be
expressed as

( f1  f 2 )r p
fh 0 # f2  (1-20)
r p 1
The estimated fractional error for f2 is defined as

H ˜rp
E2 (1-21)
r p 1
Richardson extrapolation is based on a Taylor series representation. If there are shocks and other
discontinuities present, the Richardson extrapolation is invalid in the region of the discontinuity.

44
Grid Convergence Index (GCI)
Roache suggested a grid convergence index (GCI) to provide a consistent manner of reporting results
of grid convergence studies and provide an error band on the grid convergence of the solution.
One significant issue in numerical computation is determining what level of grid resolution is
appropriate. This is a function of flow conditions, type of analysis, geometry, and other variables.
The GCI is a measure of the percentage the computed values are away from the values of the
asymptotic numerical value, which indicates an error band on how far the solution is from the asymptotic
value and how much the solution would change with a further refinement of the grid. A small value of the
GCI indicates that the computation is within the asymptotic range.
The GCI on the fine grid is defined as

Fs H
GCI fine (1-22)
(r p  1)

where Fs is a factor of safety. The refinement may be spatial or in time. The factor of safety is
recommended to be Fs = 3.0 for comparisons of two grids and Fs = 1.25 for comparisons over three or
more grids. The higher factor of safety is recommended for reporting purposes and is quite conservative
of the actual errors.
The GCI for the coarser grid is defined as

Fs H r p
GCI fine (1-23)
(r p  1)

It is important that each grid level yield solutions that are in the asymptotic range of convergence for
the computed solution. This can be checked by observing two GCI values as computed over three grids,

GCI 23 r p GCI 12 (1-24)

Required Grid Resolution

If a desired accuracy level is known and results from the grid resolution study are available, one can
then estimate the grid resolution required to obtain level of accuracy,
1/ p
* § GCI * ·
r ¨¨ ¸¸ (1-25)
© GCI 21 ¹
Effective Grid Refinement Ratio
If one generates a finer or coarser grid and is unsure of the value of grid refinement ratio to use, one
can compute an effective grid refinement ratio as
(1 / D )
§ N1 ·
reffective ¨¨ ¸¸ (1-26)
© N2 ¹
where N is the total number of grid points used for the grid and D is the dimension of the flow domain.
This effective grid refinement ratio can also used for unstructured grids.

45
1.4.1.4 Results and Discussions
Heat and Tritium Diffusion Analysis in the PCHEs
Figures 1-22 and 1-23 are a graphical representation of the concentration and temperature profiles
using the Primary Inlet Conditions in the two reference configurations (standard and off-set). Analysis of
flux data for seven different temperatures and two different pressures showed a 1% decrease in tritium
flux and a negligible decrease in heat flux when calculated using the reference geometries (horizontal
pitch=1.464 mm; vertical pitch=0.96 mm; diameter=1.2 mm. The heat flux for the off-set configuration
was equal to 99.98% of the initial value for the reference configurations. Flux values were extrapolated to
find the value as the mesh size approaches zero using Richardson’s method, summarized in the previous
section. The error associated with this extrapolation is 1.5% for the initial configuration and 1.7% for the
offset configuration when calculating tritium flux.
(a) Temperature Profiles.

(b) Concentration Profiles.

Figure 1-22. Calculated temperature and tritium concentration profiles

(for the standard configuration).

46
 (a) Temperature profile.

(b) Concentration profile.

Figure 1-23. Calculated temperature and tritium concentration profiles (for the off-set
configuration).

This analysis showed that the majority of resistance to heat transfer is found in the fluid boundary
layer, not in the metal itself. In contrast, the majority of resistance to tritium flux is found in the metal
itself. Knowing this, the tritium flux can be reduced through the metal without greatly reducing heat
transfer.

47
 Figures 1-24 and 1-25 show profiles of tritium diffusion flux and the main penetration paths for
standard and off-set configurations, respectively. As shown in these figures, the tritium flux on the flow
channel is not uniform along the surface. Most of the tritium diffusion is concentrated on the center part
of the semicircle area. On the other hand, there is very small tritium penetration on the channel side
because of very small concentration gradient in this direction. The dead spot area is highly dependent on
the channel horizontal pitch.
(a) Diffusion flux.

(b) Diffusion path.

Figure 1-24. Diffusion flux and path of tritium in the PCHE (for the standard configuration).

48
 (a) Diffusion flux.

(b) Diffusion path.

Figure 1-25. Diffusion flux and path of tritium in the PCHE (for the off-set configuration).

Figure 1-26 shows the tritium flux for different horizontal pitches and plate thickness. These
calculations represent the flux at the primary coolant inlet side of a counter-current PCHE heat exchanger.
This figure shows the tritium penetration flux is significantly decreased by increasing the plate thickness.
However, the change of horizontal pitch does not highly affect the tritium penetration flux. According to
our calculation, increased horizontal pitch slightly increases the overall tritium penetration rate by
increasing the tritium concentration gradient with a longer horizontal pathway on the side of the channels.

49
 (a) Standard configuration.

(b) Off-set configuration.

Figure 1-26. Tritium flux for different horizontal pitches.

The decreasing trend of the 0.69 mm vertical pitch on Figure 1-26 (b) still shows higher tritium flux
compared with that of higher vertical pitches. The decrease is contributed by the longer horizontal
pathway of the channels. The tritium concentration gradient at higher than 0.96 mm vertical pitch is
uniformed, resulting in slight changes in the flux.
Figure 1-27 shows the tritium flux for a different plate thickness. As shown, the tritium flux is
exponentially reduced with increased vertical pitch (plate thickness). At the small vertical pitch, the off-
set configuration showed slightly smaller tritium penetration, but they were almost identical for the large
thickness.

50
 3.5e-5

3.0e-5 Standard

Avergage Tritium Flux (PCi / m2 hr)

Off-set

2.5e-5

2.0e-5

Horizontal Pitch = 1.332 mm

1.5e-5

1.0e-5

5.0e-6

0.0
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Vertical Pitch (mm)

Figure 1-27. Tritium flux for different vertical pitch (plate thickness).

In summary, heat transfer and tritium penetration analyses have been performed for standard and off-
set channel configurations of the PCHE. Depending on such factors as horizontal and vertical pitch, it was
learned that the off-set configuration can be favorable over the standard configuration. It was also learned
that flux is weakly dependent on the horizontal pitch—the flux slowly increases as the horizontal pitch
slowly increases. This is because of an increased dead space as shown in Figures 1-24 and 1-25. The one
exception to this is in the off-set configuration. If the horizontal pitch to vertical pitch ratio is very large,
the flux starts to decrease as the horizontal pitch increases. The vertical pitch and flux relationship was
much simpler; as the vertical pitch (plate thickness) increases, the flux decreases, which makes sense
because there is a longer flux pathway.
Tritium flux was seen to vary widely as the model geometry changed. The majority of the resistance
to tritium flux was through the metal, so that as the shape of the metal changed, the flux changed
respectively. The flux varied from 23.77 to 262.11% of the reference flux value (13.3 μCi/m2h) as the
geometry changed.
Heat flux changed less as the model geometry changed. The majority of the resistance to heat flux
was through the helium boundary layer, so the flux only changed a little as the shape of the metal changed.
The heat flux varied from 84.28 to 102.89% of the reference value (149.22 MW/m2) as the geometry
changed.
The off-set configuration tended to have slightly less flux for a given horizontal and vertical pitch
because the flux pathway tended to be slightly longer, as can be seen in Figures 1-24 and 1-25. This effect
was the greatest when the horizontal pitch was large and the vertical pitch was small, because, in this case,
the flux pathway was almost lateral instead of longitudinal.

1.4.2 ABAQUS Modeling for Stress in the PCHE

Stress analysis was performed on the Heatric’s PCHE proposed for the VHTRs using ABAQUS ®
(ver. 6.75). The purpose of this study was to observe the change in maximum stress to variation in design
parameters. For the maximum stress, parametric studies were performed on the normal and off-set
channel arrangements as function of the channel’s vertical and horizontal pitches. Effects of outer wall
pressure and the channel’s sharp corner fillet size were also studied on the normal channel arrangement.

51
1.4.2.1 Geometry and Dimensions
Figure 1-28 shows the cross sectional views of the PCHE for original and offset arrangements. For
the offset arrangement, the rows were offset from each other by half the horizontal pitch, which should be
easily accomplished by shifting plates during construction.

Figure 1-28. Normal (left) and offset (right) channel configurations.

The reference channel dimensions and operating conditions are shown in the Table 1-33. The
reference diameter used was the value recommended for nuclear application by Heatric during a
workshop meeting with MIT for discussion on the use of PCHE for a gas-cooled fast reactor (Gezelius
2004, p. 167). Meanwhile, the pitch and plate thickness were based on standard design ratios to the
diameter employed by Heatric and other PCHE modelers (Gezelius 2004, p. 88). The heat exchanger
body temperature was set to 850°C, well above the minimum temperature of 800°C required for SI cycle
(Lillo et al. 2005, p. 1). The primary loop pressure was set according to operating pressure of the core
(Natesan et al. 2006, p. 85). The secondary loop pressure was set a little below the primary loop channel
to observe the stress induced by the pressure difference between the primary and secondary channels. It
has been suggested to immerse the entire IHX in a vessel operating at a pressure similar to IHX loops to
relieve stress by reducing the pressure difference between the internal channels and outside of the IHX.
Hence, the outside pressure was set to 6 MPa, the average of the primary and secondary loop pressures.
Table 1-33. Default channel dimensions and operating conditions.
Default channel dimensions
Diameter [mm] 1.2
Horizontal pitch [mm] 1.464
Vertical pitch [mm] 0.96
Default operating conditions
Heat exchanger body temperature [C] 850
Primary (hot) loop channel pressure [MPa] 7
Secondary (cold) loop channel pressure [MPa] 5
Outside pressure [MPa] 6

1.4.2.2 Material Properties

The stress analysis was performed for Alloy 617, one of the primary candidate for the PCHE
(Natesan et al. 2006, p. i). Basic mechanical properties for Alloy 617 at several temperatures are shown in
the Table 1-34. Properties for the 850°C are interpolated from 800°C and 900°C properties. The
maximum stress observed from this study was well below the yield stress, thus the stress strain response
was assumed to be linear with constant Young’s modulus.

52
Table 1-34. Alloy 617 mechanical properties (Special Metals Corporation, 2006, p. 2-3).
Temperature Young’s Modulus Yield Stress
[°C] [GPa] Poisson’s Ratio [MPa]
800 149 0.3 290
850 153 0.3 280
900 157 0.3 200

Figure 1-29 shows the stress versus time-to-1% strain and time to rupture from creep for Alloy 617.
Considering 1% strain must occur before the rupture, the time-to-1% strain can be used as the
conservative limit for the allowed stress. Based on Figure 1-29, the allowed stress for 850°C is
approximately 20 MPa for 11.4 years. By extrapolating the 850°C trend line to 50 years, the maximum
allowed stress was estimated as 10 MPa.

Figure 1-29. Stress versus time-to-1% strain and time to rupture from creep for several
temperatures (Schubert 1984, p. 328).

1.4.2.3 Assumptions
It is acknowledged the actual heat exchanger has some pressure drop associated with flow across the
hot and cold loops, and there is a temperature profile in both the cross sectional plane and axial direction.
However, the analysis was done in 2-D, which assumes there is no stress variation in axial direction, and
material properties were evaluated assuming uniform temperature. These are reasonable considering the
model represents a differential segment on the front of the heat exchanger where the temperature is
highest, which is the location most vulnerable to creep strain failure. When modeling such a small slice of
axial segment, variation of temperature and pressure in axial direction can be neglected. The heat
exchanger body temperature can be assumed to be uniform because the temperature gradient from hot to
cold channels at a given axial position is small (a couple of degrees Celsius), which makes negligible
difference to material properties. The cross-sectional temperature profile generated using COMSOL is
shown in the previous section.

53
 For future reference, the ABAQUS will need to a 3-D model to perform thermal stress analysis,
including axial profile. The project started out with 3-D thermal stress analysis as part of the study, but it
was postponed and not included in this report because of time constraints.

1.4.2.4 Stress vs. Number of Heat Transfer Units

Model Geometry
The heat exchanger model is composed of heat transfer units (HTU) that are each composed of a hot
and a cold channel surrounded by the outer wall. From a model size stand point, it is impractical to
include all of the HTUs present in an actual IHX, so a convergence study was done to determine the
minimum number of HTUs needed.
ABAQUS models were analyzed, including HTUs 1, 4 (2x2), 9 (3x3), 25 (5x5), 49 (7x7), and 81
(9x9). For example, Figure 1-30 shows the ABAQUS models for HTUs 1, 9 and 25. Since the model is
symmetric about the center, the model was cut in half to reduce the model size. As mentioned earlier, the
channel diameter is 1.2 mm, horizontal pitch is 1.464 mm, and vertical pitch is 0.96 mm. The outer wall is
10.18 mm thick vertically and 10.132 mm thick horizontally. Although not visible on the figure, the
models are 0.1 mm thick in axial direction (in to the page).

Figure 1-30. ABAQUS models for different numbers of HTUs.

Boundary conditions were given to constrain the front and back in axial direction, which effectively
makes the analysis identical to 2-D. It is acknowledged that the thickness is not necessary in the 2-D
analysis. Originally the models were created to perform such analyses as axial temperature profile for
thermal stress, which required 3-D models, but thermal stress analysis was eventually postponed because
of time constraint and limitations on the model size. Hence, the models’ 3-D feature remains as a vestigial
but unnecessary part of the 2D analysis performed.

54
 As shown in Figure 1-31, round fillets were used (0.02
mm in radius) on the channels’ sharp corners. This is very
important to this particular stress analysis because the
maximum stress occurs on the corner. If the corner is modeled
as a sharp corner, theoretically the stress is infinite at this
location, and stress analysis result will continue to increase
with continued mesh refinement (ABAQUS 2004, p.4.49).
Normally, omitting such fine detail has a negligible effect on
the overall response of the model, but stress a profile close to
the singularity will be inaccurate (ABAQUS 2004, p.4.49).
Since exact stresses in the corners are required, fillets were
used on the corners.
Boundary Conditions
Pressure applied to the hot channel walls, cold channel
walls, and outer walls were 7 MPa, 5 MPa, and 6 MPa, Figure 1-31. Close up view of a channel
respectively. The symmetric boundary condition with respect corner.
to x-axis direction was applied to the cut region across the
center, which constrains displacement in x direction and rotation about y and z (axial) axes. The
symmetric boundary condition with respect to z-axis direction was applied to the front and back surfaces,
which constrains displacement in z direction and rotation about x and y axes. The last boundary condition
essentially makes the model identical to a 2D model, which assumes such constraints by default.
Partitioning and Meshing
A locally refined meshing scheme was used to minimize the number of elements. The meshing
scheme used is summarized in the Table 1-35. Figure 1-32 shows the model for 25 HTUs, which has the
fine seeded borders marked in red and also shows the partition scheme. Figure 1-33 shows the finished
mesh grid around a half channel.
Table 1-35. Meshing scheme for HTU study.
Fine seed (red) [mm] 0.01
Channel corner fillet [divisions] 6
Remainder (black) [mm] 0.2
Mesh control Wedge sweep
Element type C3D6*
Geometric order Linear
Horizontal partition distance, Ph [mm] 3
Vertical partition distance, Pv [mm] 2

* 3-D continuum element with 6 node points (prism shape).

55
Results
Figure 1-34 shows the maximum Von Mises stress as a function of heat transfer units used. The
maximum stress limits move out as the number of units used increases, and it is virtually identical from
25 units and beyond. Based on this trend, the 25 HTU model was used on parametric studies for outside
pressure, vertical pitch, horizontal pitch, and temperature.
The stress distributions for the 25 HTU model are shown in Figures 1-35, 1-36, and 1-37. Figure 1-35
shows the entire model, Figure 1-36 shows the first three rows of second and third columns, and Figure 1-
37 shows a zoomed-in view of the corner with the maximum stress, which is the left corner of the second
row, second column channel. As shown in Figure 1-35, high stress occurs on channel walls, while the
outer walls are subjected to relatively low stress. Figure 1-36 shows that the stress is concentrated toward
channel corners and is higher on low-pressure channel corners. Figure 1-37 presents a close-up view of
the corner experiencing the maximum stress, which is concentrated in close proximity to the corner’s
curve. High-pressure channel walls expand outward while the low-pressure channel walls get compressed
inward. The compression from high-pressure channels invokes a pivot and lever mechanism to the low-
pressure channel corners, causing a high concentration of stress.

Maximum Stress vs Units

16.5

16
Maximum Stress [MPa]

15.5

15

14.5

14
0 10 20 30 40 50 60 70 80 90
Units

Figure 1-34. Maximum stress as function of heat transfer units used.

57
Figure 1-35. Stress distribution of 25 HTUs model, entire model view.

58
 Figure 1-36. Stress distribution of 25 HTUs model, top three rows, second and third columns.

Figure 1-37. Stress distribution of 25 HTUs model, left corner of second row, second column channel
view.

