GENERAL ATOMK3

GA-A14123
DC-7 7

EVALUATION OF STEAM
GENERATOR DESIGNS
FOR APPLICATION TO
THE 300-MW(e) GAS-COOLED
FAST BREEDER REACTOR

by
NOTICE
C. J. BAROCZY This report was prepared as an account of work
sponsored by the United States Government. Neither
the United States nor the United States Energy
Research and Development Administration, nor any of
their employees, nor any of their contractors,
subcontractors, or their employees, make any
warranty, express 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.
Prepared under
Contract EY-76-C-03-0167
Project Agreement No. 23
for the San Francisco Operations Office
U.S. Energy Research and Development Administration

GENERAL ATOMIC PROJECT 3228 DATE PUBLISHED: MAY 1977

- "RIBUT1QN OF THIS DOCUMENT IS UNttWITEO
 DISCLAIMER

This report was prepared as an account of work sponsored by an

agency of the United States Government. Neither the United States
Government nor any agency thereof, nor any of their employees,
makes any warranty, express 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. Reference
herein to any specific commercial product, process, or service by
trade name, trademark, manufacturer, or otherwise does not
necessarily constitute or imply its endorsement, recommendation, or
favoring by the United States Government or any agency thereof. The
views and opinions of authors expressed herein do not necessarily
state or reflect those of the United States Government or any agency
thereof.

D IS C L A IM E R

Portions of this document may be illegible in electronic image

products. Images are produced from the best available
original document.
 NOMENCLATURE

2 2
A * tube surface area, m (ft )
D ** hydraulic diameter, m (ft)
f « Fanning friction factor
2 2
G = mass velocity, kg/s-m (Ib/hr-ft )
2 8 2
g = gravitational acceleration =9.8 m/s (4.18 x 10 ft/hr )
h = heat transfer coefficient, W/m^* °C (Btu/hr-ft^-°F)

K = pressure loss coefficient

2
k = thermal conductivity, W/m-°C (Btu/hr-ft -°F/ft)
L = length, m (ft)
N = Prandtl number
pr
N^ = Reynolds number
P = pressure, MPa (psia)
2 2
AP = pressure drop, kg/m (Ib/ft )
2 2
Q/A = heat flux, W/m (Btu/hr-ft )
T = temperature, °C (°F)
x = steam quality
6 = free to frontal area ratio
p = density, kg/m^ (lb/ft^)

p = viscosity, kg/m*s (Ib/ft-hr)

Subscripts

g = gas
i = inside
1 = liquid
o = outside
s = saturation
w = wall

iii
« »
41

ABSTRACT

Several steam generator designs are considered as possible alternates

to the reference GCFR helical coil design. Sizing, comparison, and evalu­
ation of straight tube and helical coil steam generators indicate that, for
GCFR application, the helical coil design is more attractive. Major areas
of steam generator design and concern encompassing materials of construc­
tion, operating experience, boiling behavior in straight tubes and helical
coil tubes, and associated stresses are discussed and evaluated. Potential
problems and those areas in which knowledge is incomplete are identified.

v
K » « *
 CONTENTS

. NOMENCLATURE...................................................... iii
ABSTRACT .......................................................... v
1. INTRODUCTION AND SUMMARY.........................................1-1
1.1. Introduction ............................................... 1-1
1.2. Summary................................................... 1-1
2. GCFR NUCLEAR STEAM SUPPLY SYSTEM................................ 2-1
2.1. Introduction ............................................... 2-1
2.2. System Description ...................................... 2-1
References...................................... 2-8
3. REVIEW OF STEAM GENERATORS USED INNUCLEAR POWER PLANTS. .... 3-1
3.1. Steam Generator Types...................................... 3-1
3.2. Operating Experience ..... ............ . .......... 3-9
3.2.1. Water-CooledReactors................................3-9
3.2.2. Gas-CooledReactors.................................3-10
References...................................... . 3-11
4. THERMAL-HYDRAULIC ANALYSISOF STRAIGHTTUBE STEAM GENERATOR. . . 4-1
4.1. Water/Steam Side.......................................... 4-2
4.1.1. Heat Transfer...................................... 4-2
4.1.2. Pressure Drop....................................... 4-5
4.2. Helium Side.......................... 4-6
4.2.1. Heat Transfer...................................... 4-6
4.2.2. Pressure Drop...................................... 4-6
References.......................................................4-7
5. CRITICAL HEAT FLUX FOR HIGH-PRESSUREWATER ....................... 5-1
5.1. Straight Tubes...................... 5-1
5.2. Helical Coil Tubes.................................. 5-5
References............................. .................... .. 5-11

6. LOW FLOW BOILING STABILITY.............................. . . 6-1

6.1. Straight Tubes........................................ 6-1
6.2. Helical Coil Tubes ..... ................ ..... 6-5
References.............. 6-8

vii
7. STRUCTURAL EVALUATION................ 7-1
7.1. General Considerations ..... ........................ 7-1
7.2. Materials. ..... .................................... 7-2
7.2.1. HTGR Steam Generator Materials . ................... 7-4
7.2.2. ASME Code Requirements...................... . . 7-6
7.2.3. Material Cost...................................... 7-10
7.2.4. Permeability of Metals to Tritium. . ............ 7-12
7.2.5. Corrosion.............. 7-15
7.3. Tube Stresses. ..................... ...................... 7-18
7.3.1. Critical Heat Flux Induced Evaporator Tube
Stresses ...... ............................ 7-18
7.3.2. Tube-Tube Support Stresses . . .................. 7-21
References...................................................... 7-23
8. COMPARISON OF HELICAL COIL AND STRAIGHT TUBE STEAM GENERATORS
FOR GCFR .......................................................... 8-1
8.1. Current Straight Tube Steam Generators .................. 8-1
8.1.1. Babcock & Wilcox Unit................ 8-1
8.1.2. Clinch River Breeder Reactor . .................. 8-3
8.1.3. Super Phenix Breeder Reactor . . ................. 8-6
8.2. Helical Coil and Straight Tube Steam Generators for
GCFR........................ 8-8
8.2.1. Advantages and Disadvantages .................... 8-8
8.2.2. Bundle Sizes ............................... ... 8-13
8.3. Steam Generator Arrangements in PCRV ........... 8-20
References .......................................................8-24
9. CONCLUSIONS.......... 9-1

FIGURES

2-1. Nuclear steam supply system for GCFR demonstration plant . . . 2-2
2-2. Simplified flow diagram of nuclear steam supply system for
GCFR demonstration plant ....................................... 2-4
2- 3. GCFR steam generator................................. 2-6
3- 1. Westinghouse U-tube steam generator. . ..................... 3-2
3-2. Babcock & Wilcox steam generator........................ . . 3-3
3-3. Fort St. Vrain steam generator.................................3-4

viii
 FIGURES (Continued)

3-4. THTR steam generator................................... 3-5

3-5. Clinch River breeder reactor plant steam generator...... 3-7
5-1. Critical quality versus heat flux (straight tube) .......... 5-3
5-2. Temperature versus steam quality (helical coil) ............ 5-8
5- 3. Steam quality versus angle (helical coil) .................. 5-9
6- 1. Temperature versus length (heat transfer mode change) .... 6-4
7- 1 . Plate and tube temperatures in HTGR and GCFR boilers... 7-22
8- 1. Babcock & Wilcox steam generator........................ 8-2
8-2. Clinch River breeder reactor plant steam generator...... 8-4
8-3. Stein Industrie steam generator ............................ 8-7
8-4. Bundle length versus bundle diameter for helical coil and
straight tube steam generators. ............................ 8-18
8-5. Tubing weight and surface area versus bundle diameter for
helical coil and straight tube steam generators ............ 8-19
8-6. Helical coil steam generator configurations (Incoloy 800H). . 8-21
8-7. Helical coil steam generator configurations (2-1/4-1 Mo). . . 8-22
8-8. Straight tube steam generatorconfigurations.................. 8-23
8-9. Steam generator arrangement Alternate A (helical coil). . . . 8-27
8-10. Steam generator arrangement Alternate B (helical coil). . . . 8-29
8-11. Steam generator arrangement Alternate C (helical coil). . . . 8-31
8-12. Straight tube steam generator arrangement ................... 8-33

TABLES

3-1. Steam generator comparison between PWR, HTGR, LMFBR, and

GCFR.................... ................................... 3-8
3-2. Steam generator failures in gas-cooled nuclear reactors . . . 3-12
6-1. Range of operating conditions for low flow boiling stability
tests (straight tube)............. ........................ 6-3
7-1. HTGR steam generator materials and projected maximum
temperatures. . ............................................... 7-5
7-2. Comparison of ASME Code and RDT M3-33 requirements for steam
generator tubing........ .................................. 7-8
7-3. Allowable stresses. . ......... 7-9

ix
 TABLES (Continued)

8-1. General characteristics of helical coil and straight tube

steam generators............................................ 8-9
8-2. 300-MW(e) GCFR steam generator characteristics (helical) . . . 8-15
8-3. 300-MW(e) GCFR steam generator characteristics (straight). . . 8-16

x
 1. INTRODUCTION AND SUMMARY

1.1. INTRODUCTION

The study described herein was performed to review the steam generator
design for the Gas-Cooled Fast Breeder Reactor (GCFR) Demonstration Plant
and to evaluate alternate designs with the objective of identifying the
most attractive design configuration and tube material.

Included in this report is a description of the reference GCFR steam

generator design, followed by a review of nuclear plant steam generator
design, technology, and operating experiences. A comprehensive thermal-
hydraulic analysis of the straight tube and helical coil tube configurations
is presented together with an evaluation of steam critical heat flux
behavior in each configuration. Low flow (as low as 2% of full flow)
boiling water stability is an important, and unique, requirement for the
GCFR, and its implications are discussed for both straight tube and helical
coil tube designs. Materials characteristics such as allowable stresses,
corrosion behavior in water/steam and helium, tritium permeability, and
suitability for the helical coil tube support system are discussed and
evaluated. In each of the areas evaluated, potential problems and those
sections requiring further work are identified.

1.2. SUMMARY

Because of their incorporation into cavities in the prestressed con­

crete reactor vessel (PCRV), steam generators for the GCFR must fulfill
stringent size and geometry requirements. For example, the three once-
through steam generators must not only be compact to limit PCRV size but,
in addition, should be designed so as to contain the main bundle and
resuperheater in a single physical unit in order to limit the amount of

1-1
helium ducting within the PCRV. As a result, separate evaporator-
superheater combinations (with or without a steam drum) are not considered
practical because their use can result in longer, more complex, or more
numerous PCRV cavities. In addition to the above shortcomings, the thermal
stress problems associated with the high helium/water temperature differ­
ences and high level of steam superheat in the GCFR rule against the use
of a U-tube steam generator. While the straight tube "hockey-stick" con­
figuration is attractive, accommodation of this shape, particularly the
upper bend portion, with a PCRV closure plug presents a difficult design
problem. Consequently, the straight tube once-through steam generator
appeared to be a reasonable alternate to the present helical coil tube
reference design and, on this basis, was evaluated for possible GCFR
application.

The results of this evaluation indicate that the straight tube steam
generator can be a practical alternate to the helical coil unit as it has
the advantages of relatively simple design and assembly, reduced fabrication
cost, low water side pressure drop, and demonstrated favorable low water
flow boiling behavior. Disadvantages associated with the straight tube
steam generator include (1) a large number of tubes, (2) large, thick tube-
sheets and heads, (3) costly tube support-spacers, (4) inefficient utili­
zation of available helium pressure drop due to tube spacer pressure loss,
(5) difficulty in providing for tube thermal expansion, and (6) extra
helium pressure drop penalty due to two entrance and exit losses if a
resuperheater is used.

Even though the helical coil - straight tube steam generator compari­
son utilized a cycle without a resuperheater, which eliminated one of the
major disadvantages of the straight tube unit, i.e., two helium entrance
and exit pressure losses, the helical coil configuration possessed additional
advantages. These include (1) much greater latitude in the placement of the
helium circulator and associated ducting in the PCRV, (2) significantly less
weight and surface area, and (3) simpler accommodation of thermal expansion.

1-2
Thus, despite its greater fabrication complexity and cost, the helical
coil tube steam generator was judged to be the most attractive for GCFR
application.

Power generation industry, High-Temperature Gas-Cooled Reactor (HTGR),

and Liquid Metal Fast Breeder Reactor (LMFBR) experiences and research
programs on the mechanical behavior, corrosion resistance, and tritium
permeability of steam generator tubing were examined and evaluated in terms
of GCFR requirements. The use of 2-1/4 Cr-1 Mo steel, instead of the
present reference material, Incoloy 800H, offers advantages in terms of
reduced cost and low alloy content, high resistance to stress corrosion
cracking in chloride and caustic contaminated water/steam, and extensive
operating experience. If near-future tritium permeability tests of
2-1/4 Cr-1 Mo steel demonstrate acceptably low levels (Incoloy 800H has
low tritium permeability in both the clean and oxidized states), then
2-1/4 Cr-1 Mo steel represents the best candidate material for use as
steam generator tubing.

As part of the safe shutdown procedure for the GCFR, it is required

that the steam generator operate stably and produce superheated, or high
quality, steam at flows as low as 2%. Pertinent to this, investigation
of the boiling behavior of straight vertical tubes and helical coil tubes
reveals significant, inherent differences in critical heat flux charac­
teristics . Another important difference is the near-horizontal orientation
of the tube in a helical coil, which may lead to stratification of the
water and steam phases at low flow. For these reasons, the low flow boil­
ing stability characteristics of helical coil tubes can be expected to
differ from those of straight vertical tubes. Accordingly, a test program
is needed to verify that the required performance can be obtained.

1-3
 2. GCFR NUCLEAR STEAM SUPPLY SYSTEM

2.1. INTRODUCTION

The design of nuclear steam supply systems and components for the GCFR
relies on gas-cooled thermal reactor experience and development programs,
such as those for High-Temperature Gas-Cooled Reactor (HTGR) plants,
including Peach Bottom 1, which operated from 1967 to 1974, Fort St. Vrain,
which is now in the preoperational testing phase, and numerous European Gas-
Cooled Reactor installations.

In extending HTGR primary coolant system technology to the GCFR, cer­

tain design differences arise as consequences of the higher coolant pres­
sure, the lower temperature differences available for heat transfer, the
higher circulator power requirements, and the more stringent demands for
uninterrupted forced cooling during shutdown and decay heat removal cooling
modes. How these aspects affect the design of the steam generator and the
other GCFR system components is described in the GCFR Demonstration Plant
report (Ref. 2-1).

2.2. SYSTEM DESCRIPTION

The principal components of the GCFR nuclear steam supply system (Ref.
2-1) are contained in the PCRV, as shown in Fig. 2-1. The PCRV is a con­
crete structure reinforced with steel rods and prestressed by a system of
longitudinal and circumferential steel tendons. Inside the PCRV there are
a number of steel-lined interconnected cavities: three contain the main
helium circulators and steam generators, and three contain the auxiliary
core cooling system. These cavities are arranged around the center reactor
cavity.

2-1
74LC3637

Fig. 2-1. Nuclear steam supply system for GCFR demonstration plant

2-2
 The function of the steam generator is to transfer heat from the pri­
mary helium coolant to the secondary water/steam coolant. The helium that
cools the reactor core enters the steam generator modules at high tempera­
ture and transfers heat first to the resuperheater section and then to the
main water/steam sections. The cooled helium then passes to the helium
circulator and returns to the reactor core to complete the cycle. Main
coolant circulation is performed by three circulators of the same type as
the integral turbocirculator unit developed for the HTGR, which employs a
single-stage axial compressor driven by a single-stage impulse steam
turbine.

The secondary coolant system conveys steam produced in the steam gen­
erators to the turbine-generator for conversion of the thermal energy into
electrical energy. A simplified flow diagram for the GCFR nuclear steam
system is shown in Fig. 2-2. The superheated steam produced by the steam
generator is used to drive the main helium circulator. The exhaust steam
from the circulator turbine is resuperheated in the steam generator before
it enters the high-pressure section of the turbine generator. The low- ■
pressure exhaust steam leaving the low-pressure section of the turbine gen­
erator is condensed in the surface condenser operating at an assumed 2 in.
(5 cm) Hg pressure. The condensate, having passed through a series of feed-
water heaters, is returned to the steam generator by the boiler feed pumps.

Removal of residual and decay heat from the reactor is normally accom­
plished by the main cooling loops using the steam generators. An auxiliary
core cooling system is provided as an independent backup to the main
cooling loops.

The design of the steam generators for the GCFR will follow HTGR
design and technology as closely as possible, consistent with the specific
requirements of the GCFR. The steam generators for the GCFR are helically
coiled once-through units with downward helium flow and upward boiling
which are similar in concept to those developed and built for the Fort St.
Vrain HTGR. Although the steam generators for the Peach Bottom HTGR are of

2-3
 NET PLANT OUTPUT 300 MW(e)

PLANT EFFICIENCY 36.0%

386 PSIA RESUPER­

(9-6 MPa) HEATER

REACTOR
GENERATOR
2-4

STEAM
GENERATOR

(20 MPa)

MAIN
(5.0 cm)
CIRCULATOR
TURBINE

1305 PSIA HELIUM

(9 MPa) CIRCULATOR HP HEATERS
LP HEATERS

Fig. 2-2. Simplified flow diagram of nuclear steam supply system for GCFR demonstration plant

fc
a different type, the design and operational experiences with them have
provided valuable background in areas such as materials behavior and com­
patibility and special welding technology.

Steam generator development for the GCFR, as described in Ref. 2-2,

will focus on satisfying the requirements for the 300~MW(e) demonstration
plant, which has three reactor coolant loops. It is expected that larger
GCFRs would have more loops and/or higher loop capacity. The larger-
capacity units, however, would be designed to require little additional
development, and since a higher temperature differential (AT) is contem­
plated, the units would not be proportionately larger in linear dimensions.

Each steam generator is composed of a main steam tube bundle and a

resuperheater tube bundle, arranged one above the other, as shown in Fig.
2-3. The tube bundles utilize multistart helical coils to achieve a high
ratio of tube surface area to steam generator volume and thereby minimize
the space requirements within the PCRV.

The incoming high-pressure feedwater is converted to steam as it

passes sequentially through the economizer, evaporator, and superheater.
The high-pres sure superheated steam is used to drive the helium circulator
turbine. The circulator turbine exhaust steam is returned to the resuper­
heater to be reheated before it goes to the main turbine.

Feedwater enters the steam generator through tubes that lead from the
feedwater penetration to the plenum area below the main tube bundle (see
Fig. 2-3). From there, the tubes are routed to form the coils of the
economizer section. The tubes then spiral upward with an equal helix angle
to form the evaporator and superheater sections. Above the superheater
section, the tubes are connected to lead-out tubes that pass into the cen­
ter duct, go downward through the duct to the superheater penetration, and
from there go outside the PCRV to the helium circulator turbine. The
returning steam enters the resuperheater inlet penetration and passes
through the lead-in tubes in the center duct to the resuperheater section.
After passing through the resuperheater, the steam flows downward through

2-5
 8 1 -6"
(0.91 m)
(9.A m'
(0. 1 m)
(1.0

(1.06 m)

11-1/2"
(3-86 m)
(3.05 m)

RSH IN
(0.91 m)

11'-6" DIA
(3.5 m) -RSH
2-6

31-6" DIA DUCT

(1.06 ffl)

VERTICAL COOLING TUBES

| IZ'lflLg:
VIEW A-A

Fig. 2-3. GCFR steam generator

* >
the lead-out tubes in the center duct to the resuperheater outlet penetra­
tion and leaves the steam generator. The steam from all three steam gen­
erators is combined in a mixing header outside the containment building and
then enters a single line which leads to the turbine generator.

