AEEW-R956.txt

T Na

AEEW - R 956 =~~~
NOT-FOR-PUBLICATION (Commercial)

Safeguard this document as directed overleaf

 

UNITED KINGDOM ATOMIC ENERGY AUTHORITY

Reactor Group

AN ASSESSMENT OF A 2500 MWe MOLTEN
N . CHLORIDE SALT FAST REACTOR

EDITED BY:
J SMITH
W E SIMMONS

 

CONTRIBUTORS:
DR R C ASHER, AERE
DR G LONG, AERE
DR H A C MCKAY, AERE
DR D L REED

 

Technical Assessments and Studies Division

 

Atomic Energy Establishment, Winfrith, Dorchester, Dorset, 1974
NOT 1POR PUBLICATION (COMMERCTAL)

 

AEEW-R 956

AN ASSESSMENT OF A 2500 MWe MOLTEN

 

CHLORIDE SALT FAST HEACTOR

Edited by J Smith

and

W E Simmons

Contributors:

Technical Assessments
and Studies Division,
AEE Winfrith

Dr

Dr

Dr

Dr

R C Asher, AERE
G Long, AERE
H A C McKay, AERE

D L Reed, AEE

 

August 1974

 

In the 1interest of paper
W 10774 economy this document has
been printed to a reduced
standard.

 

 
= Nl

1

NOT FOR_PUBLICATION (COMMERCIAL)

CONTENTS

INTRODUCTION
MSFR CONCEPTS
CHEMICAL ASPECTS OF THE FUEL
REACTOR PHYSICS
SAFETY AND OPERATIONAL ASPECTS
5.1 Normal Operation
5.1.1 Startup
5.1.2 Power Control
5.1.3 Hotspots
5.1.4 Shutdown
5.1.5 Small Leakages
5.2 Fault Conditions
ENGINEERING DESIGN PHILOSOPHY
.1 Choice of Materials
Fuel Inventory
Choice of Intermediate Coolant
Choice of Power Plant
Blanket Inventory and Cooling

Reliability and Maintenance

o O O O O O Ov
Oy W S\ no

.{ Engineering Development
OUTLINE SYSTEM DESIGNS AND AUXILIARY PLANT
7.1 General Aspects
7.2 The Indirectly Cooled Concept

7.3 Design Variations of the Indirectly
Cooled Concept

7.4 Special Auxiliary Plant

7.5 PFilling, Draining and Dump Systems

il

Prgne No

N =W

11
11
11
11
11
12
12
14
14
16
17
18
19
19
20

21

25
24

25
25
NOT FOR PUBLICATION (COMMERCIAL)

CONTENTS Page No

V.00 lmerpeney Cooling System 20

7.7 Salt Cleanup and Gaseous Fission 20
Product Removal

7.8 Processing Plant 28

7.9 Remaining Plant 28

7.9.1 Containment 28

7.9.2 Preheating 29

7.9.5 Shielding 29

7.9.4 Pumps 29

7.9.5 Steam Plant 29

8. MATERIALS 50

9. COSTS 32

9.1 Capital Costs 32

9.2 Fuel Costs 33

9.3 Overall Generating Costs 35

10.  CONCTHUITON 35

11. PARAMETERS FOR TNDIRECT AND DIRECT COOLED VERSIONS 37
OF A MOLTEN SALT FAST REACTOR - SUMMARY

REFERENCES 45
ACKNOWLEDGEMENTS 46
APPENDIX I b7

IT 53

iii
4,

no

la

1b

Y

4p

NOT FOR PUBLICATION (COMMERCIAL)
TABLES

COMPARISON OF PERFORMANCE OFF REFERENCE
MOPR WITH OTHER FAST REACTORS

2500 MW(e) NET MSIFR: SUMMARY OF FEATURES

MSEFR DIRECT LFAD.CURTAIN COOLED SYSTEM - PARAMETERS
NUCLEAR PERIFORMANCE OF A PRELIMINARY UKAEA DESIGN
Ol AN INTERNALLY COOLED 6000 MW(t) MSFR USING

CL37 SALT

MSFR INTiERNALLY COOLED DESIGN PARAMETERS

FIGURES

DESIGN 1. INDIRECTLY COOLED MSFR - LEAD COOLANT

Pafe No
10

22
49
56

57

ALTERNATIVE DESIGNS FOR INTERMEDIATE HEAT EXCHANGERS

FOR LOW CONDUCTIVITY FUEL SALT

DESIGMN 2. TNDIRECTLY COOLED MSFR - LEAD COOLED WITH

ALTERNATIVE CORE AND INTERMEDIATE HEAT EXCHANGER
LAYOU'L.