59
1.4.2.5 Stress vs. Outside Pressure
Model Geometry, Operating Conditions and Meshing
The 25 HTU model in normal arrangement was used for the outside pressure study. The default
channel dimensions and operating conditions shown in the Table 1-33 were used, and only the outside
pressure was varied. A meshing scheme identical to the one used in HTU convergence study was used
(see Table 1-34).
Results
Figure 1-38 shows the maximum Von Mises stress as a function of the outside pressure. As can be
seen, the stress is minimized when outer wall pressure is set to 6 MPa, right between the hot and cold
channel pressure. This was anticipated because at 6 MPa, the pressure differences from hot and cold
channels to the outside pressure are minimized. Based on this, 6 MPa was used for the outside pressure on
all of parametric studies.

Maximum Stress vs Outside Pressure

60

50
Maximum Stress [MPa]

40

30

20

10
0 1 2 3 4 5 6 7 8 9
Pressure [MPa]

Figure 1-38. Maximum stress as function of the outside pressure for the normal configuration.

1.4.2.6 Stress vs. Vertical and Horizontal Pitches

Model Geometry, Operating Conditions, and Meshing
For these parametric studies, the 25 HTU model was used in normal and offset arrangements for
comparison. The reference channel dimensions and operating conditions were used, and only one
parameter (either vertical or horizontal pitch) was varied per case. A slightly altered meshing scheme was
used for vertical and horizontal pitch studies because of limitations to the model size, which increases
substantially with increase in pitches. In order to use same mesh scheme for the vertical and horizontal
pitch studies, the number of elements generated had to be reduced. The new meshing scheme is
summarized in the Table 1-36.

60
Table 1-36. Meshing scheme for vertical pitch and horizontal pitch studies.
Fine seed (red) [mm] 0.014
Channel corner fillet [divisions] 4
Remainder (black) [mm] 0.2
Mesh control Wedge sweep
Element type C3D6
Geometric order Linear
Horizontal partition distance, Ph [mm] 1.5
Vertical partition distance, Pv [mm] 1

Maximum Stress vs. Vertical Pitch

Figure 1-39 shows the maximum Von Mises stress as a function of the vertical pitch. As can be seen,
the maximum stress decreases with the increasing vertical pitch. The maximum stress occurs on a low-
pressure channel corner because of a pivot and lever mechanism where the corner as the pivot and the
walls are the lever. The low-pressure channel walls are compressed inward because of high-pressure
channels above and below, therefore the most important thing in reducing stress at the corner is reducing
deformation of walls at its proximity, which is accomplished by making the wall between channels
thicker. The deformation is greater for thinner walls, causing greater stress on the low-pressure channel
corners. On the other hand, increasing the thickness makes the body more resistant to deformation
induced by the pressure difference, thus, reducing the stress on the corner.

Figure 1-39. Maximum stress as function of the vertical pitch for normal and shifted configurations.

Figures 1-40 and 1-41 show the stress distributions for normal and shifted arrangement with 0.8 mm
vertical pitch. Offsetting the channels reduces the stress because the high-pressure channels are no longer
directly above and below the low-pressure channels as seen in Figure 1-41, thus relieving them from the
compression caused by the pressure difference.

61
Figure 1-40. Normal 0.8 mm vertical pitch.

Figure 1-41. Shifted 0.8 mm vertical pitch.

62
 Judging from the trend, the stress will not decrease with vertical pitch indefinitely. This is expected
since the stress is concentrated around the area near the channel. When the vertical pitch is very thin,
increasing its thickness results in significant redistribution of the stress to a larger area. In contrast, when
the vertical pitch is very thick, increasing its thickness further makes no difference to the stress
redistribution. Figures 1-42 and 1-43 show the stress distributions for normal and shifted arrangement
with 2.4 mm vertical pitch, which can be compared to Figures 1-40 and 1-41 for differences in the stress
profile of the walls between channels.

Figure 1-42. Normal 2.4 mm vertical pitch. Figure 1-43. Shifted 2.4 mm vertical pitch.

63
Maximum Stress vs. Horizontal Pitch Maximum Stress vs Horizontal Pitch
26

Figure 1-44 shows the maximum Von Normal

22
Mises stress as a function of the horizontal

Maximum Stress [MPa]

Offset
pitch. As can be seen, the stress sharply
18
declines and gradually limits out with the
increase in the horizontal pitch. The
14
sidewalls between channels essentially
hold the top and bottom walls together
10
from falling apart by pressure force exerted
in vertical directions. The size of channels
6
does not change, so the total pressure 1 1.5 2 2.5 3 3.5 4 4.5
forces in vertical direction also stay the horizontal pitch [mm]
same. On the other hand, as the sidewall
thickness is reduced, the total pressure Figure 1-44. Maximum stress as function of the horizontal
force must be counteracted by a smaller pitch for normal and shifted configurations.
cross-sectional area of thinner sidewall,
thus increasing the stress concentration. Because of this, the strain in the proximity of the channel corner
is increased, resulting in greater maximum stress. The stress is relieved with increasing horizontal pitch
because the stress gets distributed over thicker walls. Figures 1-45 and 1-46 show stress distributions for
normal and offset arrangements with 1.264 mm horizontal pitch, and Figures 1-47 and 1-48 show stress
distributions for normal and offset arrangements with 2.4 mm horizontal pitch, which can be compared
for difference in stress distribution on sidewalls of different thicknesses.
As shown in Figures 1-39 and 1-44, increasing horizontal pitch is particularly more effective for the
shifted configuration because the high-pressure channels are not directly above and below the low-
pressure channels and, as the horizontal pitch increases, the high-pressure channels move away from these
regions, further reducing the compression caused by the pressure difference. When the horizontal pitch
increases beyond twice the diameter, the high-pressure channels are completely out of the region above
and below the low-pressure channels. At this pitch, the channels right above and below each are the
channels on every other rows, which have identical pressures, relieving the pressure difference between
the channels.
As shown in Figure 1-44, eventually, increasing the horizontal pitch beyond 1.7 mm for the normal
configuration makes no difference to the maximum stress. As shown in Figures 1-47 and 1-48, when the
sidewall is very thick, the stress is fully redistributed over its thickness and negligible change would occur
to the stress distribution with further increase in horizontal pitch. At very thick horizontal pitch, the only
dominant stress mechanism on the corner is the pivot and lever effect. It is most likely that the pivot and
lever mechanism is not significantly affected by change in horizontal pitch because channels adjacent to
left and right are channels with same pressures. Thus deformation because of pressure difference like the
top and bottom walls in the study with vertical pitch is not present on sidewalls.
The minimum stress is reached sooner for normal arrangement because, as shown in Figures 1-45 and
1-47, it does not have secondary effect of shifting the high and low pressures out of alignment. For the
offset arrangement, the minimum stress should result once the high and low-pressure channels are
completely out of alignment, which is reflected in its trend.

64
Figure 1-45. Normal 1.264 mm horizontal pitch.

Figure 1-46. Shifted 1.264 mm horizontal pitch.

65
Figure 1-47. Normal 2.6 mm horizontal pitch.

Figure 1-48. Shifted 2.6 mm horizontal pitch.

66
 2. TRITIUM BEHAVIOR STUDY IN THE VHTR SYSTEM

2.1 Introduction
In the VHTR hydrogen production system, tritium permeation from the core into the produced
hydrogen is currently one of the most significant issues, because hydrogen isotopes can easily permeate
through the high temperature heat exchanger tubes and contaminate product hydrogen. If the hydrogen
produced in this system was then used as fuel in fuel cell vehicles, that tritium water would be released to
the environment. Because of this potential, tritium contamination of hydrogen is considered in this study.
The tritium generation mechanism is well described by Gainey (1976), and Ohashi and Sherman
(2006). The primary tritium birth mechanism is ternary fission of fuel (e.g., 233U, 235U, 239Pu, and 241Pu)
because of thermal neutrons. Tritium is also generated in VHTR from 6Li, 7Li, 3He, and 10B by neutron
capture reactions as the following:
6
Li (n, Į) 3H, (2-1)
7
Li (n, nĮ) 3H (2-2)
3
He (n, p) 3H (2-3)
10
B (n, 2Į) 3H (2-4)
10
B (n, Į) 7Li . (2-5)
6
Li and 7Li are impurities in the core graphite material such as the sleeve, spine, reflector, and fuel
matrix. 3He is an impurity in the reactor coolant helium. Because helium coolant leaks from the primary
loop to the containment vessel, helium is supplied to the primary coolant as a make-up with an impurity
of 3He. 10B exists in control rods, burnable poisons, and reflectors. Tritium is produced directly from 10B
via Equation (2-4) or via the chain reaction in Equations (2-2) and (2-5).
Tritium generated in the fuel particles by ternary fissions can escape into a primary coolant,
permeating several barriers of the fuel particles. In addition, tritium born from 10B and 6Li can pass into
the primary coolant. The principal chemical form for tritium in the reactor coolant was reported as HT
(1H-3H) because of the isotope exchange reaction between T2 (3H-3H) and H2 (Wichner and Dyer 1979).
Some of the tritium in the primary coolant is removed by a purification system installed in the primary
loop. Some of the tritium can escape outside the coolant by permeation through the components and
piping and by leakage with the primary helium coolant. The remaining tritium in the primary coolant
permeates through the heat transfer tubes or surfaces of the IHX and gets mixed in with the secondary
coolant.
In the secondary loop, some of the tritium is removed by the purification system or escapes outside,
just as tritium behaves in the primary loop. The remainder of the tritium in the secondary coolant
permeates through heat transfer surfaces and gets mixed into the tertiary coolant. Transportation of tritium
into the tertiary coolant is the same as for the secondary coolant. It permeates through the heat transfer
surfaces of the process heat exchangers and is mixed into the process chemicals of the hydrogen plant.
Tritium that has permeated the tertiary loop going to the hydrogen plant can react with
hydrogen-containing process chemicals through isotopic exchange reactions. For example, the high-
temperature electrolysis (HTE) process contains H2O. Therefore, HTO (tritiated water) is produced by the
isotope exchange reaction between HT and H2O. Gaseous HTO and HT escape from the HTE process
with the product hydrogen and oxygen. A part of the liquid HTO flows out from the hydrogen plant with
the drain water. The remainder of the liquid HTO circulates into the plant with the recycling water and
accumulates in the water of the HTE process. The SI process used in hydrogen production also contains

67
H2O, H2SO4, and HI chemicals from which HTO, HTSO4, and TI may be produced through isotope
exchange reactions. These tritium-containing chemicals circulate with and accumulate in process the
chemicals. Only gaseous HT and HTO can escape from the SI process with the product hydrogen and
oxygen.
Tritium behavior in high temperature gas-cooled reactors (HTGR) has been was well evaluated in
several countries in the 1970s (e.g., the Dragon reactor in England (Forsyth 1972), the Peach Bottom
HTGR in the U.S. (Wichner and Dyer 1979), and Arbeitsgemeinschaft Versuchsreaktor (AVR) in
Germany (Steinwarz, Rohrig, and Nieder 1980). Data from the operation of HTGRs and from laboratory
experiments revealed the mechanism of tritium production, transport, and release to the environment. In
his review of tritium behavior in an HTGR system, Gainey’s (1976) calculations show that tritium
releases should be well within current federal guidelines for the nuclear plant. For example, the estimated
maximum dose to an average adult for a typical 3,000-MWt HTGR with a cooling tower is
0.38 milligram/year, which is slightly more than one-tenth of the maximum annual dose allowed
(Gainey 1976). For this reason, no further laboratory-scale work on tritium was required at that time.
However, their tritium calculations were only concerned with general tritium release and did not examine
questions related to nuclear hydrogen production. NGNP will use the HTGR as a heat source to produce
hydrogen for industry or individual users..
Ohashi and Sherman (2007) recently estimated steady-state tritium movement and accumulation in an
NGNP with a hydrogen plant using a high temperature electrolysis process and a thermochemical water
splitting sulfur-iodine process using the numerical code THYTAN. Estimated tritium concentrations in
product hydrogen and in process chemicals in the hydrogen plant of the NGNP using the high temperature
electrolysis process were slightly higher than the drinking water limit defined by the U.S. Environmental
Protection Agency and the limit in the effluent at the boundary of an unrestricted area of a nuclear plant
as defined by the U.S. Nuclear Regulatory Commission. However, modified designs and operations could
bring these concentrations within specified limits. Tritium concentrations in the NGNP using the sulfur
iodine process were calculated to be significantly higher and were affected by parameters with large
uncertainties (tritium permeability of the process heat exchanger, the hydrogen concentration in the heat
transfer and process fluids, and the equilibrium constant of the isotope exchange reaction between HT and
H2SO4). These parameters, including tritium generation and the release rate in the reactor core, should be
more accurately estimated in the near future to improve the calculations for the sulfur-iodine process.
Decreasing the tritium permeation through the heat exchanger between the primary and secondary circuits
may be an effective measure for decreasing tritium concentrations in product hydrogen, hydrogen plant,
and tertiary coolant.
Based on Ohashi and Sherman’s calculation, this study focused on the predictions of the transient
tritium behaviors in the NGNP system, finally evaluating how much tritium is contained in the produced
hydrogen. To accomplish this, the following tasks were performed in FY 2008:
x Developed a dynamic tritium behavior analysis tool by MATLAB SIMULINK (Mathworks, Inc.
2004)
x Followed-up THYTAN verification and validation work by Ohashi and Sherman (2008)
x Developed THYTAN core input for considering transient tritium source
x Evaluated the effective thickness for tritium permeation in the PCHE type heat exchangers
x Developed a Tritium Penetration Analysis Code for VHTR and Hydrogen Generation System.

68
 A dynamic tritium behavior analysis code was developed by the Matlab Simulink software package.
This code can solve the generations of tritium source and tritium transport equations, including the
penetration of tritium through the heat exchanger wall. The following phenomena are currently taken into
consideration in this code:
x Tritium birth by ternary fission in the fuel particle and through neutron absorption reactions of 6Li,
7
Li, 10B, and 3He in the core and tritium release to the primary coolant helium
x Tritium and hydrogen permeation through the heat transfer surfaces of the heat exchanger (e.g., IHX),
the chemical reactor, and the recuperator
x Tritium and hydrogen permeation assuming a co-axial pipe in the primary and intermediate loops
x Tritium and hydrogen permeation to the outside through the outer walls of components and piping
x Tritium and hydrogen removal by the purification system installed in the primary and secondary
loops
x Tritium and hydrogen leakage to the atmosphere or to another loop that accompanies general bulk
helium leakage.

2.1.1 Basic Equations

The following governing equations were solved in the MATLAB SIMULINK (Mathworks, Inc.
2004) code for analysis of the tritium behaviors in the NGNP system. The basic equations include mass
balance, core tritium source, tritium permeation, leakage, and purification system. The ion exchange
model is not yet considered in this code. The following equations are based on the report by Ohashi and
Sherman (2006).

2.1.1.1 Mass Balance

The current tritium analysis code basically solves the mass balance of tritium-containing chemicals
and hydrogen. The mass balance of tritium-containing chemicals and hydrogen in each block is calculated
using the following basic equations:
dC i , j
Vj
dt
¦ F total , j 1 ˜ Ci , j 1   Ftotal , j ˜ C i , j  S i , j (2-6)

Si, j  Rcore,i , j  RHX ,i , j  Rcopipe,i , j  Rcomp ,i , j  Rleak ,i , j  RPF ,i , j  Rreaction ,i , j (2-7)

where
Vj = the volume of block j [m3 (STP)]
Ci, j = the volume fraction of chemical i in block j [m3 (STP)/m3 (STP)]
t = time [s]
Ftotal , j = the volumetric flow rate of all chemicals in block j [m3 (STP)/s]
S i, j = the total amount of volume change rate of chemical i in block j by considering
generation, release, permeation, removal, leakage, and isotope exchange reactions
[m3 (STP)/s]
Rcore,i , j = volumetric release rate from the core to the primary coolant [m3 (STP)/s]
R HX ,i , j = volumetric permeation rate at heat exchanger (i = H2 and HT) [m3 (STP)/s]
Rcopipe,i , j = volumetric permeation rate at co-axial pipe (i = H2 and HT) [m3 (STP)/s]

69
 Rcomp ,i , j = volumetric permeation rate to outside (i = H2 and HT) [m3 (STP)/s]
Rleak ,i , j = volumetric leak rate with helium leakage (i = H2, HT and HTO) [m3 (STP)/s]
R PF ,i , j = volumetric removal rate by purification system (i = H2, HT and HTO) [m3 (STP)/s]
Rreaction ,i , j = volumetric reaction rate by isotope exchange reactions
(i = H2, HT, HTO, HTSO4 and TI) [m3 (STP)/s].

2.1.1.2 Core Model

There are five different sources for tritium generation in the reactor core: ternary fission, birth from
6
Li, birth from 7Li, birth from 3He, and birth from 10B. The detail equations are expressed as follows.
Ternary fission

d N T (Ter ) 
K ˜ P ˜ Y  O ˜ N T (Ter ) (2-8)
dt
where
N T (Ter ) = number of tritium atoms from ternary fission
K = fission rate per thermal megawatt [fission/MW/s]
P = reactor power [MW]
Y = average yield per fission [1/fission]
O = tritium decay constant [1/s].

Birth from 6Li

d N Li 6 
I th ˜ V Li 6T ˜ N Li 6 (2-9)
dt

d N T ( Li 6 ) 
I th ˜ V Li 6T ˜ N Li 6  O ˜ N T ( Li 6) (2-10)
dt
where
N Li 6 = number of 6Li atoms
N T ( Li 6 ) = number of tritium atoms from 6Li
I th = thermal neutron flux [neutrons/cm2/s]
V Li6T = effective cross section for 6Li (n, Į) 3H [cm2].