Helium from the reactor flows into the steam generator through a hori­
zontal duct near the top of the reactor cavity. The connection between
this duct and the steam generator inlet is a bellows disconnect arrange­
ment , as shown in Fig. 2-3. More recent designs replace the bellows type
seal with an alternate seal arrangement. The hot gas enters a plenum
above the tube bundles, flows downward across the resuperheater tubes, and
then flows across the tubes of the main bundle. After transferring its
heat, the helium leaves the main bundle at the economizer end. From there,
the helium flows radially outward in the steam generator cavity into an
annulus between the steam generator shell and the PCRV thermal barrier and
then upward into a plenum above the steam generator, from which it enters
the circulator inlet.

The tube coil geometry consists of equal-length tubes arranged in

columns of equal height and approximately equal helix angle. The longi­
tudinal pitch of a coil is constant within each row but varies between
rows. The transverse pitch is constant over the full bundle diameter.

The arrangement of the tube bundles and their connections to external

piping are such that all sections of the steam generator can be drained.

The GCFR steam generator modules will be equipped with instrumentation

consisting of temperature sensors and pressure transducers to measure the
inlet and outlet temperatures and pressures of the helium. Also, sampling
tubes will be provided to supply primary coolant to the moisture-monitoring
system. In addition, one of the modules will be provided with more exten­
sive instrumentation for confirming the design during preoperational
testing. This instrumentation will measure water/steam and tube wall tem­
peratures, helium-side pressure drops, vibration amplitudes, and sound
levels.

2-7
REFERENCES

2-1. "300 MW(e) Gas-Cooled Fast Breeder Reactor Demonstration Plant,"

General Atomic Company, unpublished data.
2-2. Khoe, J. H., "Development Plan for the GCFR Steam Generator," ERDA
Report GA-A13236, General Atomic Company, March 3, 1975.

2-8
 3. REVIEW OF STEAM GENERATORS USED IN NUCLEAR POWER PLANTS

As part of the present study, an evaluation was made of possible

alternate steam generator configurations for use instead of the helical
coil reference design. To provide background information and some per­
spective, the steam generators in current use in nuclear power plants were
examined, and the major features of, and experiences with, each type are
described in this section.

3.1. STEAM GENERATOR TYPES

Pressurized water reactor (PWR) systems utilize the U-tube steam gen­
erator (Westinghouse and Combustion Engineering) or the straight tube,
once-through steam generator (Babcock & Wilcox). The U-tube steam genera­
tor shown in Fig. 3-1 is of the recirculating type which produces steam of
up to 50% quality. After passing through driers, essentially dry steam, at
saturation temperature, goes to the turbine. In contrast, the straight
tube, once-through steam generator produces slightly superheated [about
28°C (50°F)] steam (see Fig. 3-2). In both types of steam generators, the
steam is produced on the shell side as the primary fluid flows inside the
tubes.

Gas-cooled reactor systems have used (1) the U-tube steam generator
(Peach Bottom), which initially produces wet steam and, after moisture
separation and additional heating, provides highly superheated steam, and
(2) the helical coil, once-through steam generator (Fort St. Vrain, large
HTGR, and German THTR), which produces highly superheated steam. Typical
units are shown in Fig. 3-3 (Fort St. Vrain) and Fig. 3-4 (THTR).

The Clinch River breeder reactor plant uses a "hockey-stick" shaped

steam generator which is essentially a fixed tubesheet, straight tube steam

3-1
 1000 PSIA (8.9 MPa)
545°F (285°C)

STEAM SEPARATORS

TUBE BUNDLE

TUBE SUPPORTS

LOWER SHELL
PREHEATER OUTLET

FEEDWATER NOZZLE
I 6" SCH. 80
(MAIN A AUXILIARY
440°F FEEDWATER)
(226°C)
if! t 1 - PREHEATER SECTION
TUBE SHEET
PREHEATER OUTLET

yjT 2120 PSIA (14.6 MPa)

V\ 558°F (292°C)
PRIMARY COOLANT INLET
2160 PSIA (14.9 MPa) COOLANT CHANNEL
618°F (325°C)

Fig. 3-1. Westinghouse U-tube steam generator

3-2
 REACTOR
COOLANT
INLET
2174 PSIA (15 MPa)
630°F (332°C)
HANDHOLE

SUPERHEAT
REGION

BOILING
REGION

HEAT TRANSFER
REGIONS AT
100% POWER

STEAM ANNULUS
NUCLEATE
BOILING
REGION

STEAM
OUTLETS (2)
HANDHOLES 1135 PSIA (7.8 MPa)
595°F (313°C)
AUXILIARY FEEDWATER INLET
FEEDWATER SUBCOOLEO
INLETS (2) BOILING
471 °F (244°C) REGION

2 REACTOR /
COOLANT ------------------- *'>
OUTLETS 2166 PSIA (14.9 MPa) ----- -MANWAY
573°F (300°C) HANDHOLE
DRAIN NOZZLE —^

Fig. 3-2. Babcock & Wilcox steam generator

3-3
 Fort St. Vrain steam generator
o m;
S3
oo —toi&-
55 FT. 2-1/4 IN. w u *So u as ~
.(16.8 ra) \ \
I
25 FT. 7 IN. 15 FT. 6 IN. (4.7 m) — 12 FT. 1 IN.'
(7.8 m) CONCRETE SHIELD (3.7 m)
FLANGE
i

Fig. 3-3.
 HP It BUNDLE

Fig. 3-4. THTR steam generator (from Ref. 3-1)

3-5
generator with a large radius bend near the outlet end to accommodate ther­
mal expansion (see Fig. 3-5). Although the original version produced
highly superheated steam in the once-through mode, the Clinch River breeder
reactor steam generator utilizes an evaporator that produces a 50%-quality
steam mixture which, after moisture removal, becomes highly superheated in
a separate superheater. In this design, steam flows inside the tubes while
sodium flows on the outside of the tubes.

The major characteristics of the, above units and the reference design
and updated cycle GCFR steam generators are shown in Table 3-1. In addi­
tion to other distinctions among the units, such as once-through or recir­
culating mode of operation, the PWR units generally have high overall heat
transfer coefficients and low temperature differences, whereas the other
units have low to medium overall heat transfer coefficients combined with
medium to high temperature differences.

Although Table 3-1 shows that wide variations exist in the number of
tubes and surface area for the various steam generators, it can be seen
that, except for the reference design GCFR unit, the steam generator
effectivenesses, in terms of area per unit heat transferred, are compara­
ble. If the updated GCFR reference cycle is used, then the effectiveness
of the steam generator is increased so that it is closer to, but still less
than, the other units.

However, it is emphasized that the total cost of a steam generator

includes not only the amount of tubing surface and weight, but also the
effects of materials and their specifications, temperature-pressure condi­
tions, and fabrication cost, which is largely geometry or configuration
dependent. Thus, the cost per unit of surface area may differ considerably
among steam generators with similar surface effectiveness.

3-6
 VIBRATION
SUPPRESSORS-^^ REMOVABLE
STEAM HEAD

STEAM OUTLET

TUBESHEET

TUBE (5/8-IN. O.D.

X 0.109-IN. WALL)
65 FT.
(15.9 mm X 2.7 mm
(19.8 m
BUTT-WELD (IBW)
SODIUM OUTLET MACHINED
(2 NOZZLES) BOSS
DRAIN NOZZLE

REMOVABLE STEAM/WATER HEAD

STEAM/WATER INLET

Fig. 3-5. Clinch River breeder reactor plant steam generator (from Ref. 3-2)

3-7
 TABLE 3-1
STEAM GENERATOR COMPARISON BETWEEN PWR, HTGR, LMFBR, AND GCFR

Company or Project Westinghouse Babcock & HTGR GCFR GCFR Clinch River
Wilcox Ref. Des. Mod. Cycle Breeder
Reactor

Type U-tube Straight Helical Helical Helical Hockey-stick

Thermal duty, MW(t) 914 1900 506 291 291 325
2 2
Surface area, ft (m ) 48,000 145,000 33,010 35,670 26,065 17,084
(4461) (13,478) (3068) (3315) (2422) (1588)
No. of tubes 4674 15,743 328/351 218/297 180 757 x 2 (EV)
757 (SH)
Tube dia., in. (cm) 0.75 (1.9) 0.625 (1.6) 1.0/1.125/1.75 0.75/1.0 0.75/1.0 0.625 (1.6)
(2.5/2.8/4.4) (1.9/2.5) (1.9/2.5)

Surface effectiveness, 52.5 (4.88) 76.3 (7.09) 65.2 (6.06) 122.5 89.5 (8.32) 52.5 (4.88)
ft2/MW(t) [m2/MW(t>] (11.38)
Tube material Inconel 600 Inconel 600 Incoloy 800 Incoloy 800 Incoloy 800 2-1/4 Cr-1 Mo
2-1/4 Cr-1 Mo
Primary coolant Water Water Helium Helium Helium Sodium
Pressure, psia (MPa) 2157 (14.9) 2170 (15.0) 690 (4.8) 1255 (8.7) 1255 (8.7) 150 (1.0)
Temperature, °F (°C)
In 618 (325) 630 (332) 1366 (741) 1022 (550) 1010 (543) 921 (494)
Out 558 (292) 574 (301) 624 (329) 591 (310) 648 (342) 649 (343)
Secondary coolant Water/steam Water/steam Water/steam Water/steam . Water/steam Water/steam
Pressure, psia (MPa) 1000 (6.9) 1135 (7.8) 2515 (17.3) 2900/1285 2900/1600(a) 1450 (10.0)
(20/8.9)(a) (20/11)(a)

Temperature, °F (°C)
In 440 (227) 471 (244) 370 (188) 412 (211) 406 (208) 468 (242)
Out 544 (284) 595 (313) 955 (513) 876/9280) 955/801(a) 905 (485)
(469/498)(a) (513/427)(a)

Operation mode Recirculation Once-through Once-through Once-through Once-through Recirculation

("s')
To main turbine.

* *
3.2. OPERATING EXPERIENCE

3.2.1. Water-Cooled Reactors

The most extensive steam generator operating experience is that for

PWR systems. A very detailed documentation of operating experience, up to
1973, has been made by Stevens-Guile (Ref. 3-3), and a more recent, but
limited, survey was performed by Baschek and Kocourek (Ref. 3-4). Accord­
ing to Ref. 3-3, 19 of the 34 reactors in operation up to 1971 had experi­
enced tube failures with corrosion of some type as the most frequent cause,.
followed by vibration, manufacturing faults, and damage caused by debris.
Since corrosion is the dominant mode of tube failure, water chemistry on
the secondary (steam) side is thought to be responsible (Ref. 3-4). How­
ever, a chemical coneentration mechanism must be present locally in order
to produce tube failure by corrosion; such conditions may be brought about
by poor flow distribution or by alternating boiling and all-liquid condi­
tions. Thus, tube failure by corrosion can be the result of a complex pro­
cess involving water chemistry, materials, and thermal-hydraulic design.
It is emphasized that the tube failures described above have been observed
on the outside tube surfaces of recirculation-type (steam quality less than
unity) steam generators. To date, operating experience with the once-
through (superheated steam at the exit) Babcock & Wilcox steam generators
in use at six plants has been trouble-free. Since the GCFR steam generator
is of the once-through type and, in addition, the steam flows inside the
tubes rather than on the outside, the probability of flow-maldistribution-
caused chemical concentration in the water is low. Furthermore, the use of
conventional once-through steam generator water treatment, involving the
all-volatile method, will minimize tube fouling at the critical heat flux
location.

While the relatively low operating temperature level, about 343°C

(650°F), of PWR steam generators ’’ould permit the use of comparatively
inexpensive low-alloy steel tubing from stress considerations, Inconel 600
is almost universally used because of its very low overall corrosion rate
and its excellent resistance to stress corrosion cracking. However, there

3-9
have been numerous instances of corrosion failure of Inconel 600 steam gen­
erator tubes in the low fluid velocity region just above the tubesheet and
in the upper (U-bend) portion of the bundle at antivibration supports.

In addition, many steam generators of the U-tube type have experienced

separation of the Inconel alloy cladding from the low-alloy steel tube-
sheet. These failures have been eliminated by applying the cladding by the
weld deposit method instead of the previously used explosive bonding
method.

It is not clear whether the steam generator tube failures were due to
low fluid velocity or improper water chemistry, but support for the latter
cause is evident in the change, by Westinghouse, from the previously
preferred phosphate water treatment method to the zero-solids method (Ref.
3-5). The latest Westinghouse steam generator designs (Ref. 3-5) use not
only zero-solids water treatment but also a modified feedwater inlet flow
distribution system which largely eliminates the very low fluid velocity
regions on the outside of the tubes. Sludge tends to form on the tube-
sheets and tubes in the low fluid velocity regions, and contaminants, which
exist at tolerable levels in the bulk fluid, can concentrate under the
sludge to form potentially aggressive solutions. Westinghouse expects that
the change in water treatment method and elimination of low fluid velocity
sections on the steam side of the steam generators will prevent future tube
failures due to corrosion.

3.2.2. Gas-Cooled Reactors

An extensive survey of operating experience, up to 1974, for steam

generators used in gas-cooled nuclear reactor plants has been made by
Borne lburg and Schuetzenduebel (Ref. 3-6).

According to this study, the primary causes of failure in gas-cooled

reactor steam generators were:

3-10
 1. Defective tube welds which resulted in water leaks. This condi­
tion was corrected by the increased use of machine welds, shop
(rather than site) fabrication, and more stringent quality
control.

2. Vibration problems due to high turbulence, poor flow distribu­

tion, or vortex shedding; or acoustic problems such as standing
waves in shell side cavities. These problems, which in some
cases caused tube failure, were corrected by the use of gas
baffles, plates, or screens. Because of the low density and high
acoustic velocity of helium, vibration and acoustic problems are
expected to be minimal in GCFR and HTGR units as compared with
those experienced using CC^ as coolant.

3. Gas side corrosion. After a considerable period of operation

with CO2 coolant, mild steel parts in the primary circuit became
corroded. For corrective action, the affected parts were
replaced by stainless steel parts. Because helium is an inert
gas, such corrosion is not anticipated in either the GCFR or
HTGR primary system.

A summary of gas-cooled reactor plant steam generator failures is

shown in Table 3-2 (from Ref. 3-6).

REFERENCES

3-1. Fricker, Hans W. , "Design and Manufacturing Experience for the German
Thorium High-Temperature Reactor 300-MW(e) Steam Generator," Nucl.
Technol. 28, No. 3, 339 (1976).
3-2. Patriarca, P., et al. , "U.S, Advanced Materials Development Program
for Steam Generators," Nucl. Technol. 28, No. 3, 516 (1976).
3-3. Stevens-Guile, P. D., "Steam Generator Tube Failures: A World Survey
of Water-Cooled Nuclear Power Reactors to the End of 1971," Atomic
Energy of Canada Ltd. Report AECL-4449, April 1973.

3-11
 TABLE 3-2
STEAM GENERATOR FAILURES IN GAS-COOLED NUCLEAR REACTORS

Tube Leak Vibration Other Problems and

Name Causes(a) Problems(k) Miscellaneous Remarks

Calder Hall WD
Chapelcross WD
Berkeley WD Design fault in superheater; only 1
tube leak reported
Bradwell WD, CC
Hunterston A WD, CC Poor flow distribution
Hinkley Point A CC
Trawsfynydd WC Cracked bellows; WC caused by
improper water treatment
Dungeness A SF, CC Yes
Sizewell CC
Oldbury CC
Wylf a —
Latina SF, CC Yes Vibration caused leaks owing to
stress failure
Tokai-Mura WD, CC Yes Failure of insulation cover plates;
has experienced frequent failures
G2, G3 Marcoule CE, CC Cracked bellows
EDF1, Chinon WD, CC Faulty design caused expansion
cracks of tube joints
EDF2, Chinon CE, CC
EDF3, Chinon WD, CE, CC Yes Leaking bellows
St. Laurent 1 WD • Erosion at feedwater inlet orifices
St. Laurent 2 —
Erosion at feedwater inlet orifices
Bugey I —

Windscale AGR Poor QC Only 2 tube leaks reported

AVR Pebble Bed —

THTR —
Still under construction
Dragon WC Combination of faulty design and
improper water treatment; first
set of SGs was replaced
Peach Bottom Contamination SGs were retubed
Fort St. Vrain —
In commissioning phase
Bohunice
Brennilis EL4 Faulty design First set of SGs was replaced
Lucens —
Vandellos —

Niederaichbach Yes Cause of leaks not yet established

Hartlepool —
In construction phase
Hinkley Point B — In commissioning phase
Hunterston B —
In commissioning phase
Dungeness B — In commissioning phase
Heysham ——
In construction phase

(a)
WD - weld defects. No longer a problem because of better welding techniques
and more stringent quality controls.
CC = CO corrosion. No problem when using helium.
WC = water corrosion. Can be avoided by proper water treatment.
SF = stress failure.
CE = corrosion-erosion. Can be avoided by proper design, verified by tests.
QC = quality control. More effective NDT techniques are now available.
^^Vibration problems are still possible, but are less likely to occur because
of improved understanding.

3-12
 3-4. Baschek, H., and E. Kocourek, "Operational Experience with Heat
Exchangers in Nuclear Power Stations with Light Water Reactors,"
Combustion 47, No. 3, 14 (1975).
3-5. Fletcher, W. D., and D. D. Malinowski, "Operating Experience with
Westinghouse Steam Generators," Nucl. Technol. 28, No. 3, 356 (1976).
-v

3-6. Bomelburg, H. J., and W. G. Schuetzenduebel, "Operating Experience

with Steam Generators in Gas-Cooled Reactors," General Atomic Report
GA-A13036, September 13, 1974.

3-13
 4. THERMAL-HYDRAULIC ANALYSIS OF STRAIGHT TUBE STEAM GENERATOR

After the U-tube and straight tube steam generators had been compared
as possible alternates to the helical coil configuration, the straight tube
configuration was chosen for the following reasons: (1) both the Babcock &
Wilcox steam generator and the Clinch River breeder reactor steam generator
are essentially straight tube designs, (2) extensive manufacturing and
operational experience exists for straight tube heat exchangers, (3) a
straight tube steam generator is more compatible with the size and geometry
limitations imposed by the available space in a PCRV than a recirculating
type U-tube steam generator, (4) fabrication of a straight tube steam gen­
erator is simpler and less expensive than for a helical coil steam genera­
tor, (5) a straight tube steam generator HTGR design study has been per­
formed by GA for application to commercial gas-cooled reactor nuclear power
plants.

Although the comparison between helical coil and straight tube steam
generators (Section 8) utilized a computer code, initially the straight
tube steam generator designs were hand calculated. This served a useful
purpose in that it provided a method for incorporating the low mass
velocity critical heat flux data into the calculation, and also supplied
the basis for computing the tube support spacer helium pressure drop. This
information was then utilized in the computer code calculations.

While some internal differences exist between the computer code (Sec­
tion 8) and hand-calculated examples, the sizing results calculated by
both methods compared well. The calculation method is described in the
following sections.

4-1
4*1. WATER/STEAM SIDE

4.1.1. Heat Transfer

The straight tube steam generator thermal performance was analyzed by

means of a temperature (helium and steam) versus steam enthalpy plot using
short sections, corresponding to an enthalpy change of about 116,300 J/kg (50
Btu/lb), and the appropriate overall heat transfer coefficient. From the
known enthalpy change, helium to steam temperature difference, and overall
heat transfer coefficient, the required heat transfer surface area and
corresponding tube length were determined for each short section.