DESIGN 5. INDIRECTLY COOLED MSFR - PRELIMINARY DESIGN

O HELIUM COOLED VERSION.

DESIGN 4. DIRECT COOLED MSFR WITH JET PUMP ASSISTLD

CIRCULATION

SIMPLIFIED DIAGRAM OF DIRECT SYSTEM SHOWING LEAD
CURTAIN AND POLOIDAL FLOW

GENERAL VIEW OF 2500 MWe MSFR (INDIRECT SYSTEM BASED

ON DE3IGN 1)

6 &7 MSFR RBUIILDING LAYOUT (BASED ON DESIGN 1)

8

9
10

11

12

SIZE COMPARISON OF MSKFR WITH CFR AND MSBR
FLOW DIAGRAM -~ INDIRECT SYSTEM

"I,OW DIAGRAM-I'UEL AND BLANKET SALT CLEANUP AND OFF-
GASSING

FLOW DIAGRAM - DIRECT SYSTEM

INTERNALLY COOLED MSKFR

iv
NOT FOR PUBLICATION (COMMERCIAL)

 

1. Introduction

As a result of some initial studies in 1964 and 1965,
it was concluded that the line of investigation of most interest
to the UK on molten salt reactors and which would also be
¢ omplementary to the US investigations would be on a fast reactor
version. A preliminary study of a fast system using the U233%/Th
cycle and fluoride salts did not indicate encouraging results and
it was therefore decided that a Pu/U238 cycle would be examined.
This 1nvolved the use of chloride salts and work on salt chemistry
began in 1965 which in 1970 was extended to include additional
materials aspects, The assessment and fluid flow study work was
carried out mainly in 1971 and 1972. This report presents an
overall summary. It is in the nature of a survey report as the
limited effort available has meant that the depth to which questions
could be investigated has been restricted.

The initial question which is bound to be asked is why
consider a fluid fuel, with all the implications of a highly
active circuit If it is considered that the ultimate refine-
ment of the solid fuel system will appear as some form of fast
reactor, then to make further progress beyond this, some basic
change 1n concept has to be made, and it seems clear that the
most fundamental would be to escape from the trammels of fuel
fabrication with fine tolerances, expensive active transport
and also from central processing if suitable on-line methods
can be developed to treat the fuel directly.

A basic safety feature, in principle, is that a fluid
fuel can be disposed of from the reactor to ever-safe containers,
while good negative temperature coefficlents should limit
excursions and potentially offer a self-regulating system. The
current feasibllity studies have sought to demonstrate how far
these potentialities might be realised and to give a preliminary
appraisal of the economic and performance prospects. The table
below summari ses the possible advantages and disadvantages which
have become apparent as the investigations have been in progress,
and it is hoped this will form a useful reference against which
to judge the success achieved to date.

Table I - Advantages and Disadvantages of Molten Salt Fast
Reactors (MSFRs)

 

 

I. Potential Advantages

1. No loss of neutrons occurs in fuel cladding or
structural material in core,

2. Continuous addition or removal of fuel is possible
under power,

3. Simple control of power may be achievable by coolant
conditions and power demand. It seems possible to
avoid control rods or equivalent systems.
NOT IPOR PUBLICATION (COMMISKCEAL)

 

b, Strong negative temperature coefficient gilves
basic safety.

5. High heat capacity of fuel restricts temperature
rise on loss of normal cooling. A dump system
with independent cooling system can be used for
decay heat removal,

6. Reactor vessel is small. Prefabrication of the
vessel at works should be possible, thus reducing
construction time and interest charges. In the
event of a fault in the vessel fuel may be removed
and the vessel size is such that its replacement
can be contemplated.

7. Building layout is compact due to the small size
of the reactor vessel and primary circuit and
absence of elaborate fuel handling route.