Birth from 7Li

d N Li 7 
I f ˜ V Li 7 H 3 ˜ N Li 7 (2-11)
dt

d N T ( Li 7 ) 
I f ˜ V Li 7T ˜ N Li 7  O ˜ N T ( Li 7 ) (2-12)
dt

70
where
N Li 7 = number of 7Li atoms, excluding 10B source
N T (Li 7 ) = number of tritium atoms from 7Li, excluding birth from 10B
If = fast neutron flux [neutrons/cm2/s]
V Li7T = effective cross section for 7Li (n, nĮ) 3H [cm2].

Birth from 3He

d N He3 
3  f ˜ N He 3  I He ˜ V He 3T ˜ N He 3
$
f ˜ N He (2-13)
dt

d N T ( He3) 
I He ˜ V He3T ˜ N He3  O ˜ N T ( He3) (2-14)
dt

Wcore
I He ˜ I th (2-15)
Wtotal

where
N He3 = number of 3He atoms
N T ( He3) = number of tritium atoms from 3He
f = fractional supply rate of helium coolant [1/s]
$
N He 3 = number of 3He atoms in the supply helium
V He3T = effective cross section for 3He (n, p) T [cm2]
I He = average thermal neutron flux experienced by the total primary helium inventory [n/cm2/s]
Wcore = helium inventory in core [kg]
Wtotal = total primary helium inventory [kg].

Birth from 10B

d N B10 
I th ˜ V B10 Li 7  I f ˜ V B10T  ˜ N B10 (2-16)
dt
d N Li 7 ( B10) 
I th ˜ V B10 Li 7 ˜ N B10  I f ˜ V Li 7T ˜ N Li 7 ( B10) (2-17)
dt

d N T ( B10) 
I f ˜ V Li 7T ˜ N Li 7 ( B10)  I f ˜ V B10T ˜ N B10  O ˜ N T ( B10) (2-18)
dt

where
N B10 = number of 10B atoms
N Li 7 ( B10) = number of 7Li atoms from 10B
N T ( B10) = number of tritium from 10B
V B10 Li 7 = effective cross section for 10B (n, Į) 7Li [cm2]
V B10T = effective cross section for 10B (n, 2Į) 3H [cm2].

71
Tritium release rate

The tritium release rate from the core to the primary coolant, Rcore, HT , j [m3 (STP)/s], is calculated by
using the following equations:

N T (total ) ­ d N T (Ter )  ½ ª ­ d N T ( Li 6)  d N T ( Li 7 )  ½º
dt ¦ ®¯D Ter
dt
¾
¿
¦ «¬D Li ®
¯ dt

dt
¾»
¿¼
­ d N T ( He3) ½ ­ d N T ( B10)  ½
 ¦ ®¯D He3
dt
¾
¿
¦ ®¯D B10
dt
¾
¿ (2-19)

N T (total ) 1 R ˜ T0
Rcore, HT , j u u
dt NA P0 (2-20)
where
N T (release) = number of tritium atoms released to the primary coolant
D Ter = fractional release ratio of tritium produced from ternary fission
D Li = fractional release ratio of tritium produced from 6Li and 7Li
D He3 = fractional release ratio of tritium produced from 3He
D B10 = fractional release ratio of tritium produced from 10B
NA = Avogadro constant
R = gas constant (8.314)
T0 = standard temperature (273.15 K)
P0 = standard pressure (1.0 1325 × 105 Pa).

2.1.1.3 Permeation Model

The permeation rate of H2 at the heat exchanger, R HX , H 2 [m3 (STP)/s], and at the co-axial pipe,
Rcopipe,H 2 [m3 (STP)/s], is estimated using the following equation:

RHX , H 2, j or Rcopipe, H 2, j 
A
l
˜ k p,H ˜  PH 2,h  PH 2,l  (2-21)

where
A = heat transfer area or surface area [m2]
l = thickness of heat transfer tube or component casing [m]
k p, H = permeability of hydrogen [m3 (STP)/m/s/Pa0.5]
PH 2, h = partial pressure of H2 at high pressure side [Pa]
PH 2,l = partial pressure of H2 at low pressure side [Pa]

l is calculated by the following equation:

§r ·
l ro ˜ ln¨ o ¸ (2-22)
© ri ¹

72
where
ro = outer radius of heat transfer tube [m]
ri = inner radius of heat transfer tube [m].

Permeability is calculated by using the Arrhenius equation:

§E·
kp f ˜ exp¨ ¸ (2-23)
© RT ¹

where
f = pre-exponential factor of permeability [m3 (STP)/m/s/Pa0.5]
E = activation energy [J/mol]
R = ideal gas constant [J/mol-K]
T = temperature [K].

The permeation rate of H2 though the outer wall of the component and piping, Rcomp, H 2 , is calculated
by excluding the H2 partial pressure at the low pressure side as follows:

A
Rcomp, H 2, j ˜ k p , H ˜ PH 2,h . (2-24)
l

The permeation rate of HT at the heat exchanger, R HX , H 2 [m3 (STP)/s], and at the co-axial pipe,
Rcopipe, H 2 [m3 (STP)/s], is estimated by considering the effect of the existence of hydrogen on the
adsorption-dissociation and recombination-desorption step as follows:

A § P ·§¨ PHT ,h PHT ,l ·

RHX , HT , j or Rcopipe, HT , j  ˜ k p ,T ˜ ¨¨1  H 2,l ¸
¸¨ P  ¸ (2-25)
l © PTotal ,l ¹© H 2,h  PHT ,h PH 2,l  PHT ,l ¸
¹
where
k p ,T = permeability of tritium [m3 (STP)/m/s/Pa0.5]
PHT ,h = partial pressure of HT at high pressure side [Pa]
PHT ,l = partial pressure of HT at low pressure side [Pa]
PTotal ,l = total pressure at low pressure side [Pa].

The permeation rate of HT through the outer wall of the component and piping, Rcomp, HT , is
calculated by using the following equation:

A PHT ,h
Rcomp, HT , j ˜ k p ,T ˜ . (2-26)
l PH 2,h  PHT ,h

2.1.1.4 Leak Model

The leak rate of helium from the loop, Rleak , He [m3 (STP)/s], can be expressed by the following
equation:

Rleak , He Vtotal ˜ LR (2-27)

73
where
Vtotal = total inventory in loop [m3 (STP)]
LR = helium leak rate [1/s].

The leak rate of H2, HT, and HTO with helium leakage in node j, Rleak ,i , j , is calculated by using the
following equation:

Vj
Rleak ,i , j Vtotal ˜ LR ˜ C i ˜ (i = H2, HT and HTO) (2-28)
Vleak

where
Ci = average concentration of chemical i in nodes with helium leak
Vleak = sum of the inventory of nodes with helium leak.

2.1.1.5 Purification System Model

The removal rate, R PF , of hydrogen and tritium-containing chemicals in the primary and
intermediate loop by the purification system is expressed by the following equation:

R PF ,i , j FPF , He ˜ K i ˜ C i , j (i = H2, HT, and HTO) (2-29)

where
FPF , He = helium flow rate at purification system [m3 (STP)/s]
Ki = fractional efficiency of purification system for removing component i.

2.1.2 Development of Dynamic Tritium Behavior Analysis Code by MATLAB

Simulink

2.1.2.1 MATLAB Simulink

In this work, a portion of the dynamic tritium behavior analysis code was developed by the MATLAB
Simulink software package. MATLAB is a high-level technical computing programming language
created by MathWorks. It provides easy matrix manipulation, algorithm s implementation, user interface
creation, and has numerous built-in numerical methods. Simulink is a commercial tool for modeling,
simulating, and analyzing multidomain dynamic systems. Its primary interface is a graphical block
diagramming tool and a customizable set of block libraries. It offers tight integration with the rest of the
MATLAB environment and drives MATLAB or is scripted from it. It supports linear and nonlinear
systems modeled in continuous time, sampled time, or a hybrid of both. Systems can also be multirate,
i.e., have different parts that are sampled or updated at different rates.
With Simulink, users can move beyond idealized linear models to explore more realistic nonlinear
models, factoring in friction, air resistance, gear slippage, hard stops, and other things that describe real-
world phenomena. For this reason, the use of MATLAB Simulink for the current work is quite adequate.
For modeling, Simulink provides a graphical user interface (GUI) for building models as block diagrams,
using click-and-drag mouse operations. With this interface, models can be drawn just as they are with
pencil and paper (or as most textbooks depict them). This leads to an easy and fast code development
process compared to the previous simulation packages that required researchers to formulate differential
equations and difference equations in a language or program. Simulink includes a comprehensive block

74
library of sinks, sources, linear and nonlinear components, and connectors. It can also be customized,
allowing users to creating their own blocks.
Models are hierarchical, so they can be built using both top-down and bottom-up approaches. The
system can be viewed at a high level, or easily gone through to see increasing levels of model detail. This
approach provides insight into how a model is organized and how its parts interact.

2.1.2.2 Basic Structure of the Dynamic Tritium Analysis Tool

The current dynamic tritium analysis code is basically composed of three different types of blocks:
Component Blocks, Model Blocks, and Supportive Blocks. The Component Blocks represents the actual
reactor or system components, which includes all the models and equations to solve the tritium behaviors
in the each component. It is composed of lots of Model Blocks that actually solve various models
including mass balance, tritium source, tritium permeation, purification system, and tritium leakage. The
Supportive blocks include the Detector Block and Data Storage Block, which provide data handling and
manipulation features. The overall categorization of these blocks is as follows:
x Model Blocks
- Mass Balance Model Block
- Core Model Block
 Tertiary Fission Block
 Birth from 6Li Block
 Birth from 7Li Block
 Birth from 3He Block
 Birth from 10B Block
- Permeation Model Block
- Leak Model Block
- Purification System Model Block
x Component Blocks
- Reactor Block
 Mass Balance Model Block
 Core Block
 Permeation Model Block
 Leak Model Block
 Purification System Block
- Pipe Block
 Mass Balance Model Block
 Permeation Model Block
 Leak Model Block
 Purification System Block
- Storage Vessel Block
 Mass Balance Model Block
 Permeation Model Block
 Leak Model Block

75
  Purification System Block
- Heat Exchanger Block
 Mass Balance Model Block
 Permeation Model Block
- Flow Splitter Block (no model blocks)
- Flow Mixer Block (no model blocks)
- Boundary Flow Block (no model blocks)
x Supportive Blocks
- Detector Block

2.1.2.3 Blocks Summaries

The basic blocks developed in this work for analysis of tritium behaviors in NGNP systems are
summarized in this section.
x Model Blocks
- Mass Balance Model Block

* Block Diagram

<Mass Balance Model Block>

Mass Balance Model Block solves mass balance equations (see Equation 2-6). This model block has
three inputs and two outputs.
Resource : species source [m3(STP)/s]
Flow rate IN : total inlet flow rate into the component [m3(STP)/s]
Concentration IN : species concentration of species i in inflow [m3(STP)/ m3(STP)]
Flow rate Out : total outlet flow rate into the component [m3(STP)/s]
Concentration IN : species concentration of species i in the component [m3(STP)/ m3(STP)].
To consider the effect of leak, purification and permeation sources, Leak Model, Purification Model
and Permeation Model Blocks are connected to this Block through the Resource port. The detail model
configuration in the block is illustrated in Figure 2-1.

76
Figure 2-1. Internal structure of the Mass Balance Model Block.

- Core Model Block

* Block Diagram

<Core Model Block>

The Core Model Block solves tritium source equations (see Equation (2-8) through (2-20)). It is
composed of five different sub-blocks based on the source type (see Figure 2-2). This model block has
only one output.
Output-1 : tritium sources from core [m3(STP)/s]
This block is linked to the reactor block by connecting the ports as shown below. Then, it gives
tritium sources to the reactor component.

Connecting Core Model Block

to the Reactor Block (example)

Figure 2-2 shows the internal block diagram of the Core Model Block. The Core Model Block is
composed of many sub-blocks that solve each tritium source equation. The core model parameters can be
modified by double-click the sub-block for each tritium source (see Figure 2-3). The detail structures of
the subcomponents are illustrated in Figures 2-4 through 2-8.

77
Figure 2-2. Internal block diagram of Core Model Block.

78
Figure 2-3. Parameter setting window for the tritium source from tertiary fission.

Figure 2-4. Submodel for tritium source from tertiary fission.

79
Figure 2-5. Submodel for tritium source from Birth from 6Li.

Figure 2-6. Submodel for tritium source from Birth from 7Li.

80
Figure 2-7. Submodel for tritium source from Birth from 7He.

Figure 2-8. Submodel for tritium source from Birth from 10B.

81
 - Permeation Model Block

* Block Diagram

<Permeation Model Block>

The Permeation Model Block solves the tritium and hydrogen penetration model (see Equations (21)
through (26)). This model block has two inputs, one switch, and one output.
Input-1 : tritium partial pressure [Pa]
Input-2 : hydrogen partial pressure [Pa]
Switch-1 (top) : activation of permeation model (true or false)
Output-1 : tritium or hydrogen permeation source [m3(STP)/s]

- Leak Model Block

* Block Diagram

<Leak Model Block>

The Leak Model Block solves tritium source or sink from leakage of coolant (see Equation (27),
(28)). This model block has two inputs, one switch, and one output.
Input-1 : average concentration of species [m3(STP)/ m3(STP)]
Input-2 : volume of the component [m3(STP)]
Switch-1 (top) : activation of leak model (true or false)
Output-1 : tritium or hydrogen leak rate [m3(STP)/s]

82
 - Purification System Model Block

* Block Diagram

<Purification System Model Block>

Purification System Model Block solves tritium removal rate by purification system (see Equation
(29)). This model block has one input, one switch, and one output.
Input-1 : average concentration of species [m3(STP)/ m3(STP)]
Switch-1 (top) : activation of purification model (true or false)
Output-1 : tritium or hydrogen purifying rate [m3(STP)/s]

x Component Blocks
Component Blocks represent the reactor system components in tritium behavior modeling. The
reactor system is modeled by a combination of these Component Blocks. The Component Blocks are
developed by assembling and modifying the basic Model Blocks. Details about the Component Blocks
are presented below.
- Reactor Block

* Block Diagram

<Reactor Block>

Reactor Block solves tritium behavior in the nuclear reactor. It consists of several model blocks. This
block contains mass balance model blocks, core model block, permeation model block, leak model block
and purification system model block. This component block has two inputs and two outputs.
Input-1 : tritium sources to core [m3(STP)/s]
Input-2 : in-flow variables (connected from upstream components)

83
 Output-1 : tritium permeation or leak rate to outside (connected to other components)
Output-2 : out-flow variables (connected to downstream components).
This component has lots of input parameters. By double-clicking the block icon, we can activate the
input window to modify them. The detail configuration of the Reactor Block is shown in Figure 2-9. (For
more information, please check the block inside.)

Figure 2-9. Internal block diagram of Reactor Block.

- Pipe Block

* Block Diagram

<Pipe Block>

The Pipe Block solves tritium behavior in the pipe. It consists of several model blocks including mass
balance model block, permeation model blocks, leak model blocks, and purification model blocks. This
component block has two inputs and two outputs.
Input-1 : tritium or hydrogen sources into the pipe [m3(STP)/s]
Input-2 : in-flow variables (connected to other components)
Output-1 : tritium permeation or leak rate to outside [m3(STP)/s]
Output-2 : out-flow variables (connected to other components)
This block has lots of input parameters. By double-clicking the block icon, we can activate the input
window to modify them.

84
 - Storage Vessel Block

* Block Diagram

<Vessel Block>

The Storage Vessel Block solves tritium behavior in the vessel. It contains several Model Blocks
including mass balance model blocks, permeation model blocks, leak model blocks and purification
model block. This component block has two inlets and two outlets.
Inlet-1 : tritium or hydrogen sources into the pipe [m3(STP)/s]
Inlet-2 : in-flow variables (connected to other components)
Outlet-1 : tritium permeation or leak rate to outside [m3(STP)/s]
Outlet-2 : out-flow variables (connected to other components)
This component has lots of input parameters. By double-clicking the block icon, we can activate the
input window to modify them.
- Heat Exchanger Block

* Block Diagram

<Heat Exchanger Block>

The Heat Exchanger Block solves tritium behavior in the heat exchanger. It consists of several model
blocks especially focused on tritium and hydrogen penetration through the wall. This component block
contains mass balance model blocks and permeation model blocks. This block has two inputs and two
outputs.
Inlet-1 : in-flow variables 1 (connected to other components)
Inlet-2 : in-flow variables 2 (connected to other components)
Outlet-1 : out-flow variables 1 (connected to other components)
Outlet-2 : out-flow variables 2 (connected to other components)

85
 This component has lots of input parameters. By double-clicking the block icon, we can activate the
input window to modify them. The detail configuration of the Reactor Block is shown in Figure 2-10.
(For more information, please check the block inside.)

Figure 2-10. Internal block diagram of the Heat Exchanger Block.

- Flow Splitter and Mixer Block

* Block Diagram

<Flow Splitter Block>

The Flow Splitter Block simply splits the flow into two subflows. It consists of one main stream input
and two downstream outputs.

* Block Diagram

86
 <Flow mixer Block>

The Flow Mixer Block simply combines the two flows into one. It consists of two main stream inputs
and one downstream output.