The heat transfer coefficients for each of the sections were deter­
mined as described below.

4.1.1.1. All-Liquid Section. The Dittus-Boelter equation (Ref. 4-1),

*U A AO “3 ^ ufO • 8 0.4
hi “ °-023 d7 Nr NPR (4-1)

was used in the all-liquid (economizer) section.

The above expression was used up to the point where the inside tube
wall temperature reached saturation; at this point, subcooled nucleate
boiling started. Since the design pressure [20.3 MPa (2950 psia)] and
temperature are close to the critical point of water, the thermal proper­
ties of water change very rapidly as saturation conditions are approached.
For example, at 20.3 MPa (2950 psia) and 367°C (693°F) (saturation), the
Prandtl number for water is 6.7; at 354°C (670°F) the Prandtl number has
dropped to 2.6. The marked increase in the Prandtl number and thermal
conductivity of water with temperature can result in doubling of the water
side heat transfer coefficient from economizer inlet to outlet. The water
properties at 343°C (650°F) were used to compute the economizer outlet heat
transfer coefficient, since it was at this bulk temperature that the inside
tube wall reached saturation temperature and subcooled nucleate boiling
commenced. The average value of the entering and leaving [343°C (650°F)]

4-2
water heat transfer coefficient was used in the calculation of the overall
heat transfer coefficient.

4.1,1.2. Subcooled and Saturated Nucleate Boiling Regions. The Jens-

Lottes equation (Ref. 4-2),

T
w s P/900
e

was used to determine the water side heat transfer coefficient in the
subcooled nucleate boiling region of the economizer and in the saturated
nucleate boiling region (up to critical heat flux) of the evaporator
section.

From Eq. 4-2, the equivalent heat transfer coefficient, in terms of

wall to saturation temperature difference, was determined and then included
in the overall heat transfer coefficient expression. Because of the high
pressure level, the water side coefficient for nucleate boiling was very
high: more than 113,507 W/m^* °C (20,000 Btu/hr-ft^-°F) for a wall to satura­

tion temperature difference of 0.55°C (1°F). The nucleate boiling coeffi­

cient is at least ten times greater than the other heat transfer coefficients
(helium side, tube wall, fouling) and thus provides negligible thermal
resistance.

4.1.1.3. Critical Heat Flux. The critical heat flux data of Herkenrath
et al. (Ref. 4-3) were used as described in Section 5. In the calculation
procedure, a trial and error method was used. Involving the selection of
various steam qualities, until mass velocity, inside tube heat flux, and
steam quality matched the data of Herkenrath et al. Because of the very
low water mass velocity (M x 10 Ib/hr-ft ), the attendant critical
quality was low (about 5% to 35%) even though the heat fluxes were low
2 o
[about 94,500 W/m (30,000 Btu/hr-ft )]. Such a low critical quality for
the straight tube steam generator is a disadvantage compared with the very
high (about 70%) average critical quality which prevails in a helical coil

4-3
steam generator. The primary reason for the difference in critical quality
between the two types of steam generators is the high water side mass
velocity in the helical unit.

4.1.1.4. Film Boiling Region. In the film boiling region, assumed to

exist from the steam quality at which critical heat flux occurs to 100%
quality, the Miropolskii correlation (Ref. 4-4) was used as it was
developed for the high-pressure [about 20.7 MPa (3000 psi)] steam region.

The film boiling coefficient, h,., , is

lb

k (GD. f p
(4-3)
1 Myg L pi
0.4
where y = 1 x)

In the above expression, c, y, and k are evaluated at the tube tem­

perature whereas P]/Pg is evaluated at saturation temperature. Since the
tube temperature is not known in the calculation, but is the result of
satisfying both the heat transfer conditions and the sum of the individual
heat transfer coefficients, a modified version of the Miropolskii correla­
tion was used. This modification utilized the heat transfer coefficient at
a steam quality of one (all gas), using bulk fluid properties at saturation
temperature and the quantities shown in Eq. 4-3 for all qualities between
the critical quality and a quality of one. In this way, the film boiling
heat transfer coefficient at a quality of one was made identical to that
existing at the start of the superheater section.

4.1.1.5. Superheat Section. The water side heat transfer coefficient in

the superheater section was determined using the Dittus-Boelter equation
(Eq. 4-1). However, since the steam properties change rapidly with
increasing temperature, the film coefficient was determined as a function
of steam temperature. As an example of the influence of steam property

4-4
variation with temperature, the inside heat transfer coefficient decreased
by about a factor of four in going from 367° to 510°C (693° to 950°F) at
20.3 MPa (2950 psia).

The method described above was also used to compute the heat transfer
coefficient on the inside of the resuperheater, which operates at a pres­
sure of 9.1 MPa (1325 psia).

4.1.1.6. Fouling. To account for fouling in the economizer and evaporator

sections, a fouling factor corresponding to a heat transfer coefficient of
11,360 W/m^"°C (2000 Btu/hr-ft^-°F) (inside) was used.

4.1.2. Pressure Drop

4.1.2.1. Single-Phase Region. The friction pressure drops in the single­

phase regions of the steam generator (all liquid in the economizer, all gas
in the superheater and resuperheater) were calculated using the Fanning
smooth tube friction pressure drop equation (Ref. 4-5) as shown below:

2fLG2
AP (4-4)
pgD

4.1.2.2. Two-Phase Region. Pressure drop in the two-phase region of the

evaporator was calculated by use of the Martine11i-Nelson method (Ref.
4-6). Since the heat flux (and steam quality) variation along the heated
length of the evaporator was not uniform, the local two-phase friction mul­
tipliers were determined and then integrated along the length to obtain the
friction pressure drop. The Martinelli-Nelson method (Ref. 4-6) was also
used to determine the two-phase hydrostatic head via the length-average
void fraction. Because of the very low mass velocity in the tubes, fric­
tion and momentum type pressure drops were very small and approximately
three-quarters of the total pressure drop in the main bundle was due to the
all-liquid hydrostatic head in the economizer.

4-5
4.2. HELIUM SIDE

4.2.1. Heat Transfer

The heat transfer coefficient for helium flowing between parallel

straight circular tubes was determined using the Dittus-Boelter equation
(Eq. 4-1) and a correction to account for tube spacing. Using the analysis
of Deissler and Taylor (Ref. 4-7) for the tube geometry (p/d = 1.6) con­
sidered here, the increase in heat transfer coefficient amounted to 11%.
Although the presence of the tube supports creates additional turbulence,
no credit for enhanced heat transfer was taken.

4.2.2. Pressure Drop

Friction pressure drop was computed using the Fanning equation

(Eq. 4-4) and the Deissler-Taylor correction (Ref. 4-7) for tube spacing,
which resulted in an 11% increase in friction factor over that calculated
for an equivalent smooth tube.

The pressure drop associated with tube support-spacers was computed

from Idel'chik’s data (Ref. 4-8) assuming a thick (1 < L/D < 6) perforated
plate. The spacer pressure loss coefficient, K, was found to be sensitive
to the free to frontal area ratio, 6, in that the loss coefficient
decreased from 0.6 to 0.3 when 6 increased from 0.5 to 0.7. When the loss
coefficient is combined with the reduced fluid velocity through the spacer,
an approximately four-fold decrease in spacer pressure drop results when 6
is increased from 0.5 to 0.7. A value of 6 = 0.7 resulted in a reasonable
spacer mechanical configuration and also kept the spacer pressure drop to
an acceptably low level which, on a 1.2m (4 ft) spacing, approximately
equaled the bundle friction pressure drop.

The inlet and exit pressure losses to the steam generator were
approximated by a one velocity head (inlet) loss and a one-half velocity
head (exit) loss, respectively. The total steam generator helium pressure

4-6
drop consisted of the sum of entrance plus exit losses, friction loss, and
tube support spacer loss.

The overall heat transfer coefficient, based on the outside tube

surface area, was determined using the individual heat transfer
coefficients as follows:

U
o 2k 1 1
+
h D ^i h
D In ^ A hf A hi
o D,
o o

To account for uncertainties, 15% excess was added to the required heat
transfer surface as determined from U , above, and a heat balance.
o’

REFERENCES

4-1. McAdams, W. H., Heat Transmission, 3d ed., McGraw-Hill Book Co., New
York, 1954, p. 219.
4-2. Ibid., p. 393.
4-3. Herkenrath, H., "Warmeubergang an Wasser bei Erzwungener Stromung un
Druckbereich von 140 bis 250 Bar," European Atomic Energy Community
Report EUR 3658d, 1967.
4-4. Collier, J. G., Convective Boiling and Condensation, McGraw-Hill Book
Co., London, 1972, p. 226.
4-5. McAdams, op. cit., p. 155.
4-6. Martinelli, R. C., and D. B. Nelson, "Prediction of Pressure Drop
During Forced Circulation Boiling of Water," Trans. ASME 70, 695
(1948).
4-7. Kays, W. M., Convective Heat and Mass Transfer, McGraw-Hill Book Co.,
New York, 1966, p. 183.
4-8. Idel’chik, J. E., "Handbook of Hydraulic Resistance," USAEC
Translation AEC-tr-6630, 1966.

4-7
 5. CRITICAL HEAT FLUX FOR HIGH-PRESSURE WATER

As part of the evaluation of the reference steam generator configura­

tion and alternatives, a unit using straight tubes, instead of helical coil
tubes, was investigated.

One important difference between the helical coil tubes with helium
cross flow and the straight tubes with parallel helium flow is that the
former arrangement permits reasonably high water velocity, whereas the
latter arrangement, using the closest practical tube spacing, results in a
low water velocity. Consequently, the water side heat transfer coefficient
for the straight tube is much lower than that for the helical tube unit and
thus constitutes a much larger portion of the total thermal resistance.
For this reason, the steam quality at which critical heat flux occurs and
film boiling begins, with an attendant reduction in heat transfer coeffi­
cient , is of importance in determining the required evaporator length.

Furthermore, significant differences in boiling behavior are identi­

fied between straight tubes, where critical heat flux generally occurs at
one axial location on the entire tube periphery, and helical coil tubes,
where critical heat flux can occur over a substantial portion of the tube
length in different parts of the tube cross section.

5.1. STRAIGHT TUBES

While numerous correlations for critical heat flux in water exist,

they are generally applicable to the pressure range below 15.2 MPa (2200
psia). Actually, the available critical heat flux data for the pressure
range of interest for GCFR — about 20.7 MPa (3000 psia) — are not only
limited but are often most applicable to fossil-fired steam tubes with par­
tial heating of the tube periphery (Ref. 5-1). For this reason, the exten­
sive critical heat flux data of Herkenrath et al. (Ref. 5-2), at a pressure

5-1
of 20.5 MPa (2972 psia), were selected as a basis because they cover the
tube diameter range used in the GCFR, have long heated lengths, and require
only moderate extrapolation to cover the heat flux and mass velocity range
of the GCFR. One important difference between the GCFR operating condi­
tions and those under which the critical heat flux data (Ref. 5-2) were
taken is that the data were obtained with a uniform heat flux using elec­
trical resistance heating, whereas in the GCFR a fixed overall temperature
difference exists at a given location in the heated length. This differ­
ence in the method of heat application may affect the magnitude of the
critical heat flux. Indeed, Campolunghi et al. (Ref. 5-3) have shown that
for Freon 12 covering a wide range of operating pressure to critical pres­
sure ratio and for very long heated lengths, the critical heat flux for
indirect heating can be substantially greater than that for direct heating.
If it is assumed that similar behavior exists for water, then the critical
heat flux-quality characteristics, described below, can be considered con­
servative in that higher heat fluxes can be expected for the indirect
heating method actually used in operation.

To determine the effect of reduced mass velocity on critical heat

flux, the data of Herkenrath et al. (Ref. 5-2) were evaluated for GCFR con­
ditions. Figure 5-1 shows critical steam quality as a function of critical
heat flux and mass velocity at a pressure of 20.5 MPa (2972 psig). It may
be noted that critical quality increases with mass velocity at a given heat
flux, whereas at a given mass velocity, critical heat flux decreases as
quality increases. The curves shown for mass velocities of less than 700
2 6 2
kg/m •s (0.52 x 10 Ib/hr-ft ) are extrapolations of the data. Because of
the great difference in mass velocity between the coiled tube and straight
tube units, the critical quality will be about 70% for the coiled unit and
range from 6% to 27% for the straight unit.

The consequence of this apparent disadvantage for the straight tube is

moderate because the favorable properties of steam at 20.3 MPa (2950 psia),
i.e., very high Prandtl number, result in film boiling heat transfer coef­
ficients which are generally larger than the helium heat transfer coeffi­
cient. For this reason, the decrease in overall heat transfer coefficient
is modest when nucleate boiling ceases and film boiling prevails.

5-2
 0.90

CRITICAL STEAM QUALITY

,

TUBE OIA. = 0.39 IN. (9.9 mm) AND 0.79 IN. (20.0mm)

TUBE LENGTH = 9.8 FT (2.99 m), 16.7 FT (5.09 m),

AND 29.9 FT (9.11 m)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

(0) (100) (200) (300)

Q/A, HEAT FLUX BTU/HR FT2 X 105 (KW/M2)

Fig. 5-1 Critical quality versus heat flux (straight tube) [based on data
of Herkenrath et al. (Ref. 5-2)]

5-3
 The use of turbulence promoters, such as twisted tapes or ribs in the
tube walls, to extend the nucleate boiling region to very high qualities
has received considerable attention. The major impetus for the use of such
devices derives from the desire to prevent overtemperature failure (burn­
out) of water tubes in fossil-fired boilers. If the critical heat flux
occurs at a low steam quality, the film boiling coefficient will be at a
near minimum value. Even though the heat flux decreases at this time (with
indirect heating and a fixed overall temperature difference), the very
large decrease in water side heat transfer coefficient associated with the
change from nucleate boiling to film boiling causes the tube wall tempera­
ture to rise appreciably. Since the film boiling heat transfer coefficient
increases with steam quality, it follows that the tube wall temperature
rise will be much less if critical heat flux occurs at a high steam quality
than if it occurs at a low steam quality.

Probably the most widely applied and best known method for improving a
critical quality is the helically ribbed tube concept described by Swenson
et al. (Ref. 5-4). At a pressure of 20.7 MPa (3000 psia) and a mass veloc-
2 5 9
ity of 951 kg/m •s (7 x 10 lb/hr-ft ), the helical ribs on the inside tube
wall improved the steam critical quality from about 3%, for smooth tubes,
to over 90%. Such ribbed tubes have been successfully used in commercial
once-through fossil-fired Babcock & Wilcox steam boilers for many years. A
detailed description of the experiences accumulated with ribbed steam gen­
erator tubes over a 13-yr period, encompassing the areas of heat transfer
performance, corrosion, and deposition behavior, is provided by Lux and
Weick (Ref. 5-5).

More recent investigations of the effectiveness of ribbed surfaces in

extending the nucleate boiling range for water are described by Nishikawa
et al. (Ref. 5-6) and by Watson et al. (Ref. 5-7).

Evaluation of the data described above, for possible use in a straight

tube GCFR steam generator, led to the conclusion that use of a nucleate
boiling regime extender device is not justified for the following reasons:

5-4
 1. To accommodate the ribs, the tube wall thickness must be
increased by an amount equal to the rib height. If the ribs are
placed only in the evaporator section, a step in tube wall
thickness with two additional tube welds is required. Without
the ribs, the tube can be a single weld-free piece.

2. Watson et al. (Ref. 5-7) demonstrated that the improvement in

critical heat flux due to ribs is marginal for the low mass
2 5 2
velocity range [less than 475 kg/m •s (3.5 x 10 Ib/hr-ft )] in
which a straight tube GCFR steam generator operates.

3. Since the evaporator length is much less than the economizer or

superheater length, the evaporator length reduction achievable by
the use of internal ribs is only about 6% of the total steam gen­
erator length.

5.2. HELICAL COIL TUBES

In the previous section, the critical heat flux and steam quality for
water flowing inside straight circular tubes were discussed. In this sec­
tion, some of the important differences between straight circular tubes and
helical coil circular tubes relative to critical heat flux behavior are
presented.

Although data are limited, critical heat flux for water flowing in
helical coil circular tubes has been investigated. Miropolskii et al.
(Ref. 5-8) initially covered the water pressure range from 290 to 4278 psi
(2 to 29.5 MPa); in a later investigation, Miropolskii and Pikus (Ref.
5-9) covered the pressure range from 1421 to 3118 psia (9.8 to 21.5 MPa)
with most of the data obtained at 1421 psia (9.8 MPa).

Both of the above investigations showed that critical heat flux

occurred first on the inner tube wall (i.e., the surface nearest the coil
center line), and that the heat transfer coefficient on the inner wall, for
all-liquid or superheated steam conditions, was lower than that on the

5-5
outer wall or that for straight tubes. Furthermore, the critical heat flux
for coiled tubes was found to be lower for subcooled conditions and greater
for net steam quality conditions, as compared with the critical heat flux
for straight tubes.

While the above investigations dealt with the gross behavior of criti­
cal heat flux in coiled tubes, Carver et al. (Ref. 5-10) examined the local
behavior and distribution of critical heat flux in coiled tubes. As the
latter investigation is the most revealing and has the most significant
implications for GCFR steam generators, it has been examined and is
described below. However, to place these findings in perspective, and to
show the possible degree of applicability to GCFR steam generators, the
following areas should be considered:

Similarities: (1) The inner and outer coil radii of the test closely
approximate those expected for the GCFR; (2) the steam pressure level
of the test [17.9 MPa (2600 psia)] is close to the expected pressure
level [20.3 MPa (2950 psia)] of the GCFR; (3) the test tube inside
diameter is less than, but similar to, that expected in the GCFR;
(4) the test mass velocity range covers much of the GCFR operating
range but not the extremely low flow range encountered during reactor
shutdown.

Differences: (1) The test heat flux was constant, whereas in the GCFR
the evaporator heat flux is expected to increase significantly with
length; (2) the lowest test heat flux is about twice as great as that
expected in the GCFR evaporator at full flow conditions; thus, tube wall
temperature increases for GCFR will be a small fraction of those shown.

In general, when critical heat flux occurs with forced flow and net
steam quality inside a straight tube, it is located on the entire tube
periphery at one position in the heated length. In contrast, according to
Carver et al. (Ref. 5-10), for helical coil circular tubes the critical
heat flux occurs first in one quadrant of the tube at a particular point in
the heated length (nucleate boiling continues in the other quadrants) and

5-6
then occurs successively in other quadrants as the heated length increases.
Evidently, the local steam quality in a quadrant in which critical heat
flux occurs is higher than the cross-sectional average steam quality.

Figure 5-2 (from Ref. 5-10) shows the variation of tube temperature
with steam quality or, equivalently, with heated length in the evaporator
section as heat flux is constant. Note that critical heat flux occurs at
cross-sectional average steam qualities varying from about 10% to 80% or,
in other words, along 10% to 80% of the evaporator section heated length.*
For a straight tube the critical heat flux would occur at a quality of
about 33%. Of course, once critical heat flux occurs at a given location,
transition or film boiling exists from thereon. Thus, for the relatively
low mass velocity conditions shown [679 kg/m •s (0.5 x 10 Ib/hr-ft )],
critical heat flux for a 1.65 m (65 in.) radius coiled tube occurs at quali­
ties both above and below that for a straight tube. However, at mass
2
velocities in the range of GCFR steam generators [about 2038 kg/m • s (1.5 x
6 2
10 Ib/hr-ft )], the critical quality for a coiled tube is not only greater
than that for a straight tube on the average but also on a local basis.
Furthermore, the rate of tube temperature increase, for otherwise similar
conditions, tends to be less for a coiled tube than for a straight tube.
This more desirable behavior for the coiled tube is probably due to the
conduction of heat around the tubs circumference from the critical heat
flux location to the nucleate boiling regions on the rest of the tube
periphery. Support for this explanation is seen in Fig. 5-2 at a steam
quality of 75%, where the sudden increase in tube temperature for the
hotter locations (270°, 330°) largely coincides with the rapid increase in
tube temperature (onset of a critical heat flux) for the cooler locations
(90°, 180°) which up to this point experienced nucleate boiling.