8. The system can basically be at low pressure apart
from pumping pressures.

9. Fuel handling by pumps and pipework should be a
simpler operation than solid fuel handling.

10. The system is sulted to on-site close coupled fuel
processing.

11. A substantial proportion of the fission products
can be continuously removed by a close coupled
process.

12, Savings in fuel element fabrication and development

costs seem possible,

—

13. Potentilal for high temperatures and higher cycle
efficiency 1s good.

Disadvantages

 

1. Higher fuel inventory than in LMFBRs is required
to give heat transfer and transport.

2. Molten salt fuel has relatively poor heat transfer
characteristics compared with sodium.

3. The high melting point of suitable fuel salts (v5E0 )
involves pre-heating, and heating to prevent freezin®
during prolonged shutdowns. The femperatures involv~d
prevent access to plant.

b, The high melting point of the fuel salt also 1mposes
an additional constraint on heat exchanger systems in
order to prevent salt freezing on heat transfer
surfaces 1f low return salt temperatures are
employed, especially i high heat transfer rates
apply on the coolant side.

2
NOT FOR PUBLICATION (COMMERCTAL)

 

5. Fresence of fission products in the fuel salt - a
hhimh standard of plant reliability and lenk
t,irhtness with remote maintenance i1is required,

6. Limitation of choice of materials of construction
due to corrosion and high temperature,

2. MSFR Concepts

 

In the main part of this report two versions of the MSFR are
considered, which have been designated as the Direct Cooled and
the Indirect Cooled systems., Figures la and 1lb illustrate the
second of these, where the fuel salt is pumped through a core
vessel of dimensions which permit criticality and in which the
fission heat is generated. It then flows through heat exchangers
on the secondary side of which is circulated a coolant which
transfers the heat to the steam generators. This is the classical
pattern of fluid fuel designs and was the form originally
considered. However, the investigations at that time considered
the circuit hold up and fuel inventory for the layouts examined
to be too high and an alternative was sought, emanating from a
suggestion by Bettis of ORNL,

In this, the Direct Cooled version (See Figures la and Ub),
the salt is both cooled and circulated within the core vessel
(in what is known as poloidal flow) by a curtain of lead drops
injected down the periphery of the vessel. This in principle
gives low fuel hold up because no external inventory except
that for auxiliary circuits is required. The investigations on
this type have focussed on materials aspects and operating
conditions, and importantly also, the nuclear and thermal
performance, In the course of the study of heat transfer and
fluid flow features, it became apparent that there could be
substantial limitations on the heat removal capability of the
direct scheme, and further problems in achieving the necessary
separation of lead and salt at the region where the lead must
pass out of the core vessel to the heat exchangers,

For this reason, the Indirect system has been re-examined
in the later part of the study, and appears to be of considerable
interest, provided that the compact layouts and high heat exchanger
volumetric ratings now postulated can be achieved in practice, thus
giving a reasonable fuel inventory. The main emphasis in this
report is concerned with this latest Indirect version, while details
of the Direct system are given in Appendix I.

During the course of this work, a report by Taube (1) became
available which suggested a further alternative in which the heat is
r emoved from the core salt by a heat exchanger system within the
core., The coolant is a salt of the same composition as the blandket
salt. A preliminary appraisal of this scheme is given in Appendix
IT.

The reactor cases studied have all been of 6000 MW thermal
output. This (equivalent to about 2500 MWe) was chosen as typical
of the size of the unit in a large grid system in the later part
of the century.
NOT TPOR PUBLTICATION (COMMERCTAL)

 

3. Chemical Aspects of the Fuel

 

The concept of employing a molten salt as the combined
fuel and primary cooclant of a nuclear reactor presents many
novel problems in both reactor design and system chemistry.
The achievements of the ORNL Molten Salt Reactor Programme
both give confidence in the feasibility of a molten salt
concept and lend support to the view that in many important
chemical respects a molten salt system can be regarded as
being in a state of chemical equilibrium. Much progress
can then be made by considering the various equilibria which
might occur both within the salt and between the salt and
its environment, particularly since many of the equilibria
involved are predictable, at least in broad outline.

When selecting a salt mixture for use in the core and
blanket of the NSFR many factors, chemical and nuclear, must
be taken into account. The reference core salt NaCl:UCl,:PuCl
containing in the region ofr 30 to 40 mole % of heavy atoés
(50 to 55 welght?) fulgils the relevant criteria, including
melting point (550~5807C), nuclear properties and chemical
stability. Should occasion demand, variations are possible
without detracting from its essential properties as a fuel
by replacing some or all of the sodium chloride with potassium,
magnesium or calcium chlorides.