- Boundary Block

* Block Diagram

< Boundary Block >

Boundary Block contains flow information. It is linked to other component blocks as a flow input.

x Supportive Blocks
Supportive Blocks handle and store the simulation data. This block can be linked to any components
and linkage lines. It then captures the data from the linked components and export it to the files or
graphical plots.

- Detector Block

* Block Diagram

< Detector Block >

The Detector Block contains displays and plots flow variables. By double-clicking the block icon, the
data on the flows or components can be accessed.

2.1.2.4 GUI and Snapshot of the Tritium Behavior Analysis Tool

Figure 2-11 shows the snapshot and user interface of the tritium analysis code. The upper part of the
window contains all the system component blocks, which can be dragged and dropped to the work
windows (the bottom window). By double-clicking the component icons, the component parameters and
input data can be set up. This simple and easy interface is expected to save lots of time for developing
input files, which were previously a text based job, consuming much time and efforts.

87
Figure 2-11. GUI and workspace of the current tritium behavior analysis code.

2.2 Follow-up of THYTAN Verification Study

This work paralleled the development of INL tritium behavior analysis code described in the previous
section by performing some verification works of THYTAN code developed by Japan Atomic Energy
Agency and used by Ohashi and Sherman (2007). In order to understand THYTAN code and the Tritium
birth mechanism in the HTGR core, we followed up by using the same process used by Ohashi and
Sherman (2006) for the Peach Bottom reactor (Scheffel, Baldwin, and Tomlin 1976; Wichner and Dyer
1979).
Figure 2-12 shows a schematic of the Peach Bottom primary loop. The primary circuit consisted
of two loops, each containing a helium compressor and steam generator. The total helium flow of
210,000 kg/h was divided equally between the two loops. Coolant temperatures at the core inlet and outlet
of the rector vessel were 345 and 714°C, respectively, and the primary loop pressure was approximately
2.4 MPa (335 psig). The reactor and steam generator were connected by a concentric duct. Coolant
exiting the reactor flows in the inner pipe of the concentric duct. The steam generators were forced
recirculation, drum type boilers that have pendant U-tube economizer, evaporator, and superheater
sections.
The tritium birth of the Peach Bottom HTGR at the Core 2 operation was evaluated by Wichner and
Dyer (1979), and their data was used by Ohashi and Sherman (1979) to verify THYTAN code. Our
calculation results about tritium birth in the reactor core were compared to the data reported by them.
Finally, the THYTAN input for the Peach Bottom core was used as basic data for the core input designed
for the VHTR with 600 MWt.

88
 Figure 2-12. Primary coolant system of the Peach Bottom reactor (Wichner and Dyer
1979).

2.2.1 Verification of Core Model

Four main tritium sources were taken into consideration in our effort to verify THYTAN code.
x Birth of tritium by ternary fission
x Birth of tritium from lithium
x Birth of tritium from 3He in the coolant
x Birth of tritium from 10B.

2.2.1.1 Birth from Ternary Fission

The following analytical solution was used to estimate the tritium birth in fuel by ternary fission.

§ 1  exp O ˜ t f ·
N T t f  K ˜ P ˜ Y ¨¨ ¸¸
© O ¹ (2-30)
where
N T t f  = atoms of tritium at time t f
K = fission rate per thermal megawatt (3.121 × 1016 fissions/MW/s)
Y = average yield per fission (1 × 10-4)
O = tritium decay constant (1.793 × 10-19 s-1).

An average power, P was calculated using the following equation.

P Prated EFPD / t f  (2-31)

89
where
Prated = rated power (115 MWt)
EFP = equivalent full power days at EOL (897 days)
tf = duration of Core 2 operation, including shutdown (1.34 × 108 s or 1550 days).

Table 2-1 summarizes the computed and reported results. The reported tritium birth by ternary fission
during 1,550 days was 1,210 Ci (= 4.43e13 Bq). The calculated value by THYTAN shows very good
agreement with the reported and analytical values.
Table 2-1. Comparison of tritium activity by ternary fission.
Activity
Computed Solution from
Reported Value Analytical THYTAN (Ohashi and Computed Solution
(Wichner and Dyer 1979) Solution Sherman 2007) by Current Work
(Ci) (Bq) (Bq) (Bq) (Bq)
13 13
1,210 4.43 × 10 4.43 × 10 4.42 × 1013 4.42 × 1013

2.2.1.2 Birth from Lithium

The birth of tritium from Lithium was estimated by the following analytical solution:

§ I th ˜ V Li 6T ˜ N 6 0 ·
N T t f  ¨¨ ¸¸ ˜ >exp I th ˜ V Li 6T ˜ t f   exp O ˜ t f @
© O  I th ˜ V Li 6T ¹ (2-32)
where
N T t f = atoms of tritium from 6Li at time t f
V Li 6T = effective cross section for 6Li (n, Į) T (4.08 × 10-22 cm2)
N 6 0 = initial amount of 6Li atoms.

Table 2-2 summarizes the reported and computed tritium activity. The parameters for the calculation
of each graphite components are listed in Tables 2-3 and 2-4 (Ohashi and Sherman 2007). The radial
reflector was not replaced at the end of the Core 1 operation. Therefore, the tritium activities in the
removal radial reflector and the permanent radial reflector are calculated sequentially.
As reported by Ohashi and Sherman (2007), there were some unknown discrepancies between
reported values and the THYTAN calculation that have not yet been revealed. Still, the computed solution
by THYTAN showed good agreement with the analytical solution for each component. Our THYTAN
solution shows very slight discrepancies with the values reported by Ohashi and Sherman because of the
different time step size. However, the discrepancies were still within 1% of the analytical solution.

90
Table 2-2. Comparison of tritium activity from 6Li at the Core 2 operation of the Peach Bottom reactor.
Activity
THYTAN
Analytical (Ohashi and THYTAN
Graphite Reported Value Solution Sherman 2007) (Current Work)
Component (Ci) (Bq) (Bq) (Bq) (Bq)
11 11 11
Sleeve 14.0 5.13 × 10 5.12 × 10 5.12 × 10 5.10 × 1011
Spine 1.0 3.66 × 1010 3.78 × 1010 3.78 × 1010 3.77 × 1011
Removal radial
16.4 6.01 × 1011 5.76 × 1011 5.75 × 1011 5.74 × 1011
reflector
Permanent radial
18.8 6.89 × 1011 6.72 × 1011 6.69 × 1011 6.68 × 1011
reflector
Axial reflector 9.2 3.37 × 1011 3.42 × 1011 3.42 × 1011 3.41 × 1011
Fuel matrix 13.1 4.80 × 1011 5.68 × 1011 5.68 × 1011 5.67 × 1011

Table 2-3. Parameters for the calculation of tritium birth from 6Li (Ohashi and Sherman 2007).
Thermal Neutron Flux Initial Amount of 6Li
2
Component (neutrons/cm /s) (moles)
13
Sleeve 2.82 × 10 6.99 × 10-4
Spine 2.82 × 1013 5.17 × 10-5
Core 1 : 2.57 × 1013
Removal radial reflector 7.56 × 10-4
Core 2 : 2.82 × 1013
Core 1 : 1.28 × 1013
Permanent radial reflector 1.12 × 10-3
Core 2 : 1.41 × 1013
Axial reflector 1.41 × 1013 6.74 × 10-4
Fuel matrix 2.82 × 1013 7.76 × 10-4

Table 2-4. Input data of THYTAN for calculation of tritium birth from 6Li (Ohashi and Sherman 2007).
Lithium Concentration Graphite Weight
Component (ppm) (kg)
Sleeve 0.007 9.37 × 103
Spine 0.001 4.85 × 103
Removal radial reflector 0.007 1.01 × 104
Permanent radial reflector 0.007 1.50 × 104
Axial reflector 0.007 9.03 × 103
Fuel matrix 0.010 7.28 × 103

91
2.2.1.3 Birth from 3He in the Coolant
The birth of tritium from 3He was estimated by the following analytical equations (Wichner and Dyer
1979).

N T1
V1  V7  ˜ Ith ˜ V He3T ˜ N3
Q  O ˜V (2-33)
for the sleeve graphite

V3 ª1  exp O ˜ t f  º
N T 2 t f  ˜ I th ˜ V He3T ˜ N 3 « »
V ¬ O ¼ (2-34)
for the removal radial reflector

V4 ª1  exp O ˜ t f  º
N T 3 t f  ˜ I th ˜ V He3T ˜ N 3 « »
V ¬ O ¼ (2-35)
for the permanent radial reflector

V6 ª1  exp O ˜ t f  º
N T 4 t f  ˜ I th,3 ˜ V He3T ˜ N 3 « »
V ¬ O ¼ (2-36)
for the axial reflector

V5 ª1  exp O ˜ t f  º
N T 5 t f  ˜ I th, 2 ˜ V He3T ˜ N 3 « »
V ¬ O ¼ (2-37)
where
N T1 = total moles of tritium circulating in the reactor with the coolant
NT 2 = total moles of tritium born in the sleeve graphite from 3He
NT 3 = total moles of tritium born in the removal radial reflector from 3He
NT 4 = total moles of tritium born in the permanent reflector from 3He
NT 5 = total moles of tritium born in the axial reflector from 3He
V1 = volume of coolant passage in core (1.77 × 106 cm3)
V2 = ex-core primary system volume (1.88 × 108 cm3)
V3 = connected porosity in sleeve graphite (3.44 × 105 cm3)
V4 = connected porosity in removal radial reflector (8.46 × 105 cm3)
V5 = connected porosity in axial reflector (3.92 × 105 cm3)
V6 = connected porosity in permanent radial reflector (2.32 × 106 cm3)
V7 = purge flow volume within the fuel elements (6.32 × 105 cm3)
V = effective helium volume of the primary system
( V1 V 2V3  V4  V5  V6 [1.94 × 108 cm3])

92
 I th = average thermal neutron flux in core and removal radial reflector,
Core 2 (2.82 × 1013 neutrons/cm2/s)
I th, 2 = average thermal neutron flux in axial reflector (1.41 × 1013 neutrons/cm2/s)
I th,3 = average thermal neutron flux in permanent radial reflector (1.41 × 1013 neutrons/cm2/s)
V He3T = effective cross section for 3He (n, p) T (2.28 × 10-21 cm2)
Q = flow to chemical cleanup system plus 10% of fuel element purge flow
( Q1 (2.40 × 104 cm3/s)) + leakage flow rate from primary system ( Q2 [25.5 cm3/s]).

The total moles of 3He in the primary system, N 3 , is governed by the relation in the following
equation:

dN 3 ­§ V  V3  V 4  V7 · § V5  V 6 · ½
V He3T ˜ N 3 ®¨ 1 ¸ ˜ I th  ¨ ¸ ˜ I th, 2 ¾  Q2 [ 3 He]i [ 3 He]
dt ¯© V ¹ © V ¹ ¿ (2-38)
where

> 3 He@i = 3
He concentration in makeup helium (5.78 × 10-11 moles/cm3 based on 0.16 ppm 3He in
helium)

> 3 He@ = 3
He concentration in primary system (= N 3 / V ).

Table 2-5 summarizes the calculated tritium birth from 3He in the Core 2 operation in the Peach
Bottom reactor. The input parameters were summarized in Tables 2-6 and 2-7. Ohashi and Sherman
(2007) reported that the values are larger than both analytical and computed solutions within 10–60%.
However, the reason is still unknown. On the other hand, the computed results show good agreement with
analytical solutions.
Table 2-5. Comparison of tritium activity from 3He at the Core 2 operation of the Peach Bottom HTGR.
Activity
THYTAN
Analytical (Ohashi and THYTAN
Reported Value Solution Sherman 2007) (Current)
Region (Ci) (Bq) (Bq) (Bq) (Bq)
11 11 11
In sleeve graphite 5.4 1.98 × 10 1.31 × 10 1.26 × 10 1.26 × 1011
In removal radial reflector 13.5 4.95 × 1011 3.20 × 1011 3.09 × 1011 3.09 × 1011
In permanent radial
15.5 5.68 × 1011 5.43 × 1011 5.24 × 1011 5.24 × 1011
reflector
In axial reflector 3.1 1.14 × 1012 9.17 × 1011 8.86 × 1011 8.86 × 1011

Table 2-6. Input data of THYTAN for calculation of tritium birth from 3He (Ohashi and Sherman 2007).
Parameter Unit Value
3 -21
Effective cross section for He (n, p) T cm 2.28 × 10
3
He concentration in makeup helium ppm 0.16
Helium inventory in primary system kg 269

93
Table 2-7. Input data for the subnode of THYTAN for the calculation of tritium birth from 3He (Ohashi
and Sherman 2007).

Thermal Neutron Flux Helium Inventory

Subnode (neutrons/cm2/s) (kg)
In circulating coolant 2.82 × 1013 3.33
In sleeve graphite 2.82 × 1013 0.477
In removal radial reflector 2.82 × 1013 1.17
In permanent radial reflector 1.41 × 1013 3.22
In axial reflector 1.41 × 1013 0.543

2.2.1.4 Birth from 10B

The tritium birth from 10B was evaluated by the following equation:

NT t f  b˜a
exp a ˜ t f   exp O ˜ t f 
N10 0  b  a  ˜ O  a 
b˜a
 exp b ˜ t f   exp O ˜ t f 
b  a  ˜ O  b 
c
 exp a ˜ t f   exp O ˜ t f 
O  a  (2-39)
where
a = E Ith VB10Li7
b = Ig VLi7T
c = Ig VB10T
E = self-shielding factor (0.0141)
I th = average thermal neutron flux (2.82 × 1013 neutrons/cm2/s)
If = average fast neutron flux (2.26 × 1013 neutrons/cm2/s)
V B10Li 7 = effective cross section for 10B (n,Į) 7Li (1.63 × 10-21 cm2)
V Li7T = effective cross section for 7Li (n, nĮ) 3H (1.53 × 10-25 cm2)
V B10T = effective cross section for 10B (n, 2Į) 3H (5.00 × 10-26 cm2)
N10 0  = initial amount of 10B atoms (20.15 moles).

Tritium from 10B is produced in the poisoned spine and in the control rod. However, since the boron
level and control rod position varies with time in the control rod, only poisoned spine was taken into
consideration here. Table 2-8 shows the calculation result and reported values. Computed values show
good agreement with the analytical solutions.

94
Table 2-8. Comparison of tritium activity from 10B in the poisoned spine at the Core 2 operation of Peach
Bottom reactor.
Activity
THYTAN
Analytical (Ohashi and THYTAN
Reported Value Solution Sherman 2007) (Current Work)
(Ci) (Bq) (Bq) (Bq) (Bq)
12 12 12
85.7 3.14 × 10 3.19 × 10 3.18 × 10 3.18 × 1012

2.2.2 Verification of Permeation Model

The tritium permeation rate of the heat transfer tubing obtained from the Peach Bottom HTGR steam
generator at the end of life was reported by Yang, Baugh, and Baldwin (1977). The reported permeation
rate, J, of the Incoloy 800 super-heater tubing, steel evaporator tubing, and low-carbon steel economizer
are as follows:
for the economizer

J 7.78 u 101 ˜ exp 4500 / T  (493 K—623 K) [ȝCi/m2/h] (2-40)

for the evaporator

J 6.97 u 10 3 ˜ exp 6830 / T  (573 K—693 K) [ȝCi/m2/h] (2-41)

for the superheater

J 1.72 u 10 3 ˜ exp 6440 / T  (673 K—973 K) [ȝCi/m2/h]. (2-42)

Calculation of the tritium permeation rate for the superheater, evaporator, and economizer was carried
out by THYTAN using both permeation rates and compared to the reported correlations. Table 2-9 shows
the size and dimension of the heat exchangers. The experimental conditions—tritium concentration in the
feed side of 0.461 ppb, feed helium pressure of 0.1013 × 105 Pa, and the permeation chamber pressure of
1.33 × 10-3 Pa—were employed as boundary conditions. The effect of impurity hydrogen was ignored.
The hydrogen concentration of 199 ppm was assumed for the boundary condition for the primary side.
Table 2-9. The heat transfer tube dimensions of the steam generator in the Peach Bottom reactor (Ohashi
and Sherman 2007).
Effective Area Outer Radius Inner Radius
Sample (m2) (m) (m)
-4 -3
Economizer (inlet) 103.47 × 10 9.525 × 10 7.300 × 10-3
Evaporator (inlet) 68.34 × 10-4 6.350 × 10-3 4.775 × 10-3
Superheater (inlet) 90.16 × 10-4 9.525 × 10-3 6.350 × 10-3

Table 2-10 summarizes the analytical solutions of Equations (2-40) through (2-42) and the computed
results. As shown in this table, the computed solution is in good agreement with the analytical solution.