Figure 5-3 (from Ref. 5-10) provides some perspective on the differ­
ences in critical heat flux behavior that can result from a helical coil
diameter variation which closely resembles that anticipated in the GCFR

*The peripheral tube temperature variation accompanying this type of

behavior and the resultant thermal stresses have been evaluated for design
conditions in HTGR steam generators (Ref. 5-11).

5-7
 PRESSURE = 17-9 MPa (2600 PSIA)
MASS FLOW = 679 KG/S-M2 (500,000 LB/HR-FT2)
I-D. HEAT FLUX = 39^ KW/M2 (125,000 Btu/HR-FT2)
COIL RADIUS = 1.65 M (65 IN.)
537-7

330 DEG
o = 180 DEG
510.0 270 DEG 330 DEG
COIL
90 DEG CENTER
90 DEG
270 DEG
482.2 180 DEG
TEMPERATURE [0C(°F)]

INSIDE SURFACE TEMPERATURE

426.6

371 .0

FLUID TEMPERATURE

315.5
(600)-20
STEAM QUALITY X {%)

Fig. 5-2. Temperature versus steam quality (helical coil)

5-8
 PRESSURE = 2600 PSIA (17.9 MPa); MASS FLOW = 500,000 LB/HR-FT2 (679 KG/m2 • S);
NOMINAL ID. HEAT FLUX = 125,000 BTU/HR-FT2 (394 KW/m2)
□ 16-IN.-RADIUS COIL (0.4 m)
O 65-IN.-RADIUS COIL (1.65 m)

0°

COIL
CENTER'—' ~ 270° — 90°
LINE

TUBE CROSS SECTION

100
FILM BOILING
90

80

70

60

50

40

30

20

10

0
0 60 30 0 330 300 270 240 210 180 150 120 90

ANGLE AROUND TUBE, 9 (DEGREES)

Fig. 5-3. Steam quality versus angle (helical coil)

5-9
design. The figure shows steam quality as a function of tube angle for
coil radii of 0.4m (16 in.) and 1.65 m (65 in.) and for a straight tube,
for the same flow and heat flux conditions as shown in Fig. 5-1. It can be
seen that the critical steam quality is always greater for the smaller coil
radius, with the most pronounced difference occurring in the 330° region
(upper left quadrant). For the lower heat fluxes which will exist under
GCFR conditions, the critical steam qualities for both radii will be
greater than those shown.

Figure 5-2 shows that the upper left quadrant (270° to 360°) experi­
ences critical heat flux at the lowest steam quality, whereas the lower
right quadrant (90° to 180°) experiences critical heat flux at the highest
steam quality. This non-uniform critical heat flux behavior is apparently
due to centrifugal effects which force water to the lower right quadrant at
the expense of the upper left quadrant. However, according to Ref. 5-4,
secondary flow within the tube is such that the critical heat flux
increases as mass velocity increases, thus duplicating the behavior of
straight tubes. At low mass velocities, buoyancy tends to produce a higher
mixture quality in the upper portion of the tube than at the bottom and
thus reinforces the effect of centrifugal force.

Note that at a cross-sectional average quality of 100% and beyond

(superheat), the tube wall temperatures for the locations in which critical
heat flux occurred first (330°, 270°) are still greater than those for the
locations (90°, 180°) in which critical heat flux occurred last. This
could be due to one of two things:

1. A thermal non-equilibrium condition in which the steam quality in

the 90° and 180° positions is actually less than 100%.2

2. A significantly lower heat transfer coefficient on the inner wall

(270° and 330°) locations compared with the outer and lower wall
locations for the condition where superheated steam exists in the
entire tube cross section.

5-10
 Clearly, the non-uniform boiling behavior of a coiled tube is a char­
acteristic that requires further investigation as to possible effects on
tube thermal stress, material fatigue life, and tube corrosion behavior.

In addition, in the extremely low flow operation mode (about 2%)

required by the GCFR steam generator during the safe shutdown procedure,
buoyancy effects may result in stratification of the flow into two-phase
and liquid regions in the first part of the evaporator and in separation
into high and low steam quality regions in the last part of the evaporator.
Such behavior is not surprising if it is realized that a helical coil tube
with a vertical longitudinal axis is, because of the very small rise angle
(^3° in the GCFR), essentially a horizontal tube. Because of these differ­
ences , it is possible that the low flow boiling stability characteristics
«
of coiled tubes also differ from those of straight vertical tubes. This
subject is discussed in Section 6.

REFERENCES

5-1. Schmidt, K. R., "Warmetechnische Untersuchungen an hock belasteten

Kesselheizflachen," Mitteilungen der Vereinigung der Grosskessel-
bezitzer, December 1959, p. 391.
5-2. Herkenrath, H. P., et al., "Warmeubergang an Wasser bei Erzwungener
Stromung un Druckbereich von 140 bis 250 Bar," European Atomic
Energy Community Report EUR 3658d, 1967.
5-3. Compolunghi, F., et al., "Burn Out Power in Once-Through Tubular
Steam Generators," Proceedings of the Fifth International Heat
Transfer Conference, September 3-7, 1974, Tokyo, Vol. IV.
5-4. Swenson, H. S., J. R. Carver, and G. Szoeke, "The Effects of
Nucleate Boiling Versus Film Boiling on Heat Transfer in Power
Boiler Tubes," J. Eng. Power, ASME Trans., Ser. A, 84, 365 (1962).
5-5. Lux, J. A., and R. H. Weick, "Corrosion and Deposition Experience
with Internally-Ribbed Tubing in High Pressure Utility Boilers,"
Proceedings of the American Power Conference, Vol. 34, 1972, p. 672.

5-11
5-6. Nishikawa, K., et al., "Flow Boiling Crisis in Grooved Boiler
Tubes," Proceedings of the Fifth International Beat Transfer
Conference, September 3-7, 1974, Tokyo, Vol. IV.
5-7. Watson, G. B., R. A. Lee, and M. Weiner, "Critical Heat Flux in
Inclined and Vertical Smooth and Ribbed Tubes," Proceedings of the
Fifth International Heat Transfer Conference, September 3-7, 1974,
Tokyo, Vol. IV.
5-8. Miropolskii, Z. L., V. Yu. Pikus, and M. E. Shitsman, "Regimes of
Deteriorated Heat Transfer at Forced Flow of Fluids in Curvilinear
Channels," Third International Heat Transfer Conference, August
1966, Chicago, Vol. II.
5-9. Miropolskii, Z. L., and V. Yu. Pikus, "Critical Boiling Heat Fluxes
in Curved Channels," Heat Transfer - Soviet Research, Vol. 1, No. 1,
January 1.59.
5-10. Carver, J. R., C. R. Kakarala, and J. S. Slotnik, "Heat Transfer in
Coiled Tubes with Two-Phase Flow," USAEC Report TID-20983, Babcock &
Wilcox, July 31, 1964.
5-11. Lasarev, M., General Atomic Company, personal communication, July
1976.

5-12
 6. LOW FLOW BOILING STABILITY

The boiling behavior of once-through steam generators operating well

below maximum design flow is an area of concern since steam flow, pressure,
and temperature fluctuations, or instabilities (static or dynamic), may
accompany such operation. Typically, a steam generator in a power plant is
required to operate over the flow range of 20% to 100%. Stable boiling
behavior over this flow range is generally obtained by providing sufficient
additional pressure drop in the all-liquid inlet region by means of
orifices.

However, as part of the planned shutdown procedure for the GCFR, it is

necessary that the helium circulator turbine be supplied with high-quality
or superheated steam from the steam generator, using reactor decay heat,
until steam is available from auxiliary boilers. In order to fulfill this
unique requirement, it is necessary for the steam generator to operate in a
stable and predictable manner at steam flows as low as 2% of full flow for
approximately 30 min. Furthermore, during the first minute of shutdown,
the feedwater flow will be stopped, thereby depleting the water inventory
until the transition to the shutdown feed pumps is accomplished.

Since proper operation of the steam generator under these extremely

low steam flow conditions is an integral part of the safe shutdown proce­
dure, it is essential that the required steam generator perfomance be
confirmed. Accordingly, as a first step, an evaluation has been made of
literature pertinent to the low flow boiling stability of water for both
straight and helical coil tubes.

6.1. STRAIGHT TUBES

As part of the liquid metal fast breeder reactor (LMFBR) steam genera­
tor program, extensive development and testing are to be performed in the

6-1
areas of materials, structural behavior, and thermal-hydraulic performance
(Ref. 6-1). Of particular significance to the GCFR steam generator program
is the already completed 4000-hr test conducted on the Atomics Interna­
tional modular steam generator (Ref. 6-2). This sodium-heated unit of
hockey-stick design consisted of 158 16 mm (5/8 in.) diameter tubes about
21 m (69 ft) long and was constructed of 2-1/4 Cr-1 Mo steel. While the
full-power performance, material, and structural behavior met all design
objectives, the observed low flow stability characteristics of the unit are
of singular interest to GCFR steam generators since these characteristics
are directly related to the reactor safe shutdown procedure. Table 6-1
shows the water pressure-flow range (down to 0.5%) covered in these tests.

According to Ref. 6-2, no water side instabilities were predicted to

occur for the conditions covered in the low flow stability tests, and none
were observed. However, the previously predicted (Ref. 6-3) two-mode (exit
steam wet or superheated) thermal and pressure drop behavior was observed.
According to the analysis (Ref. 6-3), at low flow the major portion of the
steam generator surface does not transfer heat. Experimental verification
of this behavior is described below.

Figure 6-1 (from Ref. 6-2) shows the two modes of behavior, which, for
identical water pressure and flow conditions, depend on sodium flow rate.
That is, increasing sodium flow rate beyond a certain value will cause a
change from one mode of operation to the other. In the first mode (upper
portion of the figure), where the steam exit quality is one or less, the
water is brought up to saturation temperature in the initial portion of the
steam generator. This is followed by a very long no-heat-transfer section
where the sodium and water temperatures are identical to water saturation
temperature. In the last section, the water is boiled to a steam quality
of one or less. When the sodium flow rate is increased, the second mode of
operation results, as described below.

In the second mode (lower portion of the figure), in which the exit
steam is superheated, the water is brought up to saturation, boiled to a

6-2
 TABLE 6-1
RANGE OF OPERATING CONDITIONS FOR LOW FLOW BOILING STABILITY TESTS
(STRAIGHT TUBE)(a)

Water Pressure Flow Rate

[MPa (psxa)] (% of full flow)

7.6 (1100) 0.5 to 10.5

4.9 (710) 1.0 to 19.0

3.5 (510) 0.5 to 1.6

^From Ref. 6-2.

6-3
 WET STEAM

SODIUM
TEMPERATURE [ 0F(°C}j

/WATER/STEAM
NOMINAL CONDITIONS
SODIUM FLOW = 0.32 X 10° LB/HR (0.145 X 10 KG/HR)
WATER FLOW = 10,100 LB/HR (4590 KG/HR)
PRESSURE = 1000 PSIA (6.89 MPa)

ACTIVE HEAT TRANSFER LENGTH [ FT (m)l

650

SUPERHEATED STEAM

SODIUM

I WATER/STEAM

NOMINAL CONDITIONS
SODIUM FLOW = 0.52 X 10° LB/HR (0.236 X 10° KG/HR)
WATER FLOW = 10,100 LB/HR (4590 KG/HR)
FEEDWATER PRESSURE = 1000 PSIA (6.89 MPa)

ACTIVE HEAT TRANSFER LENGTH [ FT (m)]

Fig. 6-1. Temperature versus length (heat transfer mode change) (from
Ref. 6-2)

6-4
quality of one, and superheated in the initial portion of the steam genera­
tor. This is followed by a no-heat-transfer region, for the rest of the
length, in which the sodium and superheated steam temperatures are identi­
cal to the sodium inlet temperature.

Since for each mode of operation friction and momentum type pressure
losses are very small compared with hydrostatic head, it can be seen (from
Fig. 6-1) that the total pressure drop for mode one (saturated water in the
no-heat-transfer section) will be greater than that for mode two in which
the no-heat-transfer portion contains superheated steam. Thus, according
to Fig. 6-1, when a change from mode one to mode two occurs, an increase in
tube temperature, from water saturation temperature to sodium inlet temper­
ature , will be accompanied by a substantial decrease in overall pressure
drop.

In order for a group of parallel tubes in which boiling is occurring

to have identical thermal and hydraulic characteristics, it is necessary
that the flow rate in each tube, and the heat flux distribution along each
tube, be identical. Since this condition is rarely achieved in practice,
some differences in thermal-hydraulic behavior between tubes can be
expected. In the test described in Ref. 6-2, it was found that for some
tubes the transition from mode one to mode two was very gradual with an
essentially linear increase of steam outlet temperature from saturation
temperature to sodium inlet temperature. In other tubes, the outlet steam
temperature generally corresponded to the sodium inlet temperature with an
occasional decrease to saturation temperature. Since these transitions
occurred gradually, over a several-minute period, the possibility of damage
to the steam generator was considered minimal.

6.2. HELICAL COIL TUBES

Only limited data are available on low flow boiling stability in heli­
cal coil tubes. Shultz (Ref. 6-4) describes past and present investiga­
tions in this area as related to HTGR steam generator development and
requirements, and these investigations are discussed below.

6-5
 The initial boiling stability test was designed to model a Fort St.
Vrain steam generator tube and, with subsequent modifications, was used for
HTGR steam generator low flow stability investigations. The test section
consisted of a single coil, 0.76 m (30 in.) in diameter and 45.7 m (150 ft)
long, with a 10 mm (0.40 in.) inside tube diameter. The range of test -
conditions was as follows: steam pressure, 8.3, 12.4, 16.6 MPa (1200, 1800,
2 6 6
2400 psia)| mass velocity, 81.5 to 815 kg/m -s (0.06 x 10 to 0.6 x 10
2
Ib/hr-ft ) (as low as 5% of full flow); feedwater inlet temperature, 149°C
(300°F); superheater outlet temperature, 510°C (950°F). At 16.6 MPa (2400
psia), no instabilities were observed even with very large exit (two-phase)
pressure losses; at 12.4 and 8.3 MPa (1800 and 1200 psia), instabilities
were observed. Comparison of the experimental data, for both stable and
unstable conditions, with the predictions of the dynamic stability code
DYNAM (Ref. 6-5) was satisfactory, with the code generally predicting a
greater required inlet orifice pressure drop for stable operation than that
observed. This initial test was relatively simple, with minimum instrumen­
tation and without means to determine the heat flux distribution along the
tube. To overcome these shortcomings, a much improved test, conducted in
the Zebulon Loop (France, CEA), was devised and performed.

The objectives of the Zebulon boiling stability tests were to verify

the heat transfer and fluid flow correlations, the pressure drop analysis,
and the analyses performed by the codes SUPERHEAT (static instability) and
DYNAM (dynamic instability). The major improvements in the test consisted
of more extensive instrumentation, accurately controlled heat input, four
parallel channels, and measurement of pressure drop and heat transfer quan­
tities. The Zebulon test section was constructed of Incoloy 800 tubes
which were heated on the outside surface by sodium. The water/steam flow
was upward, while the sodium flow was downward. Four parallel coils, with
an inside tube diameter of about 20 mm (0.80 in.) and diameters of 0.63,
0.81, 1.8, and 2.7 m (24.8, 31.9, 70.9, and 106.3 in.), respectively, com­
prised the test section. The economizer-evaporator length was about 48.5 m
(159 ft), and the superheater length was about 14.6 m (48 ft). The test was
operated over a pressure range of 4.0 to 18 MPa (580 to 2610 psia) and a
2 6 2
mass velocity range of 300 to 1770 kg/m •s (0.22 to 1.3 x 10 Ib/hr-ft ).

6-6
 Tests have been completed covering both dynamic instability and static
instability (for single coils as well as four coils in parallel), and com­
parisons have been made with the predictions of the DYNAM code and LOOP
code. A final report on this work is to be issued during 1977.

A third test program, involving the use of a full-scale HTGR multitube

test section, is to be conducted in the French CEA Carmen-2 loop. The pur­
pose of this test is to demonstrate the boiling stability and performance
characteristics of a full-scale steam generator tube configuration and to
compare the results with predictions of steady-state performance and with
the SUPERHEAT (static instability) and DYNAM (dynamic instability) codes.

The Carmen-2 test section consists of eight 1.8 m (71 in.) diameter
helical coils in parallel, with a main bundle length of 101.8 m (334 ft) and
entrance and exit lengths of 14.2 and 25.6 m (46.5 and 84 ft), respectively.
In constrast to the Zebulon test, which used sodium as the heating fluid,
the Carmen test will use high-pressure, high-temperature helium as the
heating fluid. Construction of the test section is in progress and the
final test plan is being prepared. Testing is scheduled to start in 1978.

The results of the previously described straight tube low-flow boiling

stability tests are encouraging if they are applied to the required low-
flow performance in GCFR steam generators,* However, it should be remem­
bered that the considerable differences in single-phase heat transfer,
pressure drop in coils of different diameter, and boiling behavior (see
Section 5) between straight vertical tubes and coiled tubes may also be
reflected as significant differences in low-flow boiling stability charac­
teristics . For this reason, the Zebulon loop HTGR boiling stability
results for helical coils will be evaluated, and pertinent portions will
be utilized in the formulation of the GCFR low flow-boiling stability test
program.

*According to Profos (Ref. 6-6), a helical coil steam generator for

use in a gas-cooled nuclear reactor plant was designed to operate at flows
as low as 2%.

6-7
REFERENCES

6-1. "The Clinch River Breeder Reactor Project: A Briefing for

Engineers," Proceedings of the Breeder Reactor Corporation, April
1975 Information Session.
6-2. Harty, R. B., "Module Steam Generator Final Project Report," Atomics
International Report AI-TR-097-330-010, September 1974.
6-3. Baroczy, C. J., "Thermal-Hydraulic Analysis of Al Modular Steam
Generator Under Low Flow-Low Pressure Operating Conditions," Atomics
International Report AI-TI-097-330-006, February 23, 1973.
6-4. Schultz, G. M., "Large HTGR Steam Generator Design Verification and
Support Programs - Heat Transfer, Fluid Flow, and Stability," Gulf
General Atomic Report Gulf-GA-A12808, Vol. II, December 15, 1973.
6-5. Efferding, L. E., "DYNAM, A Digital Program for Study of the Dynamic
Stability of Once-Through Boiling Flow with Steam Superheat," Gulf
General Atomic Informal Report GAMD-8656, June 26, 1968.
6-6. Profos, 0., "The New Steam Generating System for the French Nuclear
Power Station EL-4," Sulzer Technical Review Special Number, Nuclex
69 (1969).