5

For the feasibility of the concept it is essential that
under all 1likely conditions throughout the 1life of the reactor
the core and blanket salts should, firstly, be compatible with
the container materials and secondly, should nct undergo
reactions which bring about precipitation, especially of the
plutonium, within the reactor. Compatibility is achieved in
the initially pure fuel salt by ensuring that materials for
the containment are selected from those which form relatively
unstable chlorides, such as iron, nickel and the refractory
metals. Since the chlorides in the fuel salt are of high
stability little reaction with these metals 1s then expected,
and this 1s borne out by experiment. During reactor operation
additional chemical species are introduced into the salt and
processes are envisaged in order to maintain their concentra-
tions at some acceptable equilibrium level, Of the important
fission products, some are gaseous and can be removed
¢ ontinuously, some, notably Rb, Cs, Sr, Ba, Zr and rare
earths, farm chlorides and remain the salt while others,
notably Mo, Ru and Pd are unstable as chlorides and will
precipitate from the salt as metals. Experience in the MSRE
suggests that these will in part be discharged as a fog with
the 1inert gases and partly plate out on metal surfaces. In a
d irect cooled system an additional possibility is dissoclution
in the lead coolant.

While it is possible that certain individual fission products
might take part specific corrosion reactions, as chemical
entities the majority of fission products should not significantly
modify the corrosion behaviour of the salt and their concentra-
tion will be maintained at a level acceptable to neutron economy
by chemical reprocessing.
NOT TOR PUBLICATION (COMMERCIAL)

During fission, however, there is mis-match between the number
of chlorine atoms avallable from the fissioned chloride and the
number taken up by the fission products. The excess chlorine
produced reacts with the strongest reducing agent present,

UCl,, forming UCl In the pure form this is highly corrosive,
but’predlotlon anfi experiment both show that when UCl, is
dissolved in fuel salt at low concentrations little a%tack of

c ontainer metals is to be expected, The concentration of UCl

is readily maintained at acceptably low levels by reacting the
fuel salt at a modest rate with, for example, metallic uranium.
By a similar argument, supported by ORNL irradiations of
fluoride fuel sats at comparable heat ratings, UCl, is expected
to react rapidly with any short-lived oxidising spécies produced
under the intense fission fragment irradiation of the salt.
Neither evolution of chlorine from the melt nor enhanced attack
of the container materials under irradiation are therefore to be
expected.

Other specific chemical spec1es which are of interest in

comޤt1b111+y will be gresent in the melt. The (n,p) reaction
Cl will produoe S at a mean concentration (up to a few

thousand ppm in the salt) which depends upon the fuel rating and
the isotopic concentration of the chlorine, Account must also
be taken of the effect of introducing oxide into the melt by
1ngress of air or water vapour. The UCl, component of the fuel
is expected to combine with both the sulahur and the oxygen and
so reduce their reactivity towards structural materials,
Experimental verification of this point is required, particularly
for such materials as nickel which are knwon to be susceptible
to sulphur attack.

Consideration of the stability of corrosion product chlorides,
backed by 1aboratory experience, lead to the conclusion that
alloys based on iron, nickel and the refractory metals such as
molybdenum will be compatlble with chloride fuel and blanket
salts in which the UCl, contents is maintained at a low level,
less than a few peroen% of the UCl, content. More reactive
metals such as chromium could 1n aédltlon be admitted as minor
constituents in an alloy provided some surface leaching were
allowed for. When lead is also present, as in the direct cooled

system, the lead itself is compatible with fuel salt, but the
choice of container material is more restricted. Nickel alloys
are excluded, while only a limited range of iron alloys is
possible. The presently preferred container for the two liquids
together is molybdenum, either as bulk material or in the form
of a protective clad on an iron-based structure.

The fission process produces chemical species which might
lead to precipitation of components from the fuel salt. These
include fission products reacting to give complex chlor%ges
(e.g. Cs,.UCl,) or specific ompounds (e.g. UI,) and the
forming %ulpgldes (US). Although not all the relevant solubilities
are known it is likely that uranium compounds will precipitate
in preference to the corresponding plutonium compounds.
Preliminary measurements have conflrmeotmat the material precilpitated
from a fuel salt containing both uranium and plutonium is largely
uranium sulphide.
NOT T"OR PUBLICATION (COMMERCTAL)

 

The solubility of sulphur, however, is much less than the amount
estimated to be formed in a natural-chlorine fuelled systom.