95
Table 2-10. Comparison of permeation rate.
Computed Solution from THYTAN
Ohashi and Sherman
Analytical Solutions (2007) Current Work
Temp. Permeation Rate Permeation rate Permeation rate
(K) (ȝCi/m2/h) (ȝCi/m2/h) (ȝCi/m2/h)
673 1.202 × 10-1 1.202 × 10-1 1.202 × 10-1
Superheater
(inlet) 823 6.873 × 10-1 6.876 × 10-1 6.876 × 10-1
973 2.298 2.297 2.297
573 4.640 × 10-2 4.644 × 10 -2
4.644 × 10-2
Evaporator
(inlet) 633 1.436 × 10-1 1.437 × 10-1 1.437 × 10-1
693 3.656 × 10-1 3.658 × 10-1 3.658 × 10-1
493 8.450 × 10-3 8.454 × 10-3 8.454 × 10-3
Economizer
(inlet) 558 2.447 × 10-2 2.448 × 10-2 2.448 × 10-2
623 5.676 × 10-2 5.678 × 10-2 5.678 × 10-2

2.2.3 Verification of Leak Model

In order to verify the leak model of THYTAN, the tritium concentration in the containment vessel
was estimated by THYTAN, using the reported leak rate and the tritium concentration in the primary
loop as Ohashi and Sherman (2007) did. The numerical analysis result from THYTAN on the tritium
concentration in the containment vessel was compared with the experimental result and the reported
analytical solution. Figure 2-13 shows the calculation of leakage by THYTAN code.
0.0025

In primary coolant (by THYTAN)

Trtitium Concentration [ppb]

0.0020

0.0015 In containment
(analytic solution)

0.0010 In containment by THYTAN

In containment
(Sherman and Ohashi (2007))
(experimental data)

0.0005

0.0000
0 50 100 150 200
Time (days) In containment by THYTAN
(Sherman and Ohashi (2007))

Figure 2-13. Comparison of tritium concentrations in the containment

vessel of the Peach Bottom reactor.

96
 The input data of THYTAN is summarized in Table 2-11. Because information on the containment
vessel temperature is not available, the average containment vessel temperature is assumed to be 323 K.
The tritium concentration in the primary loop of 2 × 10 5 ȝCi/cm3 corresponds to 1.98 × 10 3 ppb under
the reported average temperature of 809 K and a pressure of 23 atm in the primary coolant. In order to
keep the tritium concentration in the primary coolant 1.98 × 10 3 ppb, the tritium release rate to the
primary coolant was adjusted by the core model. The initial concentration of tritium in the containment
vessel was set at 0.
As a result, the computed tritium concentration by THYTAN code showed good agreement with
analytical solutions. However, the computed solutions were less than the reported experimental data
observed in 1971 during the special tritium survey.
Table 2-11. Input data of THYTAN for verification of the leak model (Ohashi and Sherman 2007).
Parameter Unit Value
Leak rate from primary to containment vessel 1/h 4.15 × 10-4 a
Primary loop volume m3 232
Primary loop pressure Pa 2.33 × 106 b
Primary loop temperature K 809
Leak rate from containment vessel to atmosphere 1/h 4.17 × 10-4 c
Containment vessel volume m3 1.56 × 104 d
Containment vessel pressure Pa 1.57 × 105 e
Containment vessel temperature K 323 f

a. Based on 3.2 kg/day.

b. Based on 23 atm.
c. Based on 0.1%/day.
d. Based on a nitrogen volume of 2.04 × 104 m3 at a pressure of 1.57 × 105 Pa and assumed temperature of 323 K.
e. Based on 8 psig.
f. Assumed.

2.2.4 Verification of Purification System Model

The purification system model was verified by the following simple analytical equation:
dCi , j
Vj  RPF ,i , j
dt
 FPF ,He ˜Ki ˜ Ci , j
. (2-43)

Ci, j
The solution for is found in the following equation:

§ FPF , He ˜K i ·
Ci , j t  C 0,i , j ˜ exp¨¨  ˜ t ¸¸
© Vj ¹ (2-44)
where
C0,i , j = initial concentration of chemical i in node j [m3 (STP)/m3 (STP)].

Table 2-13 summarizes the boundary conditions for verification of the purification system model. The
HT concentration was calculated during 3,000 sec.

97
Table 2-12. THYTAN boundary conditions for verification of the purification system model (Ohashi and
Sherman 2007).
Parameter Symbol Unit Value
3
Inventory Vj m (STP) 1
Helium flow rate at purification FPF , He m3 (STP) 1.0 × 10-3
system
Fractional efficiency of purification K HT — 0.9
system for HT
Initial concentration of HT C0, HT , j m3 (STP)/m3 (STP) 1.0 × 10-9

Figure 2-14 shows the comparisons between analytical solution and computational results by
THYTAN. As shown in this figure, both the analytical solution and the THYTAN results are in good
agreement.
1.20E-009
Analytical Solution
THYTAN solution (Ohashi and Sherman, 2007)
1.00E-009
THYTAN solution (Current Work)

8.00E-010
HT concentration

6.00E-010

4.00E-010

2.00E-010

0.00E+000

0 500 1000 1500 2000 2500 3000 3500

Time (s)

Figure 2-14. Comparisons between analytical solution and THYTAN results

for verification of the purification system.

2.3 Development of THYTAN Core Input for Transient Analysis of

VHTR Tritium Behaviors
For simplicity, Ohashi and Sherman’s (2006) previous tritium analysis of NGNP systems assumed
that the tritium generation from the reactor core was to be constant based on the average tritium
generation data from other HTGRs. However, because impurities and system conditions are continuously
changing during plant operation, the tritium generation should be considered time variant functions for
dynamic simulations. Lots of additional data are required for dynamic simulation, including tritium
source terms, component sizes, and details of plant operating history and conditions. In this fiscal year,
we developed a THYTAN VHTR core input file for later transient simulation by upgrading Ohashi and
Sherman’s steady-state input.

98
 As previously described, tritium is produced in the VHTR by various sources such as ternary fissions
and activation reactions with impurities and boron in the materials. The helium coolant itself is also a
tritium source in the form of neutron absorbing nuclide 3He with its extremely low isotope abundance.
Table 2-13 summarizes the tritium production reaction and cross section (Wichner 1979).
Table 2-13. Tritium production reaction and cross section (Wichner 1979).
Production Reaction Cross section (barns) Energy range (eV)
3
He(n,p)T 2,280 0–2.38
6
Li(n,a)T 408 0–2.38
10
B(n,2a)T 0.050 >0.18u106
10
B(n,a)7Li 1,630 0–2.38
7
Li(n,n’a)T 0.153 >0.18u106

Compared with these sources, the contributions of other tritium producing reactions with nuclides
such as 9Be or 12C are negligible (Steinwarz et al. 1984).

2.3.1 Input Parameters for Tritium Source from Ternary Fission

If we assume an average yield (Y) of tritium atoms per fission, the tritium production rate from
fission can be expressed as

dN T (t )
KYP(t )  ON T (t ) (2-45)
dt
where
N T (t ) = atoms of tritium at time t
K = fission rate per thermal megawatt (=3.121X1016 fissions/sec MWt)
P (t ) = power at time t (MWt)
O = decay constant (1.793X10-9 sec-1)
Y = average yield per fission (1X10-4).
Table 2-14 summarizes the necessary information for tritium source of ternary fission.
Table 2-14. Basic data for THYTAN evaluation of ternary fission.
Unit Values
Reactor Power (P) MWt 600
Fission Rate (K) Fissions/MW sec 3.12e16
Average Yield Per Fission (Y) 1/fission 1.0e-4
Failure Fraction of Fuel Particles - 0.3

2.3.2 Input Parameters for Tritium Source from 3He in the Coolant
Tritium is produced from 3He via an (n,P) reaction with thermal neutrons. The level of 3He
contamination in commercially available sources varies, but in general, helium from natural gas wells
contains approximately 0.2 ppm, while the 3He level in the atmospheric is about 10 times higher than that
(Steinwarz et al. 1980). Hence, this report assumes the 3He supply level to be 0.2 ppm.

99
 Generally, the He coolant suffers from leakage during routine operations. In Dragon, the leak rate
reached 2.0 kg/day at the beginning of 1974 and after months of research a number of leaks were found in
the stainless steel pipe work leading to the helium purification plant (Simon et. al. 1980). To compensate
for helium leaks, fresh helium should be continuously added to the primary circuit. In this work, the
tritium leakage rate and total helium inventory were assumed to be 0.01%/day (1.57e-9 sec-1) and
4,535 kg, respectively (Yook 2007).
The inventory of the primary system includes the coolant passages between the fuel elements (V1),
and the piping, plenums, and steam-generator tubing forming the ex-core primary system volume (V2). In
addition, the connected porosity in portions of the graphite core components is readily accessible to 3He
permeation (V2, V3, V5, V6 and V7). Therefore, the effective He volume of the primary system includes

V V1  V2  V3  V4  V5  V6 (2-46)

where
V1 = Volume of coolant passages in core
V2 = ex-core primary system volume
V3 = connected porosity in sleeve graphite
V4 = connected porosity in removable radial reflector
V5 = connected porosity in axial reflector
V6 = connected porosity in permanent radial reflector

Tables 2-15 and 2-16 summarize the parameters used for estimating tritium production from 3He in
this work (Yook 2007).
Table 2-15. Basic data for THYTAN evaluation of 3He Source.
Unit Values
Primary Helium Inventory kg 4,536
Helium Supply Rate 1/sec 1.57e-9
3
He Concentration in Supply Helium ppm 0.2

Table 2-16. Parameters used for estimating tritium production from 3He.
Inventory Thermal Neutron Flux Ratio of release
(kg) (n/cm2 sec) to birth rate
Coolant Passage in the Core (V1) 109 3.73e13 1.0
Ex-core Loop (V2) 4370 0 1.0
Central Graphite Passage (V3) 0 3.73e13 1.0
Removal Passage (V4) 21 3.73e13 1.0
Axial Passage (V5) 39 3.73e13 1.0
Permanent Passage (V6) 0 3.73e13 1.0
Purge Flow Passage (V7) 0 3.73e13 1.0

Tritons from the 3He(n,p)T reaction in the primary coolant will come to rest in solids as fractions,
which vary from region to region depending on the space associated with channels and fabricated holes,
clearance annuli, and pores (GA 2006).

100
 The recoil energy of a triton is about 0.2 MeV leading to an estimated range of 0.05 cm in helium at
47.6 atm and 1,000¶C. In the case of pores, because they are generally smaller than 0.05 cm, the fraction
bound is taken as unity (GA 2006). A planar approximation is used for the clearance annuli (tolerances
between graphite blocks or between the fuel rods and the graphite blocks containing them). If the width of
the gap exceeds the recoil range,

Fraction Bound range /(2 ˜ width) , range  width , (2-47)

If the width of the gap is less than the recoil range,

Fraction Bound (range  width / 2) / range , range ! width (2-48)

For large cylindrical channels or holes,

Fraction Bound range / channel diameter (2-49)

The fraction bound for this work was determined to be 0.025 by Equation (2-49); therefore, the
release rate is 0.975.

2.3.3 Input Parameters for Tritium Source from Li in Graphite

The necessary parameter for Li source in graphite are summarized in Table 2-17. Generally, the Li
level in graphite is too low to determine an appropriate average concentration for large graphite mass
from the perspective of analytical chemistry and the sampling procedure. Furthermore, the Li content in
graphite varies with the production region and company. For example, AVR has the core and reflector
graphite with 0.5 ppm and 0.1 ppm in Li level, respectively (Steinwartz 1980), while High Temperature
Reactor-10 Megawatt (HTR-10) assumes 10 ppb (Steinwartz 1984). The Li concentration is around 5 ppb
in ATR-2E block graphite and less than 0.8 ppb in spherical graphite fuel elements (Kirch et al. 1980).
The Li concentrations for Peach Bottom were measured in the selected samples of Peach Bottom graphite,
and the average of these values yields 7.0 ppb. In this report, the value 7.0 ppb was assumed.
The release rate of tritium, which is bound interstitially in graphite, is expected to be quite slow
because a temperature of at least 1,200¶C is required for tritium desorption from graphite. A retention
fraction of 0.99 was used for this work, based on the recommendation of TRITGO (computer model)
manual (GA 2006).
The weight of the graphite was determined by the volume of the region multiplied by graphite density
(1.74 g/cm3) (Park et al. 2007).
Table 2-17. Parameters used for estimating tritium production from Li.
Impurity Weight of Thermal Fast Neutron
Concentration Graphite Neutron Flux Flux
(ppm) (kg) (n/cm2 sec) (n/cm2 sec)
Inner Core 0.007 44452 3.73e13 7.02e13
Outer Core 0.007 45359 3.73e13 7.02e13
Top Reflector 0.007 23768 3.73e13 3.68e13
Bottom Reflector 0.007 31706 3.73e13 3.68e13
Inner Reflector 0.007 128366 3.73e13 3.68e13
Side Reflector 0.007 417359 3.73e13 3.68e13

101
2.3.4 Input Parameters for Tritium Source from 10B in Graphite
Tritium formed from the boron in the control rods is not easily estimated because of the varying
quantity in the active core. In addition, the mass of boron in the control rods is not specified in any
reference and can be surmised only from the amount of excess reactivity they control.
According to Peach Bottom reactor data (Wicher et al. 1979), tritium generated in the control rods are
the main source of tritium generation in the neutron reaction. But, the most of tritium in the control rods
remains within the generated site (Yook 2007).
The input of the tritium source from 10B was developed based on the TRITGO input for VHTR by
Park et al. (2007).
Table 2-18. Parameters used for estimating tritium production from Li.
Boron Regional Thermal Fast Neutron
Concentration Weight Neutron Flux Flux
(ppm) (kg) (n/cm2 sec) (n/cm2 sec)
Inner Core 41 44,452 3.73e13 7.02e13
Outer Core 41 45,359 3.73e13 7.02e13
Top Reflector 2 23,768 3.73e13 3.68e13
Bottom Reflector 2 31,706 3.73e13 3.68e13
Inner Reflector 2 128,366 3.73e13 3.68e13
Side Reflector 2 417,359 3.73e13 3.68e13
Control Rod 1e6 251 1.20e13 1.20e13

2.4 Evaluation of the Effective Thickness for

Tritium Permeation in PCHEs
In the VHTR system, heat exchangers are the main route of tritium permeation. Therefore, when we
estimate the tritium distributions and contamination levels in hydrogen from the hydrogen production
plant, it is very important to predict accurate tritium penetration rates. Generally, the tritium penetration
rate (NHT) is determined by the following equation:

N HT
A
l
 n
˜ k ˜ PHT ,h  PHT ,l
n

(2-50)
where
A = heat transfer area or surface area [m2]
l = thickness of heat transfer tube or component casing [m]
k = permeability of HT [m3 (STP)/m/s/Pa0.5]
PHT ,h = partial pressure of HT at high pressure side [Pa]
PHT ,l = partial pressure of HT at low pressure side [Pa]
N = order of permeation.
Applying this equation to shell-and-tube type heat exchangers is quite easy and straight forward.
However, when applying it to PCHEs, we meet some difficulties for determining the thickness of heat

102
transfer tubes because of the unique shape and geometry of these channels. Tritium penetration rates vary
significantly along the channel surface as shown in the previous chapter. For this reason, the effective
thickness of the tritium penetration path in PCHEs was estimated in this section.
When using geometries other than the typical shapes such as circles and plates, it is often useful to
calculate an effective thickness for a model. This thickness represents the thickness of a plate that would
have the same flux rate as the model itself. In order to estimate the effective thickness of PCHEs,
Equation (2-50) was modified to the following diffusion equations by substituting PHTn with XHT.

k
N HT ( X HT ,h  X HT ,l ) (2-51)
t eff .
where
t eff . = effective thickness for tritium permeation [m]
n
X HT = PHT

To determine the NHT, COMSOL software and modeling used in Section 1.4.1 were also used in here.
The boundary conditions of XHT and the values of k were also determined based on this model.
Theoretically, the effective thickness is independent of diffusivity, concentration, flux, temperature,
pressure, and flow rates. The effective thickness can be obtained by the following equation:

k
t eff . ( X HT ,h  X HT ,l ) (2-52)
N HT

In the above equation, k and XHT are fixed values during the calculations for a given temperature,
pressure and boundary tritium concentration. However, NHT and A are variables related to the channel
dimensions—diameter, plate thickness, and horizontal pitches—used in the modeling. Therefore, the
different effective thicknesses will be obtained for the different channel dimensions by Equation (2-52)
because of the different NTH values calculated by COMSOL.
Similar to Chapter 1.4.1, two channel configurations were taken into consideration: standard in-line
configuration and off-set configuration. The reference PCHE configuration had a plate thickness of
0.96 mm (tp= 0.96 mm) and a pitch of 1.464 mm (p = 1.464 mm). The diameter of the semicircular ports
was 1.2 mm (dt = 1.2 mm). Calculations were performed for five horizontal pitches (1.332 mm,
1.464 mm, 1.728 mm, 1.992 mm, and 2.5 mm), and five plate thicknesses (0.69 mm, 0.96 mm, 1.32 mm,
2.54 mm, and 3.48 mm).The horizontal pitch, plate thickness, and diameter are shown on Figure 2-15.
The offset model was made by shifting every other plate by half the model’s pitch.
Table 2-19 summarizes the effective thicknesses of the reference geometry for various temperatures.
As shown in the table, the effective thickness is not affected by the temperature, and it confirms that the
permeability does not affect the effective thickness. In this table, the effective thickness in the off-set
geometry is a little bit larger than in the standard geometry, which is because the diffusion path in the off-
set design is larger than the standard design. However, the differences are very slight. One important
finding in this result is that the effective thickness (0.49 mm) in the PCHE is much smaller than the
average thickness between two channels (0.66 mm), which is because most of the tritium has penetrated
through the shortest parts of the channel distance. Therefore, determination of the tube thickness by the
average channel distance may result in much less tritium penetration through this PCHE walls.