6-8
 7. STRUCTURAL EVALUATION

7.1. GENERAL CONSIDERATIONS

The stress evaluation of the GCFR steam generator is a complex area

involving materials, temperature, pressure, structural support and design,
tube geometry, and the flow conditions of helium and wacer/steam. Addi­
tional complexity is introduced by the interdependence of many of these
areas. For example, selection of the tube geometry affects the tube sup­
port plate thickness, the tube bundle and shell length, and the vibratory
characteristics of the tubes. In a similar fashion, the selection of a
particular tubing material can significantly influence not only the above-
mentioned areas but the tube to tube support stress levels as well. Thus,
the stress evaluation of the steam generator involves a process of tube
geometry selection (for thermal-hydraulic performance) and evaluation of
its compatibility with other steam generator components with regard to
stress and structural requirements. This process is repeated until stress
limits, component life, and structural requirements are met for all oper­
ating conditions.

The GCFR steam generator will be designed to meet the requirements of

the ASME Boiler and Pressure Vessel Code, Section III (Nuclear Power Plant
Components), Class 1, and ASME Code Case 1592, "Class 1 Components in
Elevated Temperature Service, Section III." Steam generator parts governed
by these criteria include the main bundle tubing, resuperheater tubing,
tubesheets, feedwater and steam pipes, and primary closures.

The support structure for the steam generator will be designed to meet
the requirements of the ASME Code, Section III, Subsection NF, "Component
Supports," or its intent, using Code Case 1592 criteria, if the part is not
covered. Included in this grouping are the following parts: tube support

7-1
plates or straps, support gussets, support flanges, shrouds, flow restric­
tor, restraining parts, and internal baffles.

Application experience with the ASME Code for high-temperature heat

exchangers, such as the HTGR steam generator, is described by Pai et al.
(Ref. 7-1). Difficulty in the application of Subsection NF to HTGR steam
generators, which operate at high temperature, was experienced because
material creep was not considered in the Code. For consistency, the
philosophy of Subsection NF was combined with modified Code Case 1592
criteria in order to provide appropriate component support classifications.

Jakub (Ref. 7-2) has reviewed Section III of the ASME Code as applied
to HTGR steam generators and, in order to overcome existing shortcomings or
omissions, has recommended that additional Code rules be made to cover the
areas of heat exchanger internal structures, differentiation of critical
and non-critical items, alternate in-service inspection methods for tubing
and structural items inside the outer pressure boundary, elevated-
temperature structures, and for specific methods to address corrosion,
cavitation, and vibration.

An assessment of the structural performance of the HTGR steam genera­

tor has been made by Gwaltney (Ref. 7-3). Included in this preliminary
structural assessment were vibration; seismic, elastic, and inelastic
loadings; materials; and analysis methods. Much of this work will be
applicable to the design of the GCFR steam generator, which, because its
maximum helium temperature is considerably lower than in the HTGR, is
expected to have more readily solvable structural design problems. A
discussion and evaluation of many of the aspects involved in the structural
design of the GCFR steam generator are given below.

7.2. MATERIALS

The selection of materials for use in the GCFR will draw not only on
the knowledge accumulated from the design, fabrication, and operation of
the Peach Bottom and Fort St. Vrain HTGRs, but also on the extensive

7-2
material development and testing programs associated with the HTGR and
those now in progress under the Clinch River breeder reactor project and
foreign LMFBR projects.

As a result of the considerably lower maximum temperatures in the GCFR

[ (helium - 550°C (1022°F) , tubing *= 521 °C (970°F) ] compared with those in
the HTGR [helium - 741°C (1366°F), tubing “ 705°C (1310°F)], greater lati­
tude in the choice of structural and tubing material is possible. Because
Incoloy 800 (for high-temperature regions) and 2-1/4 Cr-1 Mo steel (for
lower-temperature regions) have been of prime interest for HTGR steam gen­
erators, the mechanical properties of these materials and the effects of
primary and secondary coolant on them have been well established. The
demonstrated low tritium permeability of Incoloy 800, its high-temperature
strength, and its excellent corrosion resistance were the basis for its
selection as the reference material for GCFR steam generator tubing. How­
ever, the GCFR steam generator tube temperatures are sufficiently low so
that 2-1/4 Cr-1 Mo steel could be used with a significant attendant cost
savings. The numerous considerations entering into the choice of a steam
generator tubing material are discussed below.

Included in this material evaluation were the following areas:

1. Materials presently used, and those contemplated for future use,

in HTGR steam generators.

2. ASME Code material requirements and allowable stresses.

3. Material cost.

4. Permeability of metals to tritium.

5. Corrosion.

7-3
7.2.1. HTGR Steam Generator Materials

In the construction of HTGR steam generators, Incoloy 800, 2-1/4 Cr-1

Mo steel, Type 304 stainless steel, and low-carbon steel are used for the
various heat transfer and structural portions. The lower metal temperature
[up to 527°C (980°F)] heat transfer tubing and structural parts utilize
2-1/4 Cr-1 Mo steel; the higher metal transfer [up to 838°C (1540°F)] heat
transfer tubing and structural parts use Incoloy 800 and Type 304 stainless
steel. Mazandarany and Rittenhouse (Ref. 7-4) have reviewed these mate­
rials, with particular emphasis on the behavior of Incoloy 800 and 2-1/4
Cr-1 Mo steel. Table 7-1 (from Ref. 7-4) shows the materials and expected
temperature range in the various sections of an HTGR steam generator.

As one part of an effort to reduce the cost of HTGR steam generators,

the feasibility of substituting other materials for those presently used is
being investigated. Some of the avenues of investigation include:

1. The possible use of high-chrome alloys such as Type 420 stainless

steel (12 Cr) and Type 422 stainless steel (12 Cr, 1 Mo, 1 W, 0.8
Ni, 0.25 V) in place of the presently used 2-1/4 Cr-1 Mo steel.
It is anticipated that the higher allowable stress for the high-
chrome alloys in combination with the smaller corrosion allow­
ance, compared with 2-1/4 Cr-1 Mo steel, will make this substitu­
tion economically attractive. As a longer-term goal, the high-
chrome alloys may, if proven satisfactory by tests, serve as a
replacement for the relatively expensive Incoloy 800 in the high-
temperature sections.

2. The use of Type 304 or 316 stainless steel instead of Incoloy 800
in the high-temperature structural portions of the steam genera­
tor and possibly in the superheater and reheater tubing sections.
If feasible, this material substitution can result in a cost
saving.

7-4
 TABLE 7-1
HTGR STEAM GENERATOR MATERIALS AND PROJECTED MAXIMUM TEMPERATURES
(FROM REF. 7-4)

Material and Temperature

Component ASME Specification [°F (°C)]

Reheater
Tube bundle Incoloy 800H, SB-1 63 1310 (710)
Tubesheet and tubesheet Incoloy 800H, SB-408 1000 (538)
extension (hot)
Tubesheet and tubesheet Low-alloy steel, 640 (338)
extension (cold) SA-508
Support structure Incoloy 800H, SB-409 1350/1540(a)
(732/838)
Superheaters
No. 1 - tubing 2-1/4 Cr-1 Mo, 980 (528)
SA-213, T22
No. 2 - tubing Incoloy 800H, SB-1 63 1065 (575)
No. 3 - tubing (upper and Incoloy 800H, SB-1 63 1250 (677)
lower)
Tube support plate (upper Incoloy 800H, SB-409 1250 and 750
and lower) (677 and 399)
Tubesheet Incoloy 800H, SB-408 950 (510)
Tubesheet support Incoloy 800H, SB-408 950 (510)
Tubesheet extension Incoloy 800H, forged 950 (510)
piping
Economizer-evaporator
Economizer tubing 2-1/4 Cr-1 Mo, 720 (382)
SA-213, T22
No. 1 evaporator tubing 2-1/4 Cr-1 Mo, 850 (454)
(lower) SA-213, T22
No. 2 evaporator tubing 2-1/4 Cr-1 Mo, 900 (482)
(upper) SA-213, T22
Shroud and thermal barrier
cover plates
EES inner shroud 2-1/4 Cr-1 Mo, 712 (378)
SA-387-D
Incoloy 800H, SB-409 1275 (690)
EES outer shroud Type 304 stainless steel 650-1270
(343-688)
Top-head outer shroud Incoloy 800H, SB-409 1275 (690)
Thermal barrier cover Incoloy 800H, SB-409 1416 (769)
plates Type 304 stainless steel 1200 (649)
Carbon steel 660 (349)

(a)Temperature excursions of up to 1540°F (G38°C) during power cycles.

^■^Economizer-evaporator, superheaters 1 and 2.

7-5
7.2.2. ASME Code Requirements

As part of the GCFR design criteria, it is required that the steam

generator conform to the requirements of the ASME Code, Section III
(Nuclear Power Plant Components), Class 1. For service above 427°C
(800°F), the rules governing permissible materials and allowable stresses
are contained in ASME Code Case 1592 (April 29, 1974), "Class 1 Components
in Elevated Temperature Service, Section III."

7.2.2.1. Permissible Materials. ASME Code Case 1592 lists the permissible
materials, which include Type 304 and Type 316 stainless steel, 2-1/4 Cr-1
Mo steel, and Incoloy 800H (Ni-Fe-Cr). This list differs from that pre­
viously issued (ASME Code Case 1331-8) on the same subject in that Inconel
625 has been omitted and the allowable stresses for Incoloy 800 (now iden­
tified as Incoloy 800H) have been increased, provided that the material
meets certain strength, composition, and grain size criteria.

From the above information it can be seen that serious consideration

of a substitute material, such as Type 422 stainless steel, for the 2-1/4
Cr-1 Mo steel and Incoloy 800 presently used in HTGR steam generators must
be predicated on expected ASME Code approval of the substitute material.
Since ASME approval is primarily based on the evaluation of substantial
quantities of long-term, high-temperature creep-rupture data, the approval
process for Type 422 stainless steel, a material normally utilized for
turbine and compressor blades and rotors, may be long and costly.

Based mainly on the need for greater resistance to the decarburizing

effect of high-temperature [r^510°C (n.950°F) ] sodium than is offered by
2-1/4 Cr-1 Mo steel, Patriarca et al. (Ref. 7-5) discuss the status of a
similar program (possible substitution of 9% to 12% chrome alloys or
Incoloy 800 for 2-1/4 Cr-1 Mo steel) for application to LMFBR steam genera­
tors. However, it is estimated (Ref. 7-5) that ASME Code approval for the
presently unqualified high-chrome steels may take 5 to 10 yr.

7-6
 A detailed description of materials, fabrication, and inspection
development for the Clinch River breeder reactor plant steam generators is
provided by McClung et al. (Ref. 7-6). Many of the described areas,
including the use of 2-1/4 Cr-1 Mo steel and the tube-to-tubesheet stub
butt weld, are of potential applicability to the GCFR steam generator. For
example, the objectives of vacuum arc remelt (VAR) or electro slag remelt
(ESR) for tubing and tubesheets were (1) to reduce the size and quantity of
inclusions in the tubesheet stubs, (2) to minimize defective welds between
tubes and tubesheet stubs, and (3) to reduce residual impurities in the
material that are likely to produce post-weld embrittlement.

The use of a bore side butt weld process for the critical tube-to-
tubesheet stub weld produced a weld which was fully inspectable and also
eliminated crevices between the tube and tubesheet. Demonstration of the
suitability of this type of tube-to-tubesheet weld is evidenced by the
satisfactory 4000-hr steaming test of the Atomics International modular
steam generator and by a successful several-thousand-hour test of a 45-
MW(t) mockup of a straight tube Super Phenix steam generator (Ref. 7-7).
The same type of weld was used in the fabrication of the stainless steel
intermediate heat exchangers for the Fast Flux Test Facility and for the
Dutch/German SNR-300 fast breeder reactor straight tube steam generators.

Table 7-2 (from Ref. 7-6) compares the ASME Code and RDT M3-33
requirements for 2-1/4 Cr-1 Mo steel. As described in Section 7.2.3, 2-1/4
Cr-1 Mo steel tubing meeting RDT M3-33 specifications costs approximately
40% more than that meeting ASME Code specifications.

1.2.2.2. Allowable Stresses. The allowable stresses for the permissible

materials, according to the ASME Code, Section III, are listed in Code Case
1592. Table 7-3 lists the allowable stresses for Incoloy 800H, 2-1/4 Cr-1
Mo steel, and Type 304 stainless steel as a function of temperature and
service life.

Sq is defined as the maximum allowable value of general primary mem­

brane stress intensity to be used as a reference for stress calculations

7-7
 TABLE 7-2
COMPARISON OF ASME CODE AND RDT M3-33 REQUIREMENTS FOR STEAM GENERATOR TUBING (FROM REF. 7-6)

ASME Code, SA 213 RDT M3-33

Material Electric furnace VAR/ESR

Composition (%) 0.15 C max. 0,07-0.12 C, low residuals
Heat treatment Anneal or normalize and temper Anneal
Surface condition Free from scale Light treatment oxide, 63 yin.
Dimensions (in.
tolerance)
Outer diameter ±0.004 +0.005/-0

Wall thickness ±0.020/-0 +0.010/-0

Check analysis 1 per heat %5 of tubes

Tensile properties
Ultimate tensile >60 60 to 85
strength (ksi)
0.2% yield stress (ksi) >30 k30
Elongation (%) >20 ^30
Hydrostatic test In lieu of NDE, 1 or up to 24-ksi In addition to NDE, 4.5 ksi (22.5-ksi
fiber stress fiber stress)
Quality assurance NA 3700 (1/1/75) RDTF 2-4, NA 3700 (1/1/75)

Nondestructive
evaluation
Ultrasonic 5% of wall, or 0.0055 in. 3% or 0.004 in. (optional 3% or
0.002 in.) statistical control

Liquid penetrant Not specified O.D. and I.D. (I.D. is optional

provision), no indications

Eddy current In lieu of ultrasonic 3% or 0.004 in. (optional

provision)
_8 2
Helium leak Not specified Less than 10 standard cm /sec

Microcleanliness Not specified Required

Decarburization Not specified <5% of wall

7-8
 TABLE 7-3
ALLOWABLE STRESSES, 1000 PSI (MPa)

Incoloy Alloy 800H 2-1/4 Cr -1 Mo Steel Type 304 SS

S „ s S ^
mt mt mt
Temp.
105 hr 3 x 10~* hr 3 x 10^ hr 105 hr 3 x 105 hr

o
[°F (°C)j 1 0 hr 10 hr 1 0 hr

u
Sq So so
800 15. 3 15.3 15.3 15.3 15.0 17.9 17.9 16.1 15.1 15.1 15.1 15.1
(427) (105.5) (105.5) (105.5) (105.5) (103) (123.4) (123.4) (111.0) (104.1) (104.1) (104.1) (104.1)

900 14.8 14.8 . 14.8 14.8 13.1 17.2 10.9 9.6 14.6 14.6 14.6 14.6
(482) (102) (102) (102) (102) (90.3) (118.6) (75.2) (66.2) (100.7) (100.7) (100.7) (100.7)

950 14.6 14.6 14.6 14.6 11.0 16.7 8.4 7.3 14.3 14.3 14.2 12.2
(510) (100.7) (100.7) (100.7) (100.7) (75.9) (115.2) (57.9) (50.3) (98.6) (98.6) (97.9) (84.1)

1000 14.4 14.4 14.4 14.4 7.8 15.9 6.3 5.2 13.7 14.0 11.1 9.3
(538) (99.3) (99.3) (99.3) (99.3) (53.8) (109.6) (43.4) (35.9) (94.5) (96.5) (76.6) (64.1)
under design conditions. Smt is defined as the allowable limit of general
primary membrane stress intensity to be used as a reference for stress
calculations for the actual service life and under the normal plus upset
conditions; the Smt values are the lower of two stress intensity values,
S (time-independent) and S. (time-dependent).
m t

Note that at 427°C (800°F), the highest allowable stress (S . ) is for

mt
2-1/4 Cr-1 Mo steel rather than for the high-alloy materials. However, as
temperature and service life increase, the allowable stress for 2-1/4 Cr-1
Mo steel decreases substantially in comparison with the modest reduction
exhibited by the high-alloy materials. At 510°C (950°F), the near-maximum
GCFR tube temperature, the allowable stress for 2-1/4 Cr-1 Mo steel is half
that for Incoloy 800H, and it becomes progressively weaker at higher
temperatures.

The minimum wall thickness, t^, required for a straight cylinder under
internal pressure as defined by the ASME Code, Section III (Nuclear Power
Plant Components) is

PD
t = _______ o------ + a
m 2 (S + 0.4 P)
m

where P « internal pressure, MPa (psi),

Dq - outside diameter of cylinder, mm (in.),
S - allowable stress, MPa (psi) (S for design conditions; S for
m * o mt
design life and temperature),
a - corrosion allowance, mm (in.).

For curved tubes with a bend radius of at least six tube diameters, the
Code recommends a 6% increase in t .
m

7.2.3. Material Cost

Based partly on its demonstrated suitability in high-temperature

service and partly on HTGR steam generator experience, Incoloy 800 was

7-10
chosen as the reference tubing material for the GCFR steam generator. How­
ever, because the maximum helium temperature in the 300-MW(e) GCFR [550°C
(1022°F)] is considerably lower than that in the HTGR [741°C (1366°F)], the
maximum tube temperature in the 300-MW(e) GCFR steam generator is only
about 521°C (970°F), compared with 710°C (1310°F) (reheater) and 676°C
(1250°F) (main bundle) for the HTGR. At temperatures up to 538°C (1000°F),
2-1/4 Cr-1 Mo steel is suitable with respect to strength and corrosion
requirements and makes the use of Incoloy 800H, with its superior high-
temperature [over 538°C (1000°F)] strength, unnecessary.

Since tubing is a major cost item for a steam generator, a cost com­
parison was made for the helical coil straight tube steam generators using
Incoloy 800H and 2-1/4 Cr-1 Mo steel. Comparative costs (Ref. 7-8) for the
tubing materials considered are shown below:

Tubing Material Cost

[$/kg ($/lb)]
Specification 2-1/4 Cr-1 Mo Incoloy 800H

ASME
SA 213 T 22 4.49
(2.04) —

SA 163 Gr 1 __
13.8
__
(6.26)
RDT
M3-3 3 6.1 15.1
(based on ASME) (2.77) (6.86)

Note that for ASME specification tubing, Incoloy 800H costs 3.07 times as
much as 2-1/4 Cr-1 Mo steel; for RDT specification tubing, Incoloy 800H
costs 2.48 times as much as 2-1/4 Cr-1 Mo steel.

Referring to Table 8-2, which compares helical coil and straight tube
steam generators for the GCFR, a helical coil steam generator with a 2.74-m
(9 ft) bundle diameter requires a relative tubing weight (2-1/4 Cr-1 Mo/
Incoloy 800H) of 1.94. Thus, the relative tubing cost (2-1/4 Cr-1 Mo/
Incoloy 800H) and actual cost saving per steam generator are as follows:

7-11
 (a)
Cost Saving
Tubing Spec. Relative Cost ($)

ASHE 1.94/3.07 =* 0.63 133,000

KDT 1.94/2.48 « 0.78 86,000

(a)
If lead-in/lead-out tubing costs are included,
the cost saving will be greater.

A similar comparison for straight tube steam generators (Table 8-3) shows
the relative tubing cost (2-1/4 Cr-1 Mo/lncoloy 800H) and the actual cost
saving per steam generator to be:

Cost Saving
Tubing Spec. Relative Cost ($)

ASHE 2.38/3.07 » 0.78 114,000

RDT 2.38/2.48 « 0.96 21,000

Thus, a significant saving in active tubing cost can be made by substitu­

tion of 2-1/4 Cr-1 Mo steel for Incoloy 80OH.