A low solubillity of sulphur, or of any of the other uranium-
bearing species, would not adversely affect the feasibility of

t he system. DBy a simple process - such as adjustment of temper-
ature and UCl, content - precipitation could be induced in a

C Lean=-up circdit. Since the species concerned are being produced
at a predictable rate the concentrations in the core could be
safely maintained close to saturation and even an inefficient
removal process would suffice. By contrast the adventitious
admission of oxygen is likely to be unpredictable. It is there-
fore necessary firstly to ensure that the fuel salt is capable
of dissolving the amounts of oxygen which might conceivably be
admitted and secondly to keep the normal oxygen content of the
melt well below the saturation level, so maintaining the capacity
of the melt to dissolve additional oxygen. Measurements of the
oxide solubility have demonstrated an adequate capacity for
dissolving oxide, but have also shown that only a small decrease
in solubility can be expected when conditions are changed from
those normnally obtaining in the reactor. The required degree

of oxygen removal is therefore not possivle by, for example,
simple adjustment of the temperature, and alternative chemical
methods have been sought. Precipitation of oxygen as alumina

by reaction with aluminium, introduced into the salt as the
liquid sodium chloraluminate, has been studied in some detail.
This reaction promises to provide not only an efficient method
of oxide removal but also a convenient means of converting
uranium and plutonium oxides produced in the reprocessing cycle
into fuel or blanket salt., In further tests the prediction

that from melts containing both UCl3 and PUCl. uranium is
precipitated as oxide in preference-to pluton%um has been
confirmed over a range of relevant conditions.

In addlition to the adjustment of the chemistry of the fuel
and blanket salts on a fairly rapid cycle as just described
(which will, incidentally, also serve to remove any inert gac
fisslon products not removed in the cover gas), and defined
as clean up 1t is necessary to carry out more drastic processes,
probably on a slower cycle defined as processing, whose basic
purpose 1is to remove fission progycts from the core and plutonium
from the blanket. If separated Cl is used, this must be recycled
without appreciable loss; and may cause prablems in the reconversion
of heavy metal to chlorides when excess chloride is required.
It 1is also desirable to recycle the sodium in the fuel salt to
avold a radiocactive disposal problem.

Two contrasting schemes have been considered to meet these
objectives, one based on mainly well-established aqueous processes
and the other on a series of pyrochemical steps. each of which
has been tested in the laboratory. For the aqueous route it is
necessary to develop processes to convert the reprocessed heavy
atom content to chloride, as well as, probably, to recover the
sodium chloride for recycling. The pyrochemical route uses
a series ol metal displacement reactions in molten chloride
media., The feasibility of such a process has been demonstrated
at ANL on a pilot-plant scale for oxide fuels, but much
engineering development is required to evolve a workable and
reliable process, especially in view of criticality restrictions.

6
NGT FOR PUBLTICATTON (COMMERCTAL)

 

The process can be operated close-coupled to the reactor and

on a much shorter cycle than the aqueous route, and the plant 1is
comrnct. At the present time there is insufficient information
to make meaninpful cost estimates for this process but the

prel iminary work that has been done indicates that capital and
operating costs may be high because of the small batch type
operations needed,

On the other hand, an aqueous reprocessing plant for high
burnup fuel cannot economically serve much less than 20 GWe
of installed nuclear capacity, although relatively small
additions at the head and tail end could be made to such plant
used for other fast systems.

To summarise, in exploring both theoretically and experiment-
ally the chemistry of molten chlorides as a fast reactor fuel
no practical impediment to their successful application has
been found, potential methods of fission product reprocessing
and salt clean-up have been identified and some specific
reactions of possible future value to the fuel cycle have been
d eveloped.

i, Reactor Physics

 

An important aim of the reactor physics investigations has
been Lo see what form the proposed reactor of 6000 MW(t) to be
built around Lhe year 2000 would have to take in order to
achieve a doubling time of 15 - 20 years, this figure being
judged against forecast doubling times of the generating
system as a whole.. However, there is not only considerable
uncertainty in this type of forecasting but variation in
r equirements in different countries.

Furthermore, for example, fuel inventory around the turn of
the century may not be such a critical criterion. A range of
alternatives was therefore explored.