103
(a) Standard in-line configuration (b) Off-set configuration
Figure 2-15. PCHE channel configurations.

Table 2-19. Effective Thickness over a Temperature Range.

Effective Thickness Effective Thickness
for Mass Flux for Mass Flux
(m) (m)
Initial Offset
500°C 0.0004876 0.000491788
600°C 0.00048787 0.000492015
700°C 0.0004882 0.000492604
800°C 0.000488612 0.000493233
900°C 0.000489025 0.00049379

In order to correlate the effective thickness as a function of channel dimensions (channel diameter,
plate thickness, and horizontal pitch), the effective thicknesses have been obtained by Equation (2-52) and
COMSOL modeling of various channel dimensions. Figure 2-16 shows the effective thickness of the
PCHE for various pitch-to-diameter ratios in vertical and horizontal directions, which is increased with
vertical pitches, and decreased with horizontal pitches. However, the effect of the vertical pitch is more
significant.

104
 (a) Standard configuration.
3.5
plate thickness (tp)/ diameter (d)
3.0 tp/d = 0.58

effective thickness (teff) / diameter (d)

tp/d = 0.80
tp/d = 1.10
2.5 tp/d = 2.12
tp/d = 2.90
2.0

1.5

1.0

0.5

0.0
1.0 1.2 1.4 1.6 1.8 2.0 2.2
horizontal pitch (p) / diameter (d)

(b) Off-set configuration.

3.5
plate thickness (tp)/ diameter (d)
tp/d = 0.58
3.0
effective thickness (teff) / diameter (d)

tp/d = 0.80
tp/d = 1.10
2.5 tp/d = 2.12
tp/d = 2.90
2.0

1.5

1.0

0.5

1.0 1.2 1.4 1.6 1.8 2.0 2.2

horizontal pitch (p) / diameter (d)

Figure 2-16. Effective thickness for various channel dimensions

(diameter, pitch and thickness).

By regressing the data shown in Figure 2-18, the effective thickness correlations were developed for
standard and offset configurations. In order to generalize the correlations, all the geometric parameters
were normalized by the channel diameter, d. Therefore, the dimensionless effective thickness (teff/d) was
finally expressed by dimensionless plate thickness (tp/d) and dimensionless pitch (p/d). The correlations
for both standard and off-set configurations are described below.

105
Standard configuration
2.92 0.19
t eff . § tp · § p· tp
0.8 ˜ ¨¨ ¸¸ ˜¨ ¸ ( <0.8) (2-53a)
d ©d ¹ ©d ¹ d

1.51 0.61
t eff . § tp · § p· tp
0.67 ˜ ¨¨ ¸¸ ˜¨ ¸ ( >0.8) (2-53b)
d ©d ¹ ©d ¹ d

Off-set configuration
1.94 0.85
t eff . § tp · § p· tp
0.48 ˜ ¨¨ ¸¸ ˜¨ ¸ ( <0.8) (2-54a)
d ©d ¹ ©d¹ d

1.35 0.26
t eff . § tp · § p· tp
0.65 ˜ ¨¨ ¸¸ ˜¨ ¸ ( >0.8) (2-54b)
d ©d ¹ ©d ¹ d

where

d = channel diameter [m]

t p = channel vertical thickness [m]
p = channel horizontal pitch [m].

106
 3. SUMMARY
Two important issues associated with VHTR systems were investigated: heat exchangers and tritium
permeation. The tasks performed and results obtained in this study are summarized in this section.

3.1 Design Options for NGNP Heat Exchanger

3.1.1 Heat Exchanger Thermal Design and Design Options
The following heat exchanger options, applicable to VHTRs, were selected from three categories
(heat exchanger types, heat exchanger serial arrangement, and PCHE channel layouts) for consideration
in this report:
x Heat exchanger types: PCHE, shell-and-tube, and helical coil
x Heat exchanger arrangements: single- and two-stage
x PCHE channel layout configurations: standard in-line and off-set configurations
x In addition, cross-flow and counter-current flow were considered in the PCHE design, and straight
pipe and U-tube configurations were considered in the shell-and-tube heat exchanger design.
The reference reactor had 600 MWt, and two different system configurations were considered for the
thermal design. The following results were obtained from this research:
x The PCHE has much smaller size (or volume) than the tubular type heat exchangers (shell-and-tube
or helical coil). In high temperature applications, the smaller volume of heat exchanger is generally
preferred because (1) the high alloy metals or ceramics are very costly, and (2) the smaller heat
exchanger can reduce the total system size. The total core volumes for exchanging heat are 5–6 m3 for
the PCHE design, 400–600 m3 for the shell-and-tube design, and 160–190 m3 for the helical coil
design. The small size of the PCHEs contributes to the high surface area density and the high heat
transfer coefficient caused by the small channel size.
x The PCHE has much smaller heat transfer area compared to other options. The reduced heat transfer
area leads to less tritium penetration through the heat exchangers. For 600 MWt design, the heat
exchanger requires about 4,500–6,500 m2 for PCHEs, 28,000–32,000 m2 for shell-and-tube heat
exchangers, and 13,000 m3 for helical coil heat exchangers. When considering the effect of heat
exchanger wall thickness together, the PCHE is estimated to show about one-half less tritium
penetration than the shell-and-tube design.
x The PCHE shows better heat transfer performance than other types, but it requires much larger
friction loss at the same flow length. Therefore, for the same pressure drop, the PCHE requires too
short a flow length, which would lead to potential thermal stress problems because the temperature
gradient in the flow direction is too large.
x The PCHE requires a large number of small modules for large duty, which would lead to a very
complex system. The PCHE design also requires a complex manifold and flow connections between
modules.
x The shell-and-tube and helical coil designs are proven technologies, and the problems and limitation
of the heat exchangers are well identified. It is therefore more easily applied to the VHTR application
in the near term. The PCHE needs more research and validation for the VHTR application.
x Tubular heat exchangers allow in-service inspections and have well-established maintenance methods,
whereas defects and failures in PCHE during operations will be hard to find and maintenance will be
more difficult compared to the tubular design.

107
x A two-stage heat exchanger arrangement has been proposed in to reduce risk and cost for VHTR
applications. The two-stage concept splits the heat exchanger into two modules—a high temperature
unit and low temperature unit—at the separation temperature of 750°C. The high temperature unit is
designed for replacement within the plant’s lifetime while the low temperature unit is designed for
lifetime operation. Alloy 800H is a potential candidate for the low temperature unit; Alloys 617 and
230 are candidates for the high temperature unit. Conceptually, this is very good, but according
calculations, the high temperature unit requires almost two-thirds of the total heat exchanger size,
which means that the most of that unit should be replaced during the plant’s lifetime. This is
impractical when compared to the single-stage design.
x An 800°C separation temperature was considered in order to reduce the size of the high temperature
unit. The size of the LMTD was even decreased compared to the 760°C case, but the duty of the high
temperature unit was significantly reduced, resulting in a significant reduction in heat exchanger
capacity. In this design, the size of the high temperature unit was about one-half that of the total heat
exchanger size. However, in this scenario, the material for the low temperature unit is required to be
Alloy 617 or 230. We therefore will need some optimization study to determine the best separation
temperature.

3.1.2 Stress Analysis of the VHTR Heat Exchangers

Simple stress analyses on the PCHE and the tubular heat exchangers were performed. Based on these
analyses, the lifetimes of the reference design IHXs and SHXs have been estimated and the thickness
required for a certain lifetime has been calculated. For conservative analyses, Alloy 617 has been selected
as the structure material. The maximum pressure difference between hot and cold channels was 1.0 MPa
for the IHX, 2.0 MPa for the SHX, and 5.0 MPa for another SHX. The following results were obtained
from this research:
x The IHX lifetime of the reference design is about 35 years for the PCHE design, and 10 years for the
tubular design at the operating temperature of 900°C. The reason for the longer lifetime of the PCHE
design is because of the larger thickness-to-channel radius ratio (t/ri). The values of t/ri for the tubular
design with commercial pipes are less than 0.3, while the values for PCHE are usually more than 0.5.
x The lifetime of the PCHE for SHX with 'Pmax = 2.0 is close to the IHX, but for the shell-and-tube
design at 900°C drops to 1.5 years. This is because of the low t/ri values. It means that the PCHE is
theoretically more secure than the shell-and-tube design in the aspect of mechanical stress.
x The lifetime of the PCHE for SHX with 'Pmax = 5.0 is about 4,000 hours at 900°C. The reference
tubular design cannot be operated at this condition. Based on this analysis, the PCHE is also
inadequate for this application because 4,000 hours is still too short a period when compared to the
whole plant’s lifetime (about 50 years).
x The tube wall or plate thickness required for some VHTR operating conditions were analyzed in this
report. According to the results, the PCHE type is more suitable for VHTR applications than the
tubular design because the tubular design would require a large tube thickness to accommodate
commercially available pipes.

3.1.3 Computational Analysis of PCHE Channel Configuration Options

Two channel configurations of the PCHEs were investigated using Finite Element Methods (FEM)
focused on heat transfer, tritium diffusion, and stress. The two channel options considered are standard in-
line and off-set. COMSOL Multiphysics software was used to analyze heat transfer and tritium diffusion.
Various horizontal channel pitches and plate thicknesses were taken into consideration. ABAQUS ® (ver.
6.75) was used to perform stress analyses. The following results were obtained from this research:

108
x The effect of the standard off-set option on the heat transfer was very slight because the heat transfer
resistance in the heat exchanger is mainly on the fluid boundary layers.
x Off-setting the channels slightly decreased the rate of tritium diffusion because the main resistance of
the tritium diffusion is in the solid metal structure. However, the reduction is negligible.
x Increasing the plate thickness significantly reduces the heat transfer and diffusion rate, and increasing
the horizontal pitch slightly increases the diffusion rate. However, the effect of the horizontal pitch
was much smaller than the plate thickness. The tritium rate increases with the horizontal pitch
because the dead spot decreases on the channel side.
x The off-set design reduced stress concentration by a maximum of 50%, which is significant, since the
lifetime of the materials are exponentially decreased by the maximum stress. The 50% stress
reduction will lead to a significant increase in unit lifetime.

3.2 Tritium Behavior Study in the VHTR System

3.2.1 Heat Exchanger Thermal Design and Design Options
In the VHTR hydrogen system, tritium permeation from the core into the produced hydrogen is a
serious concern because the tritium can easily permeate high temperature heat exchanger tubes and
contaminate the product hydrogen. In this report, several separate works have been performed for analysis
of the tritium behaviors in the VHTR hydrogen system. The following results were obtained from this
research:
x A portion of dynamic tritium behavior analysis code was developed using a MATLAB Simulink
software package. In this code, (1) tritium generations by five different sources (Ternary fission and
neutron reactions of 6Li, 7Li, 10B and 3He), (2) tritium penetration through metals, (3) leakage of
coolant, and (4) purification system were considered using the mass balance equations of various
species (HT, H2, H2O, HTO, etc.). The code is based on the graphical user interface that allows users
to drag-and-drop when analyzing and estimating tritium behaviors in complex systems.
x In parallel with code development, the THYTAN code used by Ohashi and Sherman (2006)
was successfully verified by comparing the analytical solutions with Peach Bottom data. The code
inputs were also prepared for transient analysis of the VHTR hydrogen system.
x As the first step of the transient tritium behavior analysis in the VHTR hydrogen system, the
THYTAN tritium generation input file was developed. The input data were collected from various
sources.
x A FEM analysis was performed using COMSOL Multiphysics to get a more accurate prediction of
the tritium permeation through the heat exchangers. Based on that analysis, the effective thickness of
the models was much less than the average because the majority of the tritium passes through the
middle of the models. Therefore, the effective thickness is more similar to the minimum thickness
than to the average distance. A general correlation to predict the effective thickness for tritium
permeation was developed which selected tube diameter, horizontal pitch, and plate thickness as the
main variables.

109
 4. REFERENCES
ABAQUS, Inc, 2004, “Getting Started with ABAQUS.”
AIAA, 1998, “Guide for the Verification and Validation of Computational Fluid Dynamics Simulations,”
AIAA G-077-1998, http://www.grc.nasa.gov/WWW/wind/valid/homepage.html.
ASME, 2004, “ASME Electronic Stress Tables,” Section II, Part D—Properties.
ASPEN Technology, 2004, HYSYS Process Version 2.2.2, www.aspentech.com.
Bejan., A and Kraus, D., 2003, Heat Transfer Hand Book, John Wiley & Son, ISBN 978-0-471-39015-2.
Bell, K. J., 2004, “Heat Exchanger Design for the Process Industries,” Journal of Heat Transfer, Vol. 126,
pp. 877-885.
Bell, K. J., 1981, “Preliminary design of shell and tube heat exchangers, Heat Exchangers: Thermal-
Hydraulic Fundamentals and Design,” Kakac, S., Bergles, A. E., and Mayinger, F., Eds., Hemisphere,
Washington, D.C.
Brent, A., 1989, “Fundamentals of Composites Manufacturing Materials Methods and Applications,”
Society of Manufacturing Engineers, Dearborn, MI.
Collins, J. A., 1981, Failure of Materials in Mechanical Design—Analysis, Prediction, Prevention, Wiley.
COMSOL Group, “Modeling of COMSOL Multihphysics,” 2008.
Crandal, S. H., Dahl, N. C. and Lardner, T. J., 1972, An Introduction to the Mechanics of Solids, Second
Edition, McGraw-Hill Book Company, New York.
Davis, C. B., Oh, C. H., Barnar, R. B., Sherman, S. R. and Wilson, D. F., 2005, Thermal-Hydraulic
Analyses of Heat Transfer Fluid Requirements and Characteristics for Coupling A Hydrogen
Production Plant to a High Temperature Nuclear Reactor, INL/EXT-05-00453.
Dewson, S., Li, X., 2005, “Selection Criteria for the High Temperature Reactor Intermediate Heat
Exchanger,” Proceedings of ICAPP’ 05, Seoul, Korea, May 15–19.
Dewson, S. J. and Thonon, B., 2003, The Development of High Efficiency Heat Exchangers for Helium
Gas Cooled Reactors, ICAPP03.
Dostal, V., Driscoll, M. J., Hejzlar, P., 2004, “A Supercritical Carbon Dioxide Cycle for Next Generation
Nuclear Reactors,” MIT-ANP-TR-100, March 2004.
Forsyth, N., 1974, “Tritium production and distribution in High Temperature Gas Cooled Reactors:
Tritium generation, retention, distribution, and environmental release in a national 1,500 MW(th)
HTR,” O.E.C.D. High Temperature Reactor Project DRAGON, DP REPORT 905, October 1974.
Gainey, B. W., 1976, A Review of Tritium Behavior in HTGR Systems, GA A13461, General Atomic
Company, April 30, 1976.
General Atomic Company, 1978, “Metallurgical Examination of Primary Circuit Components from the
Peach Bottom HTGR, GA A14506, General Atomic Company, February 1978.
General Atomics, 2006, TRITGO Code Description and User’s Guide, GA Project 20128, 2006.
Gezelius, K., 2004, “Design of Compact Intermediate Heat Exchangers for Gas Cooled Fast Reactors,”
Degree of Master Dissertation, MIT.
Heatric™, 2003, Workshop at MIT on 2nd October 2003, Cambridge: MA.
Hesselgreaves, J. E., 2001, “Compact Heat Exchangers,” PERGAMON, ISBN 0-08-042839-8.

110
Hewitt, G. F., 2002, Heat Exchanger Design Handbook, Begell House, INC., New York, 2002.
Kakac, S., Liu, H., 2002, Heat Exchangers: Selection, Rating, and Thermal Design, CRC Press,
ISBN 0-8493-0902-6.
Kato, R., Nishihara, T., Kumitomi, K., 2007, “Design of the Intermediate Heat Exchanger for the High
Temperature Gas-cooled Reactor Hydrogen Cogeneration System,” Japanese Nuclear Society, Vo. 6,
No. 2, pp. 141–148.
Kays, W. M. and Crawford, M. E., 1981, Convective Heat and Mass Transfer, 2nd ed., McGraw-Hill,
New York, 1981.
Kern, D. Q., 1991, Process Heat Transfer, McGraw-Hill, New York, p 143.
Kirch, N. and Scheidler, G., 1980, Control of the Tritium Path in Process Heat HTRs, IWGGCR/2, IAEA.
Lillo T. M., Williamson R. L., Reed T. R., Davis C. B., Ginosar D. M., 2005, Engineering Analysis of
Intermediate Loop and Process Heat Exchanger Requirements to Include Configuration Analysis and
Materials Needs,” INL/EXT-05-00690, Idaho National Laboratory, Idaho Falls, ID.
Martin, J., 1962, Mechanical Behavior of Engineering Materials, Prentice-Hall, Inc.
Mckellar, M. G., 1992, “Optimization of a Household Refrigerator Considering Alternative Refrigerants,”
Ph.D. Thesis, Department of Mechanical Engineering, Purdue University, August 1992.
Natesan, K. et al., 2006, Preliminary Issues Associated with the Next Generation Nuclear Plant
Intermediate Heat Exchanger Design, ANL/EXT-06/46.
National Institute of Standards and Technology, 2005, “Thermophysical Properties of Fluid Systems,”
http://webbook.nist.gov/chemistry/fluid/.
Oh, C. H., Davis, C. B., Han, J., Barner, R., Sherman, S. R., Vilim, R., Lee, Y. J., Lee, W. J., 2006,
HyPEP FY06 Report: Models and Methods, INL/EXT-06-011725, September 2006.
Oh, C. H., Kim, E. S., Sherman, S. R., Vilim, R., Lee, Y. J. and Lee, W. J., 2007, HyPEP FY-07 Annual
Report: A Hydrogen Production Plant Efficiency Calculation Program, INL/EXT-07-13078.
Ohadi, M. M. and Buckey, S. G., 2001, “High temperature heat exchangers and microscale combustion
systems: applications to thermal system miniaturization,” Experimental Thermal and Fluid Science,
Vol. 25, pp. 207–217.
Ohashi, H. and Sherman, S, 2007, Tritium Movement and Accumulation in the NGNP System Interface
and Hydrogen Plant, INL/EXT-07-12746.
Park, I. G, Kim, D. H. and Lee, W. J., 2007, A sample calculation of tritium production and distribution
at VHTR by using TRITGO code, KAERI/TR-3362.
Patterson, M, Personal Communication, 2008, “NGNP Internal Summary, Conceptual Design Studies for
NGNP,” Project Status Meeting, Feb. 7, 2008.
Perry, R. H., Green, D. W., 1997, Perry’s Chemical Engineers” Handbook: 7th Edition.
Richards, M. B., Shenoy, A. S., Brown, L. C., Buckingham, R. T., Harvego, E.A., Peddicord, K. L.,
Reza, S. M. M., Coupey, J. P., Apr. 2006, H2-MHR Pre-Conceptual Report: SI-Based Plant, General
Atomics Report GA-A25401.
Schubert, F., 1984, “Evaluation of Materials for Heat Exchanging Components in Advanced
Helium-Cooled Reactors,” IAEA Specialists Meeting on Heat Exchanging Components of
Gas-Cooled Reactors, Dusseldorf, Germany, IWGGCR-9, 309.