In addition to tubing cost factors, which favor 2-1/4 Cr-1 Mo steel

over Incoloy 800H, the long-term outlook for the tubing material should
also be considered. Because the future availability of alloying elements
such as chromium, nickel, and molybdenum is uncertain and substantial price
increases for these and other relatively scarce metals can be anticipated,
the selection of a steam generator tubing material on the basis of having
the lowest alloy content, consistent with meeting all design criteria, is
prudent. In this respect, too, 2-1/4 Cr-1 Mo steel has an advantage over
Incoloy 800H.

7.2.4. Permeability of Metals to Tritium

The Nuclear Regulatory Commission safety regulations associated with

plant licensing require prediction of the potential release of tritium to
the environment by a nuclear power plant. However, estimation of the tri­
tium release from a GCFR to the environment requires knowledge of:

7-12
 1. The tritium yield, or source term, from GCFR fuel rods.

2. The tritium permeability of steam generator tubing material.

As demonstrated in the discussion below, the incomplete state of knowledge

in each of the above areas prevents the possible substitution of less
expensive 2-1/4 Cr-1 Mo steel for the reference material, Incoloy 800H.

Tests performed at General Atomic (Ref. 7-9) with clean, new tubing
show that the hydrogen permeation rate for Incoloy 800 is about one-
twentieth that of 2-1/4 Cr-1 Mo steel or low-carbon steel. However, when
an oxide film is formed on the surface of any of these materials, the
permeation rate is reduced by a factor of 20 to 100.

Extensive tests, similar to those described above, conducted at Oak

Ridge National Laboratory (ORNL) (Ref. 7-10) confirm that the presence of
an oxide film on a metal can drastically reduce the permeability of the
metal to deuterium. In the ORNL deuterium permeability tests, the fol­
lowing metals were investigated: nickel. Type 304L stainless steel,
Hastelloy N, Incoloy 800, Croloy T9 (9 Cr-1 Mo), Croloy T22 (2-1/4 Cr-1
Mo), and Type 406 stainless steel. Deuterium permeation rates for oxidized
and unoxidized conditions were investigated for Incoloy 800, 9 Cr-1 Mo
steel, 2-1/4 Cr-1 Mo steel, and Type 406 stainless steel. The presence of
an oxide film reduced the deuterium permeation rate by a factor of about
100 at 1 atm, and by a greater factor at lower pressures, for all metals
except 2-1/4 Cr-1 Mo steel, for which the permeation rate was virtually
unchanged. It was speculated (Ref. 7-10) that the oxide film formed on the
2-1 /4 Cr-1 Mo steel specimen must have been highly defective.

Factors which influence the permeation effectiveness of the oxide film

include the film thickness, density, and composition and the size and num­
ber of defects in the film. Present evidence as to the effectiveness of an
oxide film in reducing the permeability of 2-1/4 Cr-1 Mo steel to hydrogen
or deuterium, and presumably to tritium, is inconclusive. To firmly estab­
lish the permeation characteristics of 2-1/4 Cr-1 Mo steel, additional

7-13
tests are required, preferably with metal specimens that previously have
been exposed to power plant water on the inside surface and to a helium
mixture typical of gas-cooled reactors on the outside surface.

Tritium permeability tests are being performed at General Atomic (Ref.

7-11) on carbon steel and Incoloy 800 tubing specimens which were taken
from the Peach Bottom steam generator after 7 yr of operation. Preliminary
results indicate that the permeability of carbon steel may be comparable to
that of Incoloy 800 when an oxide film barrier has formed on the water side
in service. Under these conditions, tritium release would be temperature
governed and the high-temperature Incoloy 800 section of the Peach Bottom
steam generator would have contributed the greatest (^90%) tritium release
fraction.

If the results of permeability tests on 2-1/4 Cr-1 Mo steel planned

for the near future demonstrate tritium permeability comparable to that of
carbon steel and Incoloy 800, then a major uncertainty will have been
removed from the GCFR steam generator material substitution evaluation.

As a complement to the tritium permeation investigations described

above. General Atomic proposed a program to ERDA (Ref. 7-12) which had as
its objectives the determination of (1) the molecular species of tritium
(HT, T2» or HTO) that are formed in a GCFR fuel rod, (2) the amount of
tritium expected to be vented relative to that which would permeate the
cladding, and (3) the tritium yield from Pu-239 thermal fission.

If tritium occurs as HT or T2, it can permeate the fuel cladding,

enter the helium coolant, and then permeate the steam generator tubing. On
the other hand, if tritium occurs as HTO, permeation of metal barriers will
occur only through cracks or holes. In the latter case, there would be
little permeation through the cladding and more than 90% of the tritium
appearing in the fuel region would be removed by the helium purification
system units and be disposed of in solid form. The irradiation portion of
the experiment has been completed (Ref. 7-12), and the postirradiation

7-14
examination of the test capsule (NaK coolant, Zircaloy cladding, and
charcoal trap) for tritium content has been started.

In a recent summary of available data, Langer (Ref. 7-13) shows that

the tritium yield for U-238 and Th-232 is an order of magnitude greater
than expected. As a consequence, the anticipated tritium production for
the GCFR demonstration plant is more than twice that originally projected.
This larger source term was found to be tolerable, according to the current
NRC tritium release requirements, but may be "borderline" when 2-1/4 Cr-1
Mo steel of high permeability is used. While tritium yield data exist for
U-235 in a fast flux, no similar data are available for Pu-239, Pu-240, Pu-
241, and Pu-242. If the increased yield observed for U-238 and Th-232 also
prevails for plutonium and the tritium permeability of 2-1/4 Cr-1 Mo steel
proves to be higher than that of either carbon steel or Incoloy 800, then
the use of less expensive 2-1/4 Cr-1 Mo steel may not be possible unless
other measures are taken in the plant design to limit tritium release.

7.2.5. Corrosion

7.2.5.1. Helium Side. Corrosion of Incoloy 800 and 2-1/4 Cr-1 Mo steel by
helium containing contaminants typical of HTGR coolant has been studied
extensively, and the findings generally indicate negligible effects (Ref.
7-4). However, recent studies show that long-term exposure to HTGR helium
may decrease metal creep-rupture properties by from 0% to 20%.

While much additional work must be done to determine the exact role
played by helium chemistry, temperature, metal chemistry, and stress on
metal properties, it is anticipated that the relatively low maximum helium
temperature [550°C (1022°F)] used in the GCFR will minimize the potential
for metal property changes.

7.2.5.2. Water Side. Incoloy 800 has amassed a very favorable service
record with high-temperature steam and, in many instances, has provided
satisfactory service with chloride- and oxygen-contaminated water (Ref.
7-4). However, while superior to the austenitic steels in stress corrosion

7-15
cracking resistance, Incoloy 800 has been shown to be susceptible to stress
corrosion cracking when exposed to chloride-contaminated water (Refs. 7-14,
7-15, 7-16). For example, as-welded Incoloy 800 samples were more suscep­
tible to cracking than welded specimens that had been annealed or unwelded
specimens. Welded and unwelded samples of 2-1/4 Cr-1 Mo showed no tendency
to crack under an oxygenated chloride environment (Refs. 7-15, 7-16).

At moderate caustic levels in water (100 ppm sodium hydroxide),

Incoloy 800 completely resists stress corrosion cracking (Ref. 7-16).
However, using 3% to 10% caustic solutions, tests conducted at LMFBR steam
generator conditions [316°C (600°F) water-evaporator] demonstrated that
2-1/4 Cr-1 Mo steel had excellent resistance to caustic cracking, whereas
Incoloy 800 was susceptible to cracking, except when cold worked (Ref.
7-17). At temperatures below 232°C (450°F) and with high caustic levels
(20% to 33%), 2-1/4 Cr-1 Mo steel has been caused to crack (Ref. 7-17).

Under LMFBR superheater conditions [482°C (900°F) steam] and using a

0.1% caustic solution, neither Incoloy 800 nor 2-1/4 Cr-1 Mo steel
exhibited stress corrosion cracking (Ref. 7-17). At the same temperature
but at a caustic solution level of 1.0%, all eight 2-1/4 Cr-1 Mo steel
samples failed from severe corrosion. For the latter condition, the
corrosion rate of Incoloy 800 was 1/15 to 1/30 of that of 2-1/4 Cr-1 Mo
steel; however, one of six Incoloy 800 samples did fail.

As part of the Clinch River breeder reactor steam generator materials

program, De Van and Criess (Ref. 7-18) have provided a summary of on-going
water side corrosion tests performed on 2-1/4 Cr-1 Mo steel. These tests
cover the following conditions;

1. Isothermal conditions. Evaporator or superheater temperatures in

pure or chloride-contaminated water/steam; chloride-contaminated
water/steam in which the tube specimen is alternately subjected
to evaporator or superheater temperature conditions.

7-16
 2. Heat transfer conditions. Heating of pure superheated steam at
the tube surface; nucleate boiling at the tube surface and at
crevices using chloride-contaminated water or cooling tower
(Clinch River) water.

Further tests will be conducted using vacuum arc remelt (VAR) 2-1/4 Cr-1 Mo
tubing, since it is apparently in this metallurgical condition that 2-1/4
Cr-1 Mo steel has the greatest corrosion resistance.

While the 2-1/4 Cr-1 Mo steel specimens did not crack under any of the
diverse conditions described above, general corrosion rates varying from
slight to very high were observed. Based on actual long-term steam genera­
tor operating experience and their own short-term data, De Van and Griess
estimate (Ref. 7-18) a maximum corrosion allowance of 0.43 mm (0.017 in.)
for the lifetime of an LMFBR steam generator operating under normal condi­
tions. On this basis, the 0.76 mm (0.030 in.) corrosion allowance used for
2-1/4 Cr-1 Mo steel in GCFR steam generator sizing studies is ample.

It should be noted that Griess et al. (Ref. 7-16) point out that the
conditions of the oxygenated chloride environment test were more severe
than those which could exist for any extended time in a reactor system and
were chosen to ascertain metal resistance to chloride-induced cracking.
Consequently, the high general or uniform corrosion exhibited by 2-1/4 Cr-1
Mo steel in the tests does not compromise its use in the Clinch River
breeder reactor evaporator and superheater, since the demonstrated resist­
ance of this alloy to cracking in both chloride and caustic environments
makes tube failure due to stress corrosion cracking a remote possibility.

Based partly on its favorable industrial record, the results described

above, and its satisfactory behavior in a sodium environment, 2-1/4 Cr-1 Mo
steel was chosen as the reference tubing material* for the U.S. LMFBR steam
generators, and is also used in the economizer-evaporator sections of the

*Incoloy 800 is a back-up material.

7-17
LMFBR steam generators designed in England, France, Germany, Russia, and
Japan. The superheater sections of these steam generators use either 2-1/4
Cr-1 Mo steel or austenitic steel.

Although the Phenix steam generator (economizer-evaporator, 2-1/4 Cr-1

Mo steel; superheater-reheater, Type 321 stainless steel) provided excel­
lent service, the design chosen for the Super Phenix steam generator
utilizes Incoloy 800 throughout (Ref. 7-7). This departure from typical
LMFBR material practice was predicated on concern that decarburization of
2-1/4 Cr-1 Mo steel by sodium at 525°C (977°F) might present problems and
on the fact that additional safety could be obtained by the use of Incoloy
800. However, the susceptibility of Incoloy 800 to stress corrosion
cracking in chloride-contaminated water and its unsatisfactory behavior in
hot sodium with 10% sodium hydroxide content have necessitated a test pro­
gram to clarify these areas. A satisfactory, several-thousand-hour test of
a 45-MW(t) mockup of a Super Phenix helical coil steam generator, using
Incoloy 800 throughout, has already been completed, but the water and
sodium purity levels for the test are not specified.

The dominance of 2-1/4 Cr-1 Mo steel tubing in the economizer-

evaporator sections of steam generators also prevails for gas-cooled
reactor plants as exemplified by Fort St. Vrain and the German THTR and by
HTGR designs. In each of these units, Incoloy 800 is used in the
superheater-reheater sections.

7.3. TUBE STRESSES

7.3.1. Critical Heat Flux Induced Evaporator Tube Stresses

The occurrence of critical heat flux inside the tube of a steam gener­
ator is accompanied by temperature, pressure, and possibly flow oscilla­
tions. Temperature oscillations can produce fluctuating thermal stresses
in not only the tube wall proper, but also in the protective oxide film
(magnetite for 2-1/4 Cr-1 Mo steel) which forms on the inner tube surface.

7-18
The fluctuating thermal stresses may lead to either tube failure through
fatigue or accelerated tube corrosion as a result of cracking of the oxide
layer. Thus, it is important in steam generator design to assure that the
occurrence of critical heat flux does not result in premature tube failure.

For application to gas-cooled reactor steam generators, Lasarev (Ref.

7-19) describes the results of an analytical method for predicting tube
stress and fatigue life for an evaporator tube subjected to thermal oscil­
lations associated with critical heat flux. The method relies on a tran­
sient thermal analysis in which the water side heat transfer coefficient is
assumed to vary in magnitude between that of nucleate boiling and that of
film boiling for a wide range of frequencies. The radial tube temperature
fluctuations associated with the changing heat transfer coefficients are
then used to calculate the fluctuating inside surface thermal stress for a
thick-walled cylinder. Using the fluctuating stress and a derived fatigue
strength curve, the allowable number of cycles and fatigue lifetime are
determined. Comparison of the analytical tube life prediction with that
actually observed for a Dragon steam generator was favorable.

As part of the LMFBR steam generator program, Argonne National

Laboratory (ANL) is conducting an extensive analytical and experimental
program to determine the effect of critical heat flux induced temperature
oscillations on tube wall stress, tube corrosion, and tube lifetime. The
program combines thermal-hydraulic testing of a full-scale LMFBR steam gen­
erator tube (Ref. 7-20) with a stress analysis program (Ref. 7-21). The
detailed local experimental thermal measurements are used as input to the
stress analysis program, which is similar in basis but more detailed than
that of Lasarev (Ref. 7-19). Reported test results from the ANL program
(Ref. 7-22) show that the maximum tube wall temperature oscillations occur
in the transition boiling region and have a value of about 20°C (36°F).
More recent, but preliminary, results based on maintaining the critical
heat flux at an essentially fixed location in the tube for 100 min indicate
a resultant alternating stress at zero mean stress of 113 MPa (16,385 psi)
(Ref. 7-23).

7-19
 While much of the ANL LMFBR steam generator work described above is
expected to be applicable to GCFR steam generators, the following should be
considered:

1. The GCFR critical heat flux is about one-third of that existing

in the LMFBR.

2. The design pressure level in the GCFR is about 20.3 MPa (2950
psia) compared with 12.7 MPa (1840 psia) for the LMFBR.

3. The steam quality at critical heat flux is about 70% (average)

for the GCFR compared with about 20% for the LMFBR.

4. The water mass velocities for the GCFR and the LMFBR are similar.

5. The critical heat flux occurs at one location in the straight

tube LMFBR, whereas it can occur along a substantial portion of
the evaporator length, in different peripheral locations, in a
GCFR helical coil tube.

6. Both the tube wall thickness and tube thermal conductivity will
be less for the GCFR (if Incoloy 800H is used) than for the
LMFBR.

Based on the above considerations, and particularly items 1, 2, and 3 *

the tube temperature increase at the onset of critical heat flux should be
substantially less for the GCFR than for the LMFBR. However, the asym­
metric peripheral tube temperature distribution (along much of the
evaporator length) associated with helical coil critical heat flux will
increase the stress level (Ref. 7-19) in the GCFR and may also result in
tube corrosion behavior which differs from that observed for LMFBR tubes.

7-20
7.3.2. Tube-Tube Support Stresses

Evaluation of the tube and tube support plate stresses in a helical

coil steam generator requires knowledge of the tube and tube support
material properties and geometries, the helium and tube temperature pro­
files throughout the bundle, and the tube to tube support mechanical
clearances. Since the tube support is at essentially helium temperature
and the tube wall is at a temperature intermediate to those of the helium
and the steam, the radial growth of the tube support plate will exceed that
of the tube even when each part has identical thermal expansion coeffi­
cients. This condition leads to tube stresses which cause the tube to bend
around the support plate at the holes and to straighten between the plates.
Concurrently, the tube support plate experiences shear and bending stresses
from the forces created by the differential thermal expansion between it
and the tube. These thermal stresses can be determined by the HTGR-
developed code CSTRES.

As shown in Fig. 7-1 (from Ref. 7-24), the GCFR tube support plate to
tube temperature differences and the helium temperature change are con­
siderably less than those in the HTGR. As a result, the tube support plate
stresses are much lower for the GCFR than for the HTGR, but the GCFR tube
axial stress can, under certain conditions, be quite high. According to
Ref. 7-24, the GCFR tube stress was found to be sensitive to the number and
stiffness of the tube support plates and to the tube and tube support plate
material (if the latter has a lower thermal expansion coefficient the
stresses are reduced). Another GCFR thermal stress evaluation (Ref. 7-25),
based on an improved and expanded CSTRES program, showed that tube stress
could be reduced by about 50% if the tube support plate hole clearance was
increased from 0 to 0.5 mm (0 to 0.02 in.).

Because of the mutual dependence of the tube to tube support plate

design (number of support plates and stiffness, materials of construction,
mechanical clearances, etc.) and the tube vibratory characteristics, and
associated stresses, it is necessary that analysis in both areas be
concurrent.

7-21
TEMPERATURE [°F (°C)]

PLATE

TUBE

PLATE

GCFR
TUBE

DISTANCE FROM SUPERHEATER, IN. (m)

Fig. 7-1. Plate and tube temperatures in HTGR and GCFR boilers

7-22
REFERENCES

7-1 . Pai, D. H., J. K. O’Donoghue, and J. R. Stenner, "Application Expe­

rience with Codes and Standards in the Design of Heat Exchangers for
Liquid Metal and Gas-Cooled Reactor Systems," ASME Paper 75-Pwr-16,
September 1975.
7-2. Jakub, M. T., "Considerations in Applying Current Construction Codes
to Heat Exchangers in Nuclear Power Plants," ASME Paper 75-Pwr-15,
September 1975.
7-3. Gwaltney, R, C., "An Interim Assessment of the Structural Performance
of the HTGR Steam Generator," Oak Ridge National Laboratory,
unpublished data,
7-4. Mazandarany, F. N., and P. L. Rittenhouse, "Effects of Service
Environments on the Behavior of High Temperature Gas-Cooled Reactor
Steam Generator Structural Materials," Nucl. Technol. 28, No. 3,
406 (1976).
7-5. Patriarca, P,, et al., "U.S. Advanced Materials Development Program
for Steam Generators," Nucl, Technol. 28, No. 3, 516 (1976).
7-6. McClung, R. W., et al., "Fabrication and Inspection Development for
Clinch River Breeder Plant Steam Generators," Nucl. Technol. 28,
No. 3, 374 (1976.
7-7. Robin, M. G., "French Steam Generator Experience - Phenix and
Beyond," Nucl. Technol. 28, No. 3, 482 (1976) .
7-8. Schwarzer, D., General Atomic Company, personal communication,
July 1976.
7-9. Yang, L., W. Baugh, and R. E. Bruce, "Quarterly Progress Report on
Study of Tritium Permeation Rate Through Steam Generator Materials,"
General Atomic Company, unpublished data, September 4, 1974.
7-10. Strehlow, R. A., and H. C. Savage, "The Permeation of Hydrogen
Isotopes Through Structural Metals at Low Pressures and Through
Metal with Oxide Film Barriers," Nucl, Technol. 22, 127 (April
1974).
7-11. Simnad, M. T., "Selection of Reference Steam Generator Tubing
Material," General Atomic Company, unpublished data.