Throughout the study only salts containing a mixture of
(U+Pu) Cl. and NaC?2 were considered. Early in the study it
became cléar that for the better ranges of nuclear performance
consideration would have to be given to the use of chlorine
enriched in the C1-37 isotope and it was found necessary to under-
take an evaluation of the cross-section data. For simplicity the
initial calculations assumed full enrichment to Cl-37.

To understand the basic physics performance of the MSFR a
series of calculations with simple spherical reactor systems was
made. The results of this study indicated that:-

(a) 1increase in heavy metal content within the salt improves
the nuclear performance although the salt becomes more
costly and its melting point increases. The composition
recommended as a result of this study was MO%O(U+Pu)
C1./60% NaCl (this has a melting point of 577 C), the Pu
crdction being about 10% of the heavy metal content.
10T PO PUBLICATION (COMMERCTIAL)

 

(b) acceptnble breeding gains are not obtained over the
rongme of core sizes consistent with satisfactory
fuel ratings unless the core 1s surrounded by a
blanket (initially 40% UCLl./60% NaCl) - that is a
sinpgle-region system did-ngt have adequate performance.

(¢) a blanket thickness of 1lm appeared to give a reasonable
nuclear performance without leading to excessive blanket
salt inventories.

(d) the breeding gain is increased by a factor of two when
natural chlorine is replaced by Cl=37,.

(e) a doubling time* of less than 20 years for the direct
cooled system would be obtained if cor% volumetric
ratings could:ge greater than 270 MW/m- with natural
Cl or ~100 MW/~ with Cl=37 in the salt. For the indirect
system, where a large part of the overall fuel inventory
is in the external heat removal circuit (see Section 6.2
for a discussion of factors controlling this) doubling
times in the region of 20 years can only be achileved
1f the core volume is reduced until the rating is
n360 MW/ m”, as in the designs shown.

(f) the temperature coefficient of reactivity and Doppler
coefficients are large and negative over the range of
operating temperatures considered so giving a means of
reactor control.

Tndirect Cooled MSFR

 

A lead cooled reactor with a layout as shown in Figure la
was analysed in more detail and with high C1-37 salt the doubling
time is ~25 years and with natural Cl~rUlb years. A simple
analysis showed that the system should be stable to small
reactivity perturbations if the ratio of the core volume to
total volume of core salt was greater than ! and the engineering
studies showed that this requirement eould be met with a margin.
Although it was not possible to carry out more than an elementary
analysis of control arrangements, the limited study suggested
that control of power level could be achieved by varying the
flow rate of salt through the core and heat exchangers, the
temperature of the salt remaining essentially constant at all
power levels due to the combination of the negative temperature
coefficient and the absence of the heat transfer and fuel pin
temperature differentials of a solid-fuel reactor.

The results of the studies of the transient behaviour of
this reactor with a simplified reactor model showed that the
temperature rise of the core salt due to a reactivity step
of up to $1 should be less than 300°C, and for 7 pumps failing
out of 8 will be less than QBOOC. Also the effect of changes in
the fraction of delayed neutrons re-entering the core from
the external circuit with wvariation of primary circuit flow rate
was assessed to be small.

¥ All doubling times quoted make a notional allowance for holdup
in processing plants.
NOT FOR PUBLICATION (COMMERCTAL)

 

Direct Cooled MSFR

 

The initial physics studies of this concept (see Figure lda
for the lay%ut) were based on volumetric rating within the core
of 100 MW/m” to conform with the limits on heat removal which
it was then thought might apply. The performance with natural
Cl in the salt was disappointing with a doubling time of ~60
years. To improve the doubling time chlorine was replaced by
high enrichment C1-37 and the doubling time obtalned was between
20 and 24 years depending on the various core-and blanket
configurations studied to accommodate flow requirements., To
give competitive fuel costs in terms of plutonium and C1l-37
inventory, i1t would be necessary to reduce the size of the
reactorzfuch that the volumetric rating substantially exceeded
100 WM/m”., The size eventually chosen for tge referenp%
engineering design has a core volume of 40 m” (150 MW/m
volumetric rating), giving a doubling time, taking account of
full fission product absorption, of about 20 years, which could
be reduced to 18 years if U40% of the fission products are
removed with an on-ine process and there is a low reprocessing
holdup