111
Shah, R. K. and Joshi, S. D., 1987, Convective Heat Transfer in Curved Pipes, in Handbook of Single-
Phase Convective Heat Transfer, Kakac, S., Shah, R. K., and Aung, W. Eds., John Wiley & Sons,
New York, Ch. 5.
Simon, R. A., Carp, P. D., 1980, Operating Experience with the DRAGON high Temperature Reactor
Experiment, IWGGCR/2, IAEA.
Slater, J. W., 2008, “Examining Spatial (Grid) Convergence,”
http://www.grc.nasa.gov/WWW/wind/valid/tutorial/spatconv.html.
Special Metals Corporation, 2006, “Inconel Alloy 617,”
http://www.specialmetals.com/documents/Inconel%20alloy%20617.pdf
Srinivasan, P.S., Nandapurkar, S.S., and Holland, F. A., 1970, “Friction for Coils,” Trans. Inst. Chem.
Eng., 48, T156.
Steinwarz, W., D. Stover, R. Hecker, and W. Thiele, 1984, “Distribution of tritium in a nuclear process
heat plant with HTR,” Nuclear Engineering and Design, Vol. 78, pp. 267–272.
Steinwarz, W., H. D. Rohrig, and R. Nieder, 1980, Tritium Behavior in an HTR System Based on AVR
Experience, IWGGCR2, International Atomic Energy Agency, Vienna (Austria), pp. 153 160.
Sunden, B., 2005, “High Temperature Heat Exchangers (HTHE),” Proceedings of Fifth International
conference on Enhanced, Compact and Ultra-Compact Heat Exchangers: Science, Engineering and
Technology, Hoboken, NJ, USA, September 2005.
Wichner, R. P. and F. F. Dyer, 1979, Distribution and Transport of Tritium in the Peach Bottom HTGR,
ORNL 5497, Oak Ridge National Laboratory, August 1979.
Wolf, L., Kazmi, M., Todreas, N. E., 2004, “Introduction to Structural Mechanics,” Lecture Note of
Engineering of Nuclear Reactors, 2004.
Yang, L., W. A. Baugh, and N. L. Baldwin, 1977, Study of Tritium Permeation Through Peach Bottom
Steam Generator Tubes, GA A14376, General Atomics Company, June 1977.
Yook, D. K., “A Study on the Methodology for Tritium Behavior in the Gas Cooled Reactor for
Hydrogen Production System,” Korea Advanced Institute of Science and Technology, Doctorial
Thesis, 2007.
Zukauskas, A. A., 1987, “Convective Heat Transfer in Cross Flow,” in Handbook of Single-Phase
Convective Heat Transfer, Kakac, S., Shah, R. K., and Aung, W., Eds., John Wiley & Sons, New
York, 1987, Ch. 5.

112
113
 Appendix A
High Temperature Heat Exchanger Selection and
Design Guideline

114
 Appendix A
High Temperature Heat Exchanger Selection and
Design Guideline
Heat exchanger design is generally flexible depending on the criteria and designer’s decision. This
appendix briefly summarizes the methods and guidelines for selecting and designing heat exchangers. It
especially focuses on the high temperature heat exchanger (HTHE) design, which was extensively
considered for the intermediate heat exchanger (IHX) design process described in Appendix B.
A-1. GENERAL HTHE ISSUES

HTHE technology has become important for improving performance in power generation. For this
reason, many researchers have been studying the development of a heat exchanger operating with high
temperature gas. The HTHE has some different characteristics compared with the general low
temperature heat exchangers because of high temperatures and pressures. Sunden (2005) summarizes
these characteristics as follows:
x Radiation heat transfer may have a significant role in the high temperature units.
x Tube diameters and pitch should be larger in high temperature units so that the pressure drop is kept
low. The cost of adding a fan or blower to work at high temperatures might be prohibitive.
x Even though the gases have low heat transfer coefficients, fins are generally not used in high
temperature units because (a) the gaseous stream usually carries suspended dirt particles that will foul
or fill up the space between the fins and make a finned tube worse than a plain, and (b) the gas
velocity is low because of low available pressure drop, hence, the advantage of fins is negligible.
x The materials for construction are different in both cases. High temperature units use ceramics or high
alloy and costly tubing; low-temperature units usually use low alloy tubing.
x The selection of materials, their thicknesses, and the mechanical design are governed by the thermal
stress in the high temperature units. Other factors to consider are the extent of material oxidation,
thermal shock bearing capability, erosion because of suspended dirt particles, and fouling and
corrosion because of metallic salts, sulfates, etc. Stress analysis has to be carried out for a safe and
reliable design.
x Differential expansion is an important factor in high temperature units and should be accounted for by
using expansion bellows or bayonet-type units. Floating tube sheets cannot generally by used,
because sealing gasket or packing materials do not work effectively at such high temperatures.
x Heat losses from the outside surface to the environment have to be considered in the mechanical
design of the unit and the foundation design.
x Gases, air, or liquid metals and molten salts are preferred over steam for high temperature heat
transfer because the latter requires a thick shell and tubes to contain its high pressure.
Therefore, the followings should be considered in design of IHX (Sunden 2005):
x The thermal stress during the startup, shutdown, and load fluctuations of HTHEs, can be significant.
Heat exchangers must be designed accordingly for reliability and long life.
x The thermal capacitance should be reduced for high temperature heat exchangers for shorter startup
time.
x High temperature heat exchangers require costly materials, which contributes to the high balance of
power plant cost. Heat exchanger costs increase significantly with temperatures above 675¶C.

115
 A-2. HEAT EXCHANGERS TYPES AND CLASSIFICATIONS
Heat exchangers are used in a variety of applications, including power production, process, chemical
and food industry, electronics, environmental engineering, waste heat recovery, manufacturing industry,
air conditioning, refrigeration, and space application. Heat exchangers can be generally classified as
follows (Kakac and Liu 2002):
x Recuperator/Regenerator
- Recuperations
- Regenerations
x Transfer Process
- Direct contact
- Indirect contact
x Geometry of Construction
- Tubular heat exchanger
- Double pipe heat exchanger
- High pressure (in both side)
- Low heat transfer area
- Shell and Tube heat exchanger
- Thermal expansion problem
- Cleaning
- Spiral tube type heat exchanger
- Thermal expansion problem
- Cleaning problem
- Plate heat exchanger
- Gasketed plate heat exchanger (25 bar, 250°C)
- Spiral plate heat exchanger (15 bar, 500°C)
- Lamella heat exchanger (35 bar, 500°C)
- Extended surface heat exchanger
- Plate-fin heat exchanger
- Tubular-fin heat exchanger (Gas to Liquid)
x Heat Transfer Mechanism
- Single-phase convection on both sides
- Single-phase convection on one side, two-phase convection on the other side
- Two-phase convection on both sides
x Flow Arrangement
- Parallel flow
- Counter flow
- Cross flow.
The heat exchanger type is usually determined in terms of their construction and geometrical features.
As seen in the above categorization, there are two main types of heat exchangers: tubular and plate.
Tubular heat exchangers are built of circular tubes. One fluid flows inside the tubes and the other flows
on the outside of the tubes. Tube diameter, the number of tubes, the tube length, the pitch of the tubes,
and the tube arrangement are flexible. Shell-and-tube type or helical-coil (spiral-tube) type are in this

116
category. Shell-and-tube heat exchangers are the most adaptive type of heat exchangers (see
Figure A-1(a)). They are built of round tubes mounted in large cylindrical shells. They provide a
relatively large ratio of heat transfer area to volume and weight and can be easily cleaned. They provide
great flexibility to meet almost any service requirement. Helical-coil type heat exchangers consists of
spirally wound coils placed in a shell (see Figure A-1(b)). The heat transfer coefficient is higher in a
spiral tube than in a straight tube. Spiral-tube heat exchangers are generally suitable for thermal expansion
and clean fluids, since cleaning is almost impossible.

(a) Shell-and-tube heat exchanger (b) Helical-coil heat exchanger

Figure A-1. Tubular type heat exchangers.

Plate-type heat exchangers are built of thin plates forming flow channels. The fluid streams are
separated by flat plates with smooth or corrugated fins. Compact heat exchangers, including plate-fin heat
exchangers, printed circuit heat exchangers (PCHE), and tubular-fin heat exchangers, are in this category
(see Figure A-2). A heat exchanger having a surface area density greater than about 700 m2/m3 is quite
arbitrary and referred to as a compact heat exchanger. Compact heat exchangers are widely used in
industry, especially as gas-to-gas heat exchangers or liquid-to-gas heat exchangers.

Figure A-2. Printed circuit heat exchanger (PCHE).

Table A-1 summarizes the degree of compactness for different types of heat exchangers. The
compactness of heat exchanger can be described by surface area density. This table shows that the PCHE
has much larger surface area density, 2,000 m2/m3, than the shell-and-tube heat exchangers, 100 m2/m3. It
means that conceptually, the PCHE can be much smaller an perform the same as the shell-and-tube types.

117
Table A-1. Comparisons of heat exchanger compactness.
Exchanger Type Channel Size Area Density
Shell & Tube 10–50 mm 100 m2/m3
Plate type 5 mm 200 m2/m3
Plate fin 2 mm 1,000 m2/m3
PCHE 1 mm 2,000 m2/m3

Table A-2 shows the usual operating ranges of heat exchangers. According to this table, PCHE can
withstand the highest pressure and temperature (900¶C, 1,000 bar). Shell-and-tube type can be operated
up to 650¶C and 1,400 bar. Other types of heat exchangers cannot be operated at high temperature and
pressure conditions.

Table A-2. Usual operating ranges (can be wider with special materials).
Exchanger Type Temperature [°C] Pressure [bar]
Shell & Tube -25–650 300/1,400
Gasketted Plate Frame -35–180 25
Brazed PHE -195–200 30
Welded PHEs ~ 350 70
Spiral HE ~ 400 18
Plate-fin (PFHE) ~ 65 90
PCHE -200–900 1,000

A-3. GENERAL SELECTION GUIDELINE FOR HEAT EXCHANGERS

Bell (1981) suggested the following criteria for heat exchanger selection from various types:
x It must satisfy the process specification (performance): temperature and pressure.
x It must withstand service conditions of the plant environment (reliability): temperature, pressure and
fouling.
x It must be maintainable for cleaning or replacement of a special component.
x It should be cost effective: installed, operating, and maintenance costs
x Site requirements or limitations: diameter, length, weight, and tube configurations, and lifting and
servicing capability or inventory considerations.
Therefore, the followings are considered as main design factors for the selection and design of the
IHX in a VHTR.
x Materials
x Pressure drop
x Pressure level
x Fouling
x Manufacturing techniques
x Size

118
x Cost
x Corrosion control
x Cleaning problem.
A-4. MATERIAL SELECTION FOR HIGH TEMPERATURE
HEAT EXCHANGERS

In the high temperature application, one of the most important things is material selection. There are
four main categories of high temperature materials; high temperature nickel-based alloy, high temperature
ferritic steels and advanced carbon, silicon carbide composite (SiC), and ceramics. (Sunden 2005)

High temperature nickel based material has good potential for helium and molten salts up to 750¶C.
High temperature ferrite steels shows good performance under fusion and fission neutron irradiation, to
temperature around 750¶C. Advanced carbon and silicon carbide composite has excellent mechanical
strength to temperatures exceeding 1,000¶C. It is currently used for high temperature rocket nozzles to
eliminate the need for nozzle cooling and for thermal protection of the space shuttle nose and wing
leading edges. Many options are available that trade fabrication flexibility and cost, neutron irradiation
performance, and coolant compatibility. Table A-3 compares the properties of most commonly used high
temperature materials (Ohadi and Buckley 2001). It includes Ni based alloy, Ceramic materials and
Carbon and SiC composite. Figure A-3 shows the specific strength versus temperature for various
composite materials.

Table A-3. Selected properties of most commonly used high temperature materials and fabrication
technologies (Ohadi and Buckley 2001).

119
Figure A-3. Specific strength vs. temperature (Brent 1989).

Dewson and Li [2005] carried out some material selection studies for VHTR IHXs. They selected
eight candidate materials based on ASME VIII (Boiler and Pressure Vessel Code) and compared them.
The materials included Alloy 617, Alloy 556, Alloy 800H, Alloy 880 HT, Alloy 330, Alloy 230,
Alloy heat exchanger, 253 MA. Table A-4 lists the allowable design stress (S) at 898°C, the minimum
required mechanical properties (ultimate tensile stress [UTS]), 0.2% proof stress (0.2%PS), and
elongation (El) at room temperature together with the nominal compositions of the alloys. They
extensively compared the mechanical properties, physical properties, and corrosion resistance for the
candidate materials, and finally concluded that Alloy 617 and 230 are the most suitable materials for an
IHX.

Table A-4. Candidate materials for IHXs of VHTR (Dewson and Li 2005).
Tmax S898°C UTS 0.2%PS El Nominal compositions
Alloys UNS No (oC) (MPa) (MPa) (MPa) (%) (wt%)
617 N06617 982 12.4 655 240 30 52Ni-22Cr-13Co-9Mo-1.2Al
21Ni-30Fe-22Cr-18Co-3Mo-
556 R30556 898 11.0 690 310 40
3W-0.3Al
800HT N08811 898 6.3 450 170 30 33Ni-42Fe-21Cr
800H N08810 898 5.9 450 170 30 33Ni-42Fe-21Cr
330 N08330 898 3.3 483 207 30 Fe-35Ni-19Cr-1.25Si
57Ni-22Cr-14W-2Mo-0.3Al-
230 N06230 898 10.3 760 310 40
0.05La
HX N06002 898 8.3 655 240 35 47Ni-22Cr-9Mo-18Fe
253MA S30815 898 4.9 600 310 40 Fe-21Cr-11Ni-0.2N

120
 A-5. HEAT EXCHANGER DESIGN METHODOLOGY

A selected heat exchanger must satisfy the process requirements with the allowable pressure drops
until the next scheduled cleaning of the plant. The basic logical structure of the process heat exchanger
design procedure summarized in this section is well explained by Bell (2003).
Figure A-4 shows the diagram of the logical process for heat exchanger design. The fundamental goal
of any such process is to find the optimal design among the infinite set of designs that could satisfy the
thermal/hydraulic and mechanical requirements. Because of large number of qualitative factors in the
design process, optimal needs to be considered broadly. Generally, the design process aims at the least
costly (which usually means the smallest) heat exchangers that meet the required thermal duty within the
allowed pressure drops and satisfy mechanical requirements.

Figure A-4. Basic logical structure of the process heat exchanger

design procedure (Bell 2005).