7-23
7-12. Weinberg, A. F., "Justification for the Measurement of Tritium in
Capsule GB-10, GCFR Program, AEG Contract AT(04-3)-167, Project
Agreement No. 23," General Atomic Company, unpublished data,
July 31, 1974.
7-13. Langer, S., "Selection of Reference Steam Generator Tubing Material,"
General Atomic Company, unpublished data, July 22, 1976.
7—14. Hammond, J. P., "General Corrosion," Fuels and Materials Development
Program Quarterly Program Report for the Period Ending September 30,
1971, USAEC Report ORNL-TM-3550, Oak Ridge National Laboratory,
December 1971.
7-15. Hammond, J. P., et al., "Comparative Results on Chloride Stress
Corrosion Cracking of Steam Generator Materials in Cyclic Steam
Environment," ERDA Report ORNL-5031, Oak Ridge National Laboratory,
June 1975.
7-16. Griess, J. C., J, P. Hammond, and W. A. Maxwell, "Effects of
Chloride and Caustic on the Cracking Behavior of Several Materials
Under Alternate Wet and Dry Steam Conditions," ERDA Report ORNL-TM-
4995, Oak Ridge National Laboratory, September 1975.
7-17. Indig, M. E., "Stress Corrosion Studies of LMFBR Steam Generator
Materials," General Electric Report GEAP-12536, August 1974.
7-18. De Van, J. H., and J. C. Griess, "Clinch River Breeder Reactor
Environmental Effects - General Water-Side Corrosion," Nucl.
Technol. 28, No. 3, 398 (1976).
7-19. Lasarev, M., "The Effect of an Oscillating Dryout Point on Evaporator
Tube Lifetime," Trans. Am. Nucl. Soc. 15, No. 1, 396 (1972).
7-20. Stevens, H. C., and D. M. France, "Development of a Thermal Hydraulic
Test Facility for Full-Scale LMFBR Steam Generator Tubes," Trans,
Am. Nucl. Soc. 22, 538 (1975).
7-21. Chiang, T., D. M. France, and T. R. Bump, "DNB Induced Thermal
Stress and Fatigue in LMFBR Evaporator Tubes with Oxide Scale,"
USAEC Report ANL-CT-75-24, Argonne National Laboratory, 1974.
7-22. France, D. M., T. Chiang, and R. D. Carlson, "Experimental Obser­
vation of Thermal Oscillations in LMFBR Steam Generator Tubes,"
Trans. Am. Nucl, Soc. 23, 385 (1976).

7-24
7-23. France, D. M., T. Chiang, and R. D. Carlson, "Early SGTF Test
Results for CRBR 100% and 64% Loads," Argonne National Laboratory,
unpublished data, July 8, 1976.
7-24. Thompson, W. I., "Thermal Stress Comparison of HTGR and GCFR
Boilers," General Atomic Company, unpublished data, June 15, 1973.
7-25. Hinz, R., "Preliminary Study on Tube Thermal Stresses in the 300
MW(e) Demo Steam Generator Using the CSTRES Program," General
Atomic Company, unpublished data, May 10, 1974.
 8. COMPARISON OF HELICAL COIL AND STRAIGHT TUBE
STEAM GENERATORS FOR GCFR

A comparison has been made between straight tube and helical coil
steam generators for application to GCFR requirements. This study included
the following areas: (1) the design features and characteristics of
straight tube steam generators used in PWRs, the Clinch River breeder
reactor plant, and the Super Phenix breeder plant were examined and are
described herein; (2) using the updated reference GCFR cycle, straight tube
and helical coil steam generators were sized; and (3) using the steam gen­
erator sizes determined in (2), various possible steam generator arrange­
ments in the PCRV were made.

8.1. CURRENT STRAIGHT TUBE STEAM GENERATORS

The straight tube steam generators which appeared to be of interest •

and of potential applicability to the GCFR were the Babcock & Wilcox once-
through unit used in PWRs, the Atomics International "hockey-stick" unit to
be used in the Clinch River breeder reactor plant, and the Stein Industrie
units proposed for the Super Phenix breeder reactor.

8.1.1. Babcock & Wilcox Unit

The main features of the Babcock & Wilcox unit are given in Table 3-1,
and a drawing of the unit is shown in Fig. 8-1. A description of this unit
is contained in Ref. 8-1.

Pressurized high-temperature water from the reactor flows downward

through the inside of the 15,743 16 mm (5/8 in.) diameter Inconel 600
tubes, while feedwater, which enters just above the bottom, flows upward on
the outside of the tubes as it is heated and turns into slightly super­
heated [a'28°C (O/50oF) ] steam. At the top of the unit, the superheated

8-1
 REACTOR
MOUNT
INLET
2174 PSIA (15 MPa)
630°F (332°C)
HANDHOLE

\ MANVAYS
—

SUPERHEAT
REGION

r—— FILM
J. BOIL NG

CYLINDRICAL t -----:L
REGION
BAFFLE
HEAT TRANSFER
— —
REGIONS AT
100* POWER
STEAM ANNULUS fI NUCLEATE
BOILING
-—- — REGION

1
L
STEAM
OUTLETS (2)
HANDHOLES 1135 PSIA (7.8 MPa)-
AUXILIARY FEEDWATER INLET 595°F (313°C)
FEEDWATER
? INLETS (2)
471°F (244°C)
SUBCOOlEO
BOILING
REGION
V

2 REACTOR
COOLANT —----- - ----- -
OUTLETS 2166 PSIA (14.9 MPa) MANWAT
573°F (300°C) 7 E
DRAIN NOZZLE

Fig. 8-1. Babcock & Wilcox steam generator

8-2
steam makes a 180° turn and flows downward between the annulus formed by
the inner baffle and the outer (pressure) shell until it exits at a point
which is located slightly above the feedwater inlet. Through the use of
this steam routing, and by having nucleate boiling for almost all of the
evaporator length (steam quality at critical heat flux ^90%) combined with
relatively short economizer and superheat regions, the length-average tem­
peratures of the tubing and the pressure shell are kept nearly equal. As a
result, no provision for differential thermal expansion is required for the
structure, which features tubes welded to tubesheets that, in turn, are
welded to the pressure shell.

Research and development efforts carried out in support of this

design, including the areas of thermal-hydraulics, boiling stability,
transient behavior, vibration and tube supports, and stress analysis, are
described in Ref, 8-2,

Significant differences between the above unit and GCFR requirements

are (1) the PWR primary and secondary temperature levels are several hun­
dred degrees lower than in the GCFR; (2) because of the higher temperatures
in the GCFR, material choices are more limited and thermal growth problems
are more pronounced; (3) in GCFR steam generators, steam flows inside the
tubes rather than on the outside as in the Babcock & Wilcox unit; (4)
because GCFR steam generators are located inside the PCRV, large steam
leaks resulting from a GCFR tubesheet head rupture may require double
containment.

8.1.2. Clinch River Breeder Reactor

The steam generators for the Clinch River breeder reactor plant are
based on the Atomics International modular steam generator ("hockey-
stick") , which completed a problem-free 4000-hr steaming test. The per­
formance characteristics of this steam generator are shown in Table 3-1,
and a drawing of the unit is shown in Fig. 8-2. A description of this unit
is contained in Ref. 8-3, and the initial development is described in Ref.
8-4.

8-3
 VIBRATION
SUPPRESSORS-*. REMOVABLE
STEAM HEAD

STEAM OUTLET

Jr“”“
-t_______
IZIEtl
____
X 0.109-IN. WALL)
65 FT, 1
(15.9 mm X 2.7 mm
(19.8 m)
BUTT-WELD (IBW)-
SODIUM OUTLET MACHINED
(2 NOZZLES) BOSS
DRAIN NOZZLE

REMOVABLE STEAM/WATER HEAD

STEAM/WATER INLET

Fig. 8-2. Clinch River breeder reactor plant steam generator

8-4
 This steam generator is essentially a straight tube unit with fixed
tubesheets in which differential tube thermal expansion is accommodated by
a large radius bend near the top (hot end) of the unit. For both the
superheater and evaporator, high-temperature sodium enters the shell at a
point slightly below the expansion bend, flows downward on the outside of
the tubes, and exits near the bottom of the shell. Feedwater enters the
two evaporators (757 tubes each) at the bottom tubesheet and flows upward
as it is heated and turns into a 50% quality steam mixture. After passing
through the expansion bend and exiting at the top tubesheet, the wet mix­
ture goes to a steam drum where the water and steam are separated. From
there, the saturated steam goes to a single superheater, which is physi­
cally identical to an evaporator, where the steam is superheated before it
exits through the upper tubesheet.

While the sodium temperature levels in the Clinch River unit are
similar to the helium temperatures in the GCFR and thus pose similar
restraints on material choices and thermal expansion provisions on the
units, there are significant differences. Included among these are:

1. The Clinch River units have individual evaporators and super­

heaters and utilize a steam drum to separate water from the wet
steam produced by the evaporator. In contrast, the GCFR unit is
of the once-through type which produces superheated steam from a
single module. Space and geometry restraints posed by the
incorporation of the steam generators inside the PCRV rule
against the use of the Clinch River type units.

2. While a "hockey-stick" configuration permits accommodation of

differential thermal expansion, its shape does not lend itself to
a practical steam generator closure in the PCRV.

3. The application of a straight tube design, as typified by the

Clinch River unit, to GCFR requirements results in a unit of
considerable length and relatively small diameter (see Section
8.2).

8-5
8.1.3. Super Phenix Breeder Reactor

The steam generators proposed for use in the Super Phenix reactor by
Stein Industrie are of the straight tube type with separate and individual
evaporator and superheater. Figure 8-3 (from Ref. 8-5) shows the general
arrangement of the units as well as their main characteristics.

Each evaporator and superheater unit is center supported at a point

just below the hot sodium inlet. After making a side entry into the super­
heater, the hot sodium from the reactor turns upward and flows between the
annulus formed by the outer shell and the bundle shell. At the top of the
unit, the sodium makes a 180° turn and flows downward over the outside of
the tubes. At the bottom of the unit, the sodium makes a 180° turn, flows
upward in the inner shell-outer shell annulus, and then exits just below
the center support. From there the cooler sodium goes to the evaporator,
where it follows a flow path identical to that of the superheater.

Feedwater enters the 1405 18-mm (0.70 in.) diameter evaporator tubes
and flows upward to become slightly superheated steam. After exiting from
the top of the evaporator, the slightly superheated steam is piped to the
bottom of the superheater where it flows upward through 595 21.3-mm (0.84
in.) diameter tubes and becomes highly superheated. The overall cycle
conditions and possible materials of construction are similar to those of
the GCFR.

Differential thermal expansion between the tubes and the shell is

accommodated by means of a wavy tube section located at the hot end of the
evaporator and the superheater.

The evaporator of the above unit closely resembles, in length and

diameter, the straight tube GCFR steam generator (economizer, evaporator,
superheater in one unit) described in Section 8.2. Both of these units
require tubesheets about 2.0 m (6.5 ft) in diameter and about 0.45 m (1.5
ft) thick. However, the large tube spacing in the Super Phenix (1405
tubes) compared with the GCFR (5429 tubes) and the separate evaporator and

8-6
 Characteristics
{Super Phenix)

Heat Rating 750 MW(th)

Evaporator Superheater
Shell Diameter 2.344 m 2.248 m
Overall Height 25 m 22 m
Tube Number 1405 595
Tube Diameter 18 mm 21.3 mm
Tube Thickness 3 mm 2.3 mm
Material 21A Cr-1 Mo 316 SS
Weight (dry! 125 tons 100 tons

Water Inlet 235°C

Fig. 8-3. Stein Industrie steam generator

8-7
superheater in the Super Phenix permit the use of a wavy tube expansion
section with only a modest increase in overall bundle diameter and length.
For the GCFR, with its integral economizer, evaporator, and superheater,
the bundle length increase needed to accommodate a wavy tube section would
approximate the sum of the expansion lengths [about 4.9 m (16 ft)] of the
individual Super Phenix units. When this length is combined with the
approximately 18-m (59 ft) active bundle length for the GCFR, the total
length becomes prohibitive for incorporation in the PCRV.

8.2. HELICAL COIL AND STRAIGHT TUBE STEAM GENERATORS FOR GCFR

Based on the results of the size comparison described in Section

8.2.2, the characteristics and advantages/disadvantages of helical coil and
straight tube steam generators for GCFR application have been determined.
Table 8-1 shows a comparison of the general characteristics, and the
advantages and disadvantages of the two types of steam generators are
described below.

8.2.1. Advantages and Disadvantages* 1

8.2.1.1. Helical Coil Advantages. The advantages of the helical coil

design for GCFR application are as follows:

1. A large surface area is permitted in a limited volume.

2. A large helium flow area is combined with a small water flow

area.

3. A compact, favorable arrangement for a multi-cavity PCRV is

provided.

4. Practical incorporation of a resuperheater, if required, is

possible.

8-8
 TABLE 8-1
GENERAL CHARACTERISTICS OF HELICAL COIL AND STRAIGHT TUBE
STEAM GENERATORS

Helical Straight
Coil Tube

Number of tubes Low High

Tubesheet diameter Small Large
Bundle diameter Small/medium Small
Bundle length Short/medium Medium
Tube length Very long Medium
Tube surface area Medium High
Possible bundle size range Large Small
Tube material weight Medium High
Tube spacer cost Medium High
Fabrication cost High Medium
Helium pressure drop Low Low
Water/steam pressure drop High Low
Water velocity High Low

8-9
 5. Relatively high average critical heat flux steam quality can be
achieved.

6. Variable tube spacing permits a wide range of bundle sizes.

7. Relatively high overall heat transfer coefficient can be

achieved.

8. A great tube length is combined with a short bundle height.

9. There are few tubes in parallel and small diameter tubesheets can
be utilized.

10. Individual tube plugging is permitted.

11. Thermal expansion can be accommodated by tube bends.

8.2.1.2. Helical Coil Disadvantages. The disadvantages of the helical

coil design for GCFR application are as follows:

1. Extra tubing cost accrues from the lead-in/lead-out tubing

length.

2. Assembly cost is high (for threading individual coils through

support plates, assembling wear sleeves and wedges, numerous
welds to obtain needed tube length and subheadering, coiling of
tubing).

3. Choice of materials for tubing and support plate is important to

limit stresses.

4. Critical heat flux occurs along much of the evaporator length

instead of at one location. This is a difference, as compared
with straight tubes, which can affect the cross-sectional tube
temperature distribution but is not clearly a disadvantage.

8-10
 5 At very low flows, boiling may result in flow stratification
(two-phase and all-liquid) since the tubing is nearly horizontal.

6. The need for a center opening [0.91 m (3 ft) diameter] in the

bundle increases the bundle diameter.

7. Wear protection for tubes is required at tube supports.

8.2.1.3. Straight Tube Advantages. The advantages of the straight tube

design for GCFR application are as follows:

1. Relatively simple and rapid assembly reduces fabrication cost.

2. Construction is similar to once-through PWR and LMFBR steam

generators.

3. Low water side pressure drop reduces pumping power.

4. The tube length and bundle length are equal and standard tube
lengths can be used without intermediate welds.

5. Individual tube plugging is permitted.

6. The vertical orientation provides for uniformity of flow pattern

during low flow boiling.

7. Critical heat flux occurs in one location in the tube length.

8. A large surface area is permitted in a limited volume.

9. The bundle diameter is small.

8-11
8.2.1.4. Straight Tube Disadvantages. The disadvantages of the straight
tube design for GCFR application are as follows:

1. A large number of tubes in parallel are required.

2. Relatively costly tube support-spacers are required.

3. Large-diameter, thick tubesheets and heads are required.

4. A low overall heat transfer coefficient necessitates greater

surface area (and length) than the helical coil design.

5. The required helium flow area results in excessive water side

flow area and very low water velocity.

6. The helium side pressure drop due to tube support-spacers

decreases the pressure drop available to the bundle active
surface area.

7. The critical heat flux steam quality is low.

8. The variation of length with various bundle diameters is limited.

9. Accommodation of thermal expansion (inside or outside of the

shell) is more difficult than with the helical coil design.

10. The material cost is greater (for tubing, tubsheets, tube

support-spacers) than for the helical coil design.

11. Double heads may be required for protection in the event of a

tubesheet head rupture.

12. If a resuperheater is required, two units will be needed in two

PCRV cavities per steam generator; also, the helium pressure

8-12
 drop is prohibitively high due to two inlet and exit pressure
losses.

13. The long bundle length does not permit the helium circulator to
be placed in the same PCRV cavity as the steam generator.

8.2.2. Bundle Sizes

Using the steam generator code NUSIZE, which is an improved version of

the GENSIZE code (Ref. 8-6), a number of helical coil and straight tube
steam generators were sized to fulfill the performance requirements of the
updated reference cycle which uses three steam generators. Included in the
comparison were helical coil units of three different bundle diameters and
a straight tube unit. In each case, Incoloy 800H or 2-1/4 Cr-1 Mo steel
was used as tubing material.

The basis for the calculations included the following: (1) an updated
reference cycle [0.28 MPa (40 psi) helium pressure drop] using vendor's main
turbine characteristics and without a resuperheater in the steam generator;
(2) a total steam generator helium pressure drop of 0.059 MPa (8.5 psi): for
the helical coil units, 0.045 MPa (6.5 psi) in the bundle and 0.013 MPa (2.0
psi) for entrance and exit pressure losses; for the straight tube unit, 0.045
MPa (6.5 psi) in the bundle (50% to tubes, 50% to spacers) and 0.013 MPa
(2.0 psi) for entrance and exit pressure losses; (3) the use of 19 mm (0.75
in.) outside-diameter tubes for the economizer and 25.4 mm (1.0 in.) outside-
diameter tubes for the evaporator and superheater in the helical coil
units. This tube combination (item 3) provides the following advantages:

1. Since the economizer comprises about 50% of the tube (and bundle)
length, use of the 19 mm (0,75 in.) diameter tube in the
economizer maximizes the pressure drop in the all-liquid region
and thus decreases the inlet pressure drop required for boiling
static stability. As a fraction of the total steam generator
pressure drop, the economizer accounts for 44% (Incoloy 800H) and
23% (2-1/4 Cr-1 Mo steel), respectively.

8-13
 2 Since the holes in the tube supports are drilled to accommodate
25.4 mm (1 in.) diameter tubes in the upper half (approximately)
of the bundle and 19 mm (0.75 in.) diameter tubes in the lower
half, the assembly labor associated with threading the tube coils
through the support plates, from the top, is decreased as the 19
ram (0.75 in.) diameter tubes pass through the larger holes for
about one-half of the total bundle height.

3. The tube wall thickness is based on the maximum water to helium

pressure differential [13.4 MPa (1950 psi)], the maximum tube wall
temperature in a given section, the ASME Code, Section III,
requirements for (3 x 10 hr) life, and a corrosion allowance of
0.76 mm (0.030 in.) for 2-1/4 Cr-1 Mo steel.

4. A fouling factor equivalent to a heat transfer coefficient of

2 2
11,360 W/m •°C (2000 Btu/hr-ft -°F) (economizer, evaporator) and
113,600 W/m^* °C (20,000 Btu/hr-ft^-°F) (superheater) was used in

the calculation of the overall heat transfer coefficient. In

addition, as part of initial sizing, 15% excess was added to the
required surface area to account for uncertainties.