121
 The first step of the design procedure is to define the problem and provide the designer with all the
data required to solve the problem. This will include the flow rates, compositions, temperatures, pressures,
etc.; design fouling resistances based on the experience with the same or similar materials; and additional
requirements and limitations on diameter, length, weight, piping connections, supports, and construction
features.
The next step is to select a basic exchanger type; a shell-and-tube versus a welded plate, etc. In fact,
there is an increasingly wide choice of exchanger configurations available for most applications, and
some of the established practices should at least be re-examined to see if there might be attractive
alternatives. The following factors should be considered in this step (Bell 2005).
x Level of confidence in the design methodˁis the method backed up by solid performance data either
from the laboratory or the field?
x Level of confidence in the fabrication techniqueˁdoes the equipment have a good service experience
record? What lifetime can be expected?
x Level of confidence in plant operating and maintenance personnel with respect to this equipmentˁdo
they know the characteristics of this equipment, and are they able to maintain and clean it?
x Operational flexibilityˁcan this equipment or its operating conditions be adjusted or modified to
meet likely changes in plant requirement?
x Cost of the equipment.
The next step is to select a tentative but sufficient set of exchanger design parameters to allow the
rating procedure to work. For the case of a shell-and-tube heat exchanger, these would include the process
specifications, Tubular Exchangers Manufacturer’s Association (TEMA) exchanger configuration
identifier, the shell diameter, the tube diameter and wall thickness, tube layout, baffle type, baffle spacing
and cut, and etc. Additional information would include nozzle orientation data, maximum allowable
diameter and length, type of bypass sealing devices, and the requirement that TEMA construction
standards for such items as tube-to-baffle clearance would be met. If a hand design method is to be used,
these values can be estimated as close as possible.
The next step is to rate the performance of the starting design for the specified service. It takes
specified streams and their input conditions and calculates the changes in temperature and pressure
affected on those streams by the specified heat exchanger. The thermal rating output is either the outlet
temperatures and corresponding thermal duty of the exchanger if the length has been specified, or the
length required for the otherwise specified exchanger to satisfy the thermal duty given in the input data.
The pressure drops of each stream are also calculated.
In the design case, the heat duty is compared to that required and the pressure drops to those allowed.
If the duty is sufficient and both pressure drops are near but below the allowed values, the
thermal/hydraulic design is complete and the designer can move the cleaning up and verify the
mechanical design, cost estimations, etc. If all of those conditions are not satisfied, the designer or the
program moves on to the design modification step. The purpose of this step is to examine the output of
the rating program and determine what is good and acceptable and what was unacceptable in the case that
was rated. Then, the designer or the program must select what can be changed to most efficiently correct
the deficiency without too badly upsetting that which was good.

122
 A-6. SHELL-AND-TUBE HEAT EXCHANGER GENERAL
DESIGN GUIDELINE AND CONSTRAINT
The shell-and-tube heat exchanger are the most common type. They are built of round tubes mounted
in a cylindrical shell with the tubes parallel to the shell. This section summarizes the general guideline
and design constraints for this type of heat exchanger. The design options and parameters are based on the
standard suggested by TEMA.

A-6.1 Shell Type

Figure A-5 shows most common shell types standardized by TEMA. Of these types, the E shell
generally provides the lowest cost and best performance, which makes it the most widely used for most
applications. F shell is more effective than E shell in thermal performance, but it has a higher pressure
drop. This type is combined with the U tube bundles. F shell can always avoid temperature cross, but it is
difficult to seal the longitudinal baffles after maintenance. J shell split the flow arrangement to minimize
vibration problem, and X shell with cross-bundle fluid flow is an excellent choice for vacuum
condensation applications.

Figure A-5. Schematics of most common TEMA shell types (ASPEN 2005).

A-6.2 Tube Bundle Type/Tubes and Tube Passes/Tube Layout

The selection and design of tube bundle types generally focus on accommodating thermal expansion,
ease of cleaning, or the least expensive construction. The most common type is a straight tube bundle.
One design variation is the U-tube type. In this type, thermal expansion is unlimited because of the
independent expansion of tubes and shell. But it cannot be cleaned by mechanical means because of the

123
U-bend and individual tubes can only be replaced in the outer row. A fixed-tube sheet configuration has
only limited thermal expansion, and permits no access to the tubes from outside. However, cleaning is
mechanically easy. Several designs have been developed that permit the tube sheet to float, that is, to
move with thermal expansion.
Generally, a large number of tube passes are used to increase tube-side fluid velocity and the heat
transfer coefficient (within the available pressure drop) and to minimize fouling. Tube materials are
usually low carbon steel, low alloy steel, stainless steel, copper, cupronickel, aluminum, or titanium. The
thickness of the heat exchanger tubes is standardized in terms of the Birmingham Wire Gage of the tubes
(Refer to Heat Exchanger Design Handbook, Kakac 2002). TEMA lists nine standard tube sizes ranging
from 6.35 to 50.8 mm (0.25 to 2 in.) in diameter. The most common sizes used are 16 mm (0.625 in.),
19.05 mm (0.75 in.) and 25.4 mm (1 in.). For U tubes, the thinning effect at bends should be considered.
Usually, smaller tube diameters are preferred for greater area/volume density, but they are limited to 20
mm (3/4 in.) for the purpose of cleaning. Tube length affects the cost and operation of heat exchangers.
The longer the tube, the fewer tubes are needed, fewer holes are drilled, and the shell diameters are
decreased, resulting in lower cost. However, there are limits to this. Usually, shell-diameter-to-tub-length
ratio should be within one-fifth to one-fifteenth. Standard tube lengths are 2.44, 3.05, 3.66, 4.27, and 6.1
m. For gas plants and special applications, much longer lengths are possible (20 m or more), with the
maximum tube length being about 30 m, due to transportation limits.
Tube layout is characterized by the included angle between tubes, as shown in Figure A-6. Since the
layout of 30 degrees provides the greatest tube density, it is the most generally used. Pitch-to-diameter
ratio is selected between 1.25 and 1.5.

Figure A-6. Tube layouts.

A-6.3 Baffle Type and Geometry

Baffles serve two functions. One is to support the tubes for structural rigidity, preventing tube
vibration and sagging; the other is to divert flow across the bundle to obtain higher heat transfer
coefficient. They can be classified as transverse and longitudinal types. The most frequently used type is
single and double segmental baffles. Baffle spacing must be chosen very carefully. Optimum baffle
spacing is somewhere between 0.4 and 0.6 of the shell diameter and a baffle cut of 25–35% is usually
recommended. Minimum recommended spacing is 50 mm. Maximum baffle spacing is controlled by the
maximum unsupported length given in TEMA (UNS = 2*baffle spacing) and vibration analyses. Small
baffle spacing (<0.2 times shell diameter) can reduce the cross flow because of leakage. Rod or grid
baffles are formed by a grid or strip supports. In this baffle type, the flow is essentially longitudinal,
resulting in low pressure drop.

124
 A-6.4 Allocation of Streams
A decision must be made as to which fluid will flow through the tubes and which will flow through
the shell. Generally, the following things are considered:
x More seriously fouling fluid flow through the tubes.
x High pressure fluid flows through the tubes.
x The corrosive fluid must flow through the tubes; otherwise both the shell and tubes will be corroded.
x Lower heat transfer coefficients flow on the shell side, since it is easy to design outside finned tubes.
If problems arise when the above requirements are in conflict, the most economical choices must be
found by estimating trade-offs.

A-6.5 Construction
Standard pipe sizes are normally used for shell size up to 610 mm (24 in.). The diameter generally
increases in 50 mm increments for sizes over 610 mm. Most manufacturers are limited to a maximum
shell size of 1,800 mm in diameter, but there are some capable of manufacturing units over 4,300 mm in
diameter. In general, the cost of the exchanger increases significantly as the diameter increases because of
the mechanical design requirement to increase the thickness of the vessel cylinder, body flanges, and tube
sheets.

A-6.6 Vibration
Tube vibration is generally a result of one or more of the followings.
x High fluid velocity
x Large unsupported tube span
x Tube material defects
x Manufacturing process
x Exchanger design/application
x Fluid elastic instability/vortex shedding/turbulent buffering
x Damage patterns by the vibration are:
- Collision damage: impact of tubes against each other
- Baffle damage: tube wall thinning at baffles
- Fatigue: tube failure at the tube sheets
- Tube joint failure: leakage at tube to tube sheet joint
- Tube material defect propagation: failure at inherent tube defect.
Figure A-7 shows the area prone to vibration. To relieve vibration, ratio of cross velocity to critical
velocity should be less than 1.0, and vertex shedding amplitude should be less than the limit governed
primarily by the tube diameter. In addition, turbulent buffering amplitude should be less than the
recommended maximum limit based on the tube diameter.

125
 Figure A-7. Areas prone to vibration (ASPEN 2005).

A-6.7 Modification of Design Parameters

For the heat exchanger to meet the requirements, main design parameters should be modified.
Table A-5. summarizes the guideline for design parameters changes and criteria.

Table A-5. Guideline for design parameter change.

Adjusted High High
Parameters Pressure Pressure Low Low
Optimized Drop Shell Drop Tube Coefficient Coefficient Temperature Vibration
by Program Side Side Shell Side Tube Side Cross Indication
Baffle type Double/Triple – Single – – Double/triple
segmental Segmental segmental
Shell type J or X type – E or F type – E,F or G type J or X type
shell shell shell shell
Tube pattern Rotated – Triangular – – Rotated
square or Square
square
Tube Increase to 1’’ Increase to Decrease to Decrease to – Increase to
diameter or 1/25’’ 1” to 1.25‘’ 0.625’’ or 0.625’’ or 1’’ or 1.25’’
0.5‘’ 0.5’’
Baffle cut Use 30% to – Use 15% to – – –
40% 20%
Tube pitch Increase to 1.4 – Limit to – – Increase to
or 1.5 X tube TEMA std 1.4 or 1.5 X
OD spacing tube OD
Fluid Switch sides Switch sides Switch sides Switch sides – Switch sides
allocation
Arrangement Increase # of Increase # of Increase # of – Increase # of Increase # of
exch. In exch. In exch. In exch. In exch. In
parallel parallel series series parallel
# Tube – Plain – Increase # of Limit to one –
passes tube passes tube pass
Tube type Plain Ext. Internally – –
Enhanced enhanced

126
 A-7. PCHE DESIGN GUIDELINE AND CONSTRAINT

The PCHE is a compact heat exchanger manufactured by Heatric. The flow path of this heat
exchanger is created by chemical etching on the metal sheet, and they are combined though a diffusion
bonding technique. A PCHE is all welded so there is no braze material employed in construction, and no
gaskets are required. Hence the potential for leakage and fluid compatibility difficulties are reduced and
the high level of constructional integrity renders the designs exceptionally well suited to critical high
pressure applications, such as gas compression cooling exchangers on offshore platforms.
The thermal design of printed circuit heat exchangers is subjected to very few constraints. Fluids may
be liquid, gas or two-phase, multistream and multipass configurations can be assembled and flow
arrangements can be truly counter-current, co-current or cross-flow, or a combination of these, at any
required pressure drop.
Where required, high heat exchange effectiveness (over 98%) can be achieved through very close
temperature approaches in counter-flow. To simplify control or further maximize energy efficiency, more
than two fluids can exchange heat in a single core. Heat loads can vary from a few watts to many
megawatts, and these exchanger’s can weight from a few kilograms to thousands of kilograms.
Flow induced vibration, an important source of failure in shell-and-tube exchangers, is absent from
printed heat exchangers.
Lots of useful information for PCHE design was reported by Gezelius 2004 in his thesis, which
includes the summary of the workshop in 2003 between MIT and Heatric. In this part, we summarized the
guideline and criteria for designing the PCHE for IHX.
x No gasket and blazing (risk of leak is considerably reduced): two order of magnitude lower
x Very low vibration damage
x No fouling under clean gas condition
x Surface area density: about 2,500 m2/m3
x No heat transfer and friction factor correlations are available
x Semi-circular cross-section
x Width: 1.0–2.0 mm (2.0 mm shows maximum thermal performance and economic efficiency but for
nuclear application, 1.2 mm is suggested.)
x Depth: 0.5–1.0 mm
x Weight based costing: $30/kg for stainless steel, $120/kg for titanium, expected to be less than $40/kg
for nuclear application
x Carbon steel is typically not used because of the small channel diameter vulnerable to corrosion and
unsuitability for diffusion bonding.
x Average mass-to-duty ratio: 0.2 tones/MW (13.5 tones/MW in shell-and-tube design)
x No constraint to the pressure drop
x Plate thickness: 0.8D (D: channel diameter)
x Channel pitch: 1.22D (D: channel diameter)

127
x Multiport Printed Circuit Heat Exchanger module size: width: 0.5m (1.5m is max), height: 0.6 m,
depth: 0.4~0.6 m.
x Fatigue can be caused by thermal transient.
x Only pressure drop restrict the velocity.
x Minimum life is 20 years.

128
129
 Appendix B
Thermal Design Method of Helical Coil Heat
Exchanger

130
 Appendix B
Thermal Design Method of Helical Coil Heat
Exchanger
A simple thermal design method for helical coil heat exchangers is described in this appendix.
Figures B-1 and B-2 show the schematics of helical coil heat exchangers and the tube bundles for HTTR
IHX (Kato et al. 2007).

Figure B-1. Schematics of HTTR IHX (Kato et al. 2007).

131
 Figure B-2. Structural diagram of heat transfer tube
bundle of HTTR (Kato et al. 2007).

B-1. GEOMETRICAL PARAMETERS

In this type of heat exchanger, the number of tubes in the bundle can be determined by

Rout  Rin  Lm
Nt u . (B-1)
p p
where

N t = Number of tubes in the bundle (#)

Rin = Shell inner diameter (m)

Rout = Shell outer diameter (m)

Lm = Tube bundle height (m)

p = Tube pitch (m)

The Rout and Rin has the following relationship because the tube bundles connected at the inner hot
duct.

( Rout  Rin ) d 2S ˜ Rin (B-2)

132
 Therefore, minimum Rin can be expressed as follows:

Rout
Rin (B-3)
3S
The tube lengths can be calculated as follows:
1. Inner layer

Lt ,in 2SRin u N b (B-4)

2. Outer layer

Lt ,out 2SRout u N b (B-5)

3. Middle layer

Lt ,middle S ( Rin  Rout ) u N b (B-6)

where
Lt,in = Tube length in the inner layer (m)
Lt,out = Tube length in the outer layer (m)
Lt,middle = Tube length in the middle layer (m)
Nb = Number of rotations of the tube bundle (m)
The shell length is calculated by

Ls Lm u N b (B-7)

The tube outer surface area can be approximately determined as follows:

Ah Sd 0 ˜ Lt ,middle ˜ N t (B-8)

B-1.1 Heat Transfer Correlations

The heat transfer of the heat exchanger can be expressed by

Q U ˜ As ˜ 'Tln (B-9)

where

1
U|
1 1
 Rw 
hs ht

'Tln is log mean temperature.

133
 The heat transfer coefficient in the shell and tubes can be expressed as follows:
1. Shell Side—inline tube bundles in smooth pipe (Zukauskas 1987)

Figure B-3. Tube Bundle Arrangement (inline array).

0.25
0.4 0.36 § Prb ·
Nu b 0.9 ˜ c n Re b Prb ¨¨ ¸¸ (for Reb = 1—102) (B-10)
© Prw ¹
0.25
0.5 0.36 § Prb ·
Nu b 0.52 ˜ c n Re b Prb ¨¨ ¸¸ (for Reb = 102—103) (B-11)
© Prw ¹
0.25
0.63 0.36 § Prb ·
Nu b 0.27 ˜ c n Re b Prb ¨¨ ¸¸ (for Reb = 103—2X105) (B-12)
© Prw ¹
0.25
0.8 0.4 § Prb ·
Nu b 0.033 ˜ c n Re b Prb ¨¨ ¸¸ (for Reb = 2X105—2X106) (B-13)
© Prw ¹
where,
cn = a correction factor for the number of tube rows (the effect of the number of tubes becomes
negligible only when n > 16.)
Prb = Prandtl number for the bulk mean temperature
Prw = Prandtl number for the wall temperature

134
 The Reynolds number (Reb) is defined based on the average velocity through the narrowest cross
section formed by array, that is, the maximum average velocity.

U 0d0 U
Re b (B-14)
P
where

m s
U0 (B-15)
U ˜ S ( Rout  Rin 2 ) ˜ (1  d 0 / p)
2

(b) Tube Side—helical coil for turbulent flow (Schmidt’s correlation [Shah et al. 1987])
0.8
Nu c ª § a ·º§ a ·
1.0  3.6 «1  ¨ ¸»¨ ¸ (B-16)
Nu s ¬ © R ¹¼© R ¹
where
Nuc = Nusselt Number for the curved coil
Nus = Straight Tube Nusselt Number
a = Radius of the tube (=d0/2)
R = Radius of the curvature (Rin ~ Rout)
The straight tube Nusselt number for turbulent flow can be calculated by Dittus-Boelter correlation
(Kays and Crawford 1981) as
0.8 0.5
Nu s 0.022 Re b Prb (B-17)

B-1.2 Pressure Drop in Tube Bundles

1. Shell Side—Pressure drop in tube bundles in cross flow (Kakac and Liu 2002)
The pressure drop of multirow bundle is given by

§ Eu · 1 2
'Pshell ¨¨ ¸¸ F UU 0 ˜ n (B-18)
© F ¹ 2
where
Eu = Euler number
F = correction factor
N = number of tube rows counted in the flow direction

135
2. Tube Side—Pressure drop in tube bundles in cross flow (Kakac and Liu 2002)
The pressure drop in the tube can be calculated by
2
L um
'P tubes 4f U (B-19)
di 2

where
f = friction factor
L = tube length (m)
Di = tube inner diameter (m)

um = mean flow velocity in the tube (m/s)

The mean velocity can be determined by

m t
um (B-20)
U ˜ (S ˜ d i 2 ) ˜ N t
The friction factor of the helical coil for the turbulent flow is as follows (Srinivasan et al. 1970):
0.2 2
§R·
0.5
ª § R · 2 º §R· R
fc ¨ ¸ 0.0084 «Re¨ ¸ » for Re¨ ¸  700 and 7   104 . B-21)
©a¹ ¬« © a ¹ ¼» ©a¹ a

136