Table 8-2 gives the sizes and characteristics of helical coil steam
generators with bundle diameters of 2.59 m (8.5 ft), 2.74 m (9.0 ft), and
3.05 m (10.0 ft) and tubing material of Incoloy 800H or 2-1/4 Cr-1 Mo
steel. By varying the transverse tube spacing, it is possible to fulfill
the helium/water pressure drop requirements with various bundle diameter/
length combinations.

Table 8-3 shows the sizes and characteristics of straight tube steam
generators using Incoloy 800H or 2-1/4 Cr-1 Mo steel as tubing material.
Since the tube size and spacing are fixed, only one bundle diameter and
length, for each tubing material, fulfill the allowable helium pressure
drop requirement.* Furthermore, the units shown are based on the closest

*The water side pressure drop is very small and not limiting.

8-14
 TABLE 8-2 , .
300-MW(e) GCFR STEAM GENERATOR CHARACTERISTICS (HELICAL)

Superheater Evaporator Economizer Tube

Bundle Bundle Spacing, Spacing, Diameter x Diameter x Diameter x Surface Material Tube Material Tube
Diameter Length No. of Transverse Longitudinal Thickness Thickness Thickness Length Area Volume Density Weight
(m (ft)] [m (ft)] Tubes [mm (in.)] [mm (in.)] [ram (in.)] [mm (in.)] [mm (in.)] Material [m (ft)] [m2 (ft2)] [m3 (ft3)] [kg/m^ (lb/ft3)] [kg (lb)]

3.0 6.4 180 35 38/32 25 x 1.7 25 x 1.7 19 x 1.2 Incoloy 194.2 2,422 3.34 31.2 26,894
(10) (21) (1.38) (1.5/1.25) (1 x 0.0676) (1 x 0.0676) (0.75 x 0.047) 800H (637.2) (26,065) (118.1) (501) (59,163)

2. 7 9.8 178 43 38/32 25 x 1.7 25 x 1.7 19 x 1.2 Incoloy 190.8 2,364 3.27 31.2 26,325
(9) (32) (1.71) (1.5/1.25) (1 x 0.0676) (1 X 0.0676) (0.75 x 0.047) 80 0H (625.9) (25,434) (115.6) (501) (57,915)
00
£ 1-'

2.6 12,8 176 51 38/32 25 x 1.7 25 x 1.7 19 x 1.2 Incoloy 188.4 2,306 3.20 31.2 25,710
(8.5) (42) (2.02) (1.5/1.25) (1 x 0.0676) (1 x 0.0676) (0.75 x 0.047) 800H (617.9) (24,813) (112.9) (501) (56,563)

3.0 7. 3 260 36 38/32 25 x 4.4 25 x 2.2 19 x 1.8 2-1/4 Cr- 144.9 2,674 6.65 30.5 52,202
(10) (24) (1.43) (1.5/1.25) (1 x 0.172) (1 X 0.0864) (0.75 x 0.0715) 1 Mo (475.2) (28,774) (234.6) (489.5) (114,846)

2. 7 11.3 260 46 38/32 25 x 4.4 25 x 2.2 19 x 1.8 2-1/4 Cr- 143.5 2,602 6.51 30.5 51,112
(9) (37) (1.80) (1.5/1.25) (1 x 0.172) (1 X 0.0864) (0.75 x 0.0715) 1 Mo (470.6) (27,995) (229.7) (489.5) (112,448)

2. 6 15.5 260 56 38/32 25 x 4.4 25 x 2.2 19 x 1.8 2-1/4 Cr- 141 .0 2.557 6.40 30. 5 50,244
(8.5) (51) (2.20) (1.5/1.25) (1 X 0.172) (1 X 0.0864) (0.75 x 0.0715) 1 Mo (462.4) (27,510) (225.8) (489.5) (110,538)

^a^Updated reference design cycle: helium temperature = 543°C (1010°F) ins 342°C (648°F) out; water temperature = 208°C (406°?) in, 513°C (955°F) out; steam generator
AP-heliutn = 0,038 MPa (8.5 psi); total, water = 1.89 MPa (275 psi); tube wall thickness based on AP; heat duty = 291 MW(t) (per unit).
 TABLE 8-3 , .
300-MW(e) GCFR STEAM GENERATOR CHARACTERISTICS (STRAIGHT)

Tube Tube
Bundle Bundle Tube Diameter x Surface Material Tube Material Tube
Diameter Length No. of Pitch Thickness Area Volume Density Weight
[m (ft) ] [m (ft) ] Tubes [mm (in.)] [mm (in.)] Material [m2 (ft2)] [m^ (ft^)] [kg/m3 (lb/ft3)] [kg (lb)]

1.96 18.0 5429 25 16 x 1.0 Incoloy 4,875 4.65 31.2 37,370

(6.45) (59.05) (1.0) (0.625 x 0.040) 800H (52,452) (164.1) (501) (82,214)

1.93 16.8 5262 25 16 x 3.2 2-1/4 Cr- 4,418 11.36 30.5 89,177
(6.35) (55.2) (1.0) (0.625 x 0.127) 1 Mo (47,529) (400.8) (489.5) (196,191)

(A\
K "^Updated reference design cycles: helium temperature = 543°C (1010°F) in, 342°C (648°F) out; water temperature =
208°C (406°F) in, 513°C (955°F) out; steam generator AP-helium = 0.058 MPa (8.5 psi) total, water = 0.083 MPa (M 2 psi);
tube wall thickness based on AP; heat duty = 291 MW(t) (per unit).
practical tube spacing, i.e., 25.4 mm (1 in.) triangular pitch for 16 mm
(5/8 in.) diameter tubes, and thus represent the shortest possible length.

Figure 8-4 shows bundle length as a function of bundle diameter for

both helical and straight tube steam generators. For the helical coils,
note the sensitivity of bundle length to bundle diameter; i.e., on the
average, the bundle length change is about 16 times as great as the
diameter change. In addition, for the same bundle diameter, the bundle
length required for 2-1/4 Cr-1 Mo tubing is, on the average, about 17%
greater than that for Incoloy 800H tubing. Also shown in the figure are
the straight tube bundle sizes, which show, in contrast to the helical coil
units, that the required bundle length is less for 2-1/4 Cr-1 Mo than for
Incoloy 800H. This condition results because the thinner tube wall for
Incoloy 800H provides much greater water flow area and, consequently, a
lower water side (and overall) heat transfer coefficient than is the case
using 2-1/4 Cr-1 Mo steel tubing.

Figure 8-5 shows the tubing weight and surface area as a function of
bundle diameter for both the helical coil and straight tube steam genera­
tors. Referring to the right side of the figure (helical coil units), it
can be seen that the surface area required for 2-1/4 Cr-1 Mo tubing is
slightly greater than that for Incoloy 800H tubing. However, the tubing
weight is nearly twice as great for 2-1/4 Cr-1 Mo steel as for Incoloy
800H. This results from the need to add a corrosion allowance to the 2-1 /4
Cr-1 Mo tube wall thickness and from the much lower allowable stress for
2-1/4 Cr-1 Mo steel in the high-temperature (superheater) section. The
left side of Fig. 8-5 shows the straight tube steam generator surface areas
and weights. Note that the required surface areas for the straight tube
units are approximately twice as great as those for the helical units.
Similarly, the tube weights for the straight tube units are 42% to 75%
greater than the comparable tube weights for the helical coil units.
Because a uniform tube wall thickness is used throughout a straight tube
steam generator, based on the highest temperature, the tube weight is about
2-1/2 times as great for 2-1/4 Cr-1 Mo steel as for Incoloy 800H, even
though the latter unit has slightly greater surface area.

8-17
 • INCOLOY 800H

■ 2-1/4 Cr - 1
BUNDLE LENGTH [m (FT)]

INCOLOY BOOH

OD HELICAL COIL
• ■ STRAIGHT TUBE

BUNDLE DIAMETER [m (FT)]

Fig. 8-4. Bundle length versus bundle diameter for helical coil and
straight tube steam generators

8-18
 90.8
( 200)
■ 2-1/4 Cr
WEIGHT

O □ HELICAL COIL

• ■ STRAIGHT TUBE
LB)]
KG (10

• INCOLOY
10

800H AREA
[
TUBING WEIGHT

SURFACE AREA
■ 2-1/4 Cr - 1 Mo
AREA

2-1/4 Cr - 1 Mo
WEIGHT

[
10'

• INCOLOY 800H
WEIGHT
2-1/4 Cr - 1 Mo
AREA

AREA
INCOLOY 800H
WEIGHT

(6) (7) (8) (9) (10)

BUNDLE DIAMETER [m (ft)]

Fig. 8-5. Tubing weight and surface area versus bundle diameter for
helical coil and straight tube steam generators

8-19
 Despite the greater tubing weight required for steam generators using
2-1/4 Cr-1 Mo steel, the tubing cost is significantly less than that for
Incoloy 800H, as described in Section 7.2.

8.3. STEAM GENERATOR ARRANGEMENTS IN PCRV

Based on the steam generator sizing study described in Section 8.2,

several possible steam generator arrangements in the PCRV were prepared.

Figures 8-6 and 8-7 show the range of helical coil steam generator
sizes, using Incoloy 800H and 2-1/4 Cr-1 Mo steel, respectively, and their
possible placement in the PCRV. Note the wide variation in bundle length
that is possible for the bundle diameter range studied. This desirable
characteristic of helical coil steam generators permits wide latitude in
the placement of the steam generators in the PCRV, the location of the
helium circulator, and routing of the helium ducting so as to obtain the
most favorable integrated system design.

Figure 8-8 shows the straight tube steam generator configurations,

using both tubing materials. It can be seen that their great length limits
their arrangement possibilities in the PCRV. For example, the straight
tube steam generator lengths are such as to require a separate cavity for
the steam generator; in contrast, the shorter helical coil steam generators
permit the circulator to be mounted in the same cavity as the steam
generator.

Using helical coil steam generators with Incoloy 800H tubing, more
detailed layouts were prepared, with emphasis on the shortest bundle and
the longest bundle. Figure 8-9 shows Alternate A,* which has the same
bundle dimensions as the reference design unit, is top flange mounted, and
has two (one feedwater, one superheat), instead of four, penetrations at
the bottom of the steam generator cavity. The circulator, of the present

*This designation refers to the reference design, with resuperheater

and four bottom penetrations, as a basis.

8-20
 TOP Of PCJSV— TOP OP PCPV TOP 0/= PCRV

TOP Of

JOPPOPT P14N<$£- -O' SUPPORT'SLANG*

toi) SOPPOPT PldMCf

8-21

24.7&

STM 6 BN

BOTTOM Of PCPV BOTTOM Of PCRV BOTTOM Of PCPV

S.S x 42
(317™ *€.4mj (2- 7m x9.<Sm) (2.£m X/2.Sm)
ST 2AM GErNt&ATOaS

Fig. 8-6. Helical coil steam generator configurations (Incoloy 8Q0H)

 TOP Of PCPV- TOP OT PCPV

TOP Of BUNDLf

TOP Of 3UMDIB
TOP Of 6UA/DLP
8-22

STM CfN

BOTTOM Of POZV BOTTOM Of PC4V

ia\ 24' 9' X 37 <5-5 X

(3.Om ><7.3mT} (f- 7m x f/.3m) (^2. Om #
SHAM CtNtRATOgS

Fig. 8-7. Helical coil steam generator, configurations. (2-1/4 Cr-1 Mo)
 TOP OF PCgV £- TOP OP PCPV

TOP OP Bcj/VDLP 'k-

r— TOP OP 31/PJDLP-

55'
(/6.8m)
Sl/PPOIZT
plangb-

A'S’l.tUM

30 '
8-23

-- - STM TPNPN&ZATOP — STPAM GBMBSATOp

(STTAlGUf TU3B- ) (STPPJCPT ToSP)

BOTTOM OT PO?V ^ BOTTOM OP PCTV

To,4b X 59'Cs.Om x/s.On) Ts,3b' x 5 5 (7.9/h x/6.8m,)

( /NCOLOY 800 H ) ( 2'M CP - / MO )
5TbAM 0£-NbgATO£?$

Fig. 8-8. Straight tube steam generator configurations

hour-glass diffuser design, is mounted in the concrete closure plug above
the steam generator.

Figure 8-10 shows Alternate B, which is similar in bundle size to

Alternate A but is mounted at its center of gravity and uses an external
duct to return the helium from the steam generator to the circulator
plenum.

Alternate C, shown in Fig. 8-11, is a near-maximum-length unit, sup­

ported at its center of gravity. Feedwater enters at the bottom, and
superheated steam is taken from the top through a tubesheet mounted in a
concrete closure plug. The reversed flow main circulator is installed in a
separate cavity using a steel primary closure. Exit gas from the steam
generator is ducted to the bottom of the circulator cavity, and after
passing through the circulator compressor, it is directed via a separate
duct into the core inlet plenum. The separate cavity offers a number of
advantages including accessibility and ease of removal. The additional
cross duct from the separate circulator cavity to the reactor cavity
compounds the already considerable congestion of ducts and tendons, and
although there are a number of possible solutions, more extensive layout
and analysis will be required to fully confirm the feasibility of this
arrangement.

An arrangement for a center-mounted straight tube steam generator is

shown in Fig. 8-12. Feedwater enters at the bottom and superheated steam
exits from the top. Thermal expansion of the tube bundle is accommodated
by a movable tubesheet. Although not shown in this preliminary design, the
vertical thermal growth of the shell may be accommodated by bellows or by
external pipe loops.

REFERENCES

8-1, MacMillan, J. H., and T. 0. Johnson, "Once-Through Performance at 3800

MWT - The Babcock-241 Standarized Nuclear Steam System," Proceedings of
the American Power Conference, Vol. 36, 1974, p. 117,

8-24
8-2. McDonald, B. N., R. C. Post, and J. S. Scearse, "Once-Through Steam
Generator Research and Development Report-Supp. 1-Integral Economizer
OTSG," Babcock & Wilcox Report BAW-10027, Supp. 1, September 1972.
8-3. Whipple, J. C., and C. N. Spalaris, "Design of the Clinch River
Breeder Reactor Plant Steam Generators," Nucl, Technol. 28, No. 3,
305 (1976).
8-4. Harty, R. B., "Modular Steam Generator Final Project Report," Atomics
International Report AI-TR-097-330-010, September 1974.
8-5. Robin, M. G., "French Steam Generator Experience-Phenix and Beyond,"
Nucl. Technol. 28, No. 3, 482 (1976).
8-6. Swanson, L. L., "GEN-SIZE, A Once-Through Steam Generator Design
Code," General Dynamics, General Atomic Division Report GAMD-7145,
March 1966.

8-25
 (25.5 m)
£5 ' & C), a

rJr&CWd ;

■572^1/ Gzs/frZAra.#.

SUS^m

TOP OP
PCPV

5 ru3B 5£SPPO#r PJ.PTP5

SPACPO
p- cpatpp supoc/o

Of"4P'/4PIPS (/5.2cm)
iso rune-s- 24- s,ows
(4.0m) 1.575" SPAC&S COA/rOp
TO ceNTOP

oufm s//4auo
(/■?cm) .75"MIL AJVO /.OO TK
/NSULATiON (2.5cm)

/NSUL4nO,V

ttcr.'OA)
SCALP ' ,/'2"z/iO"

Fig. 8-9. Steam generator arrangement

Alternate A (helical coil)
 ■5 rase SUPPORT PIA TE-S

130 ruSBS - 24 Rons

(4cm) l576"SPdC6S cenrM
ro ceNfZR

j- CS/ur£-R SHROUD

^-Qurs/De- SHROUD
.76 W4LL /‘INSULATION
(19 cm) (Z.Scm)

/NSULA r/ON

SECT(ON
SCAUr: V= lL0"

Fig. 8-10. Steam generator arrangement

Alternate B (helical coil)
 {24. 8 m)

r" C£,vr£Z SUZOUO

s—ourtp suAoaD
7S“TM/CK (/.Jim)

to
rusts ■ IS AOtYS SPACtf
POYY P/O / = 3'-0" D/A (O/m)
AOW 4/0/6 = S'-5" D/A (2.Sm)

SUPPORT TiAN'Ct

ShSCUO ///SO-AT.O/U
/ ” TP/CK {Z.Scm)
i.d. duct

cpv/rr /msu^pt/o/t

rust- sopposr plats

<? SO’ /NrtP V4LS

refi op PtL HUM /Nl tr DOCr

PCA.V

SCAof^yp^C

OS-TAIL. A
{plTTTUaSTT CONP.'DuZPTJGA'
W/TH Lt-Yt-L OUCT ADD
JNCRcPStC P&NerCAT/OM /.D.)

Fig. 8-11 Steam generator arrangement

Alternate C (helical coil)
 (/8.0m)
rap op pcrv 35 - 3

£ ■ 2-0 0/4 -/q!-8"i.d. CAwry ■•SUPPORT PUANQSr

(3.3m)

£ STfAM GjtNS-RATOR

i/usuiaT/ON

■— s-4"i.d. cAV/ry
(2.5*) ■5 - o'-a" o/4 (.g>im)
QuruPT PO/ZTS
6 '-si 0.0. BUNDLE

tube spacers

ouct

£L£-VA TION

Fig. 8-12. Steam generator arrangement

(straight tube)
 9. CONCLUSIONS

1. Steam Generator Geometry. Although Its total cost, including

material and labor, tends to be greater, the helical coil tube
steam generator possesses significant advantages compared with
the straight tube steam generator. For GCFR requirements these
include (1) a wide variation in bundle length, for a practical
range of bundle diameters, which permits considerable latitude in
the placement of the helium circulator and helium ducting and, if
desired, simple incorporation of a resuperheater section; (2)
significantly less weight and surface area; and (3) simple pro­
visions for accommodation of tube thermal expansion. For these
reasons, and others described in this report, the helical coil
tube steam generator geometry is recommended for GCFR appli­
cation.

2. Materials. Incoloy 800H is the present reference tubing mate­

rial, and 2-1/4 Cr-1 Mo steel is under consideration as a substi­
tute. Numerous advantages would accompany the use of 2-1/4 Cr-1
Mo steel, such as reduced cost, high resistance to stress cor­
rosion cracking in chloride- and caustic-contaminated water/
steam, and a low alloy content which might keep future avail­
ability high and cost low. If near-future tritium permeability
tests of 2-1/4 Cr-1 Mo steel demonstrate acceptably low levels,
then this material is recommended for use as steam generator
tubing.

3. Boiling Behavior. The critical heat flux behavior of a coiled

tube differs considerably from that of a straight tube. Whereas
critical heat flux occurs in one location in the evaporator of a
straight tube, it can occur over a considerable length, in dif­
ferent peripheral locations, in the evaporator section of a

9-1
 helical coil. This non-uniform critical heat flux behavior
produces an asymmetrical temperature distribution on the tube
periphery and should be evaluated (a) analytically, for tube
thermal stress effects, and (b) experimentally, if the need is
demonstrated, for possible effects on corrosion behavior and
tube life.

4. Low Flow Boiling Stability. As part of the safe shutdown proce­

dure for the GCFR, it is required that the steam generator oper­
ate stably and produce high-quality steam at flows as low as 2%.
Experimental evidence, at least for vertical straight tube steam
generators, demonstrates that such operation is readily attain­
able. However, the non-uniform critical heat flux behavior of
coiled tubes, combined with the fact that a GCFR steam generator
coiled tube is nearly horizontal, can result in low flow boiling
characteristics that are substantially different from those
observed for vertical straight tubes. An experimental/analytical
program is required to establish the low flow boiling stability
characteristics of a GCFR type steam generator coil.

9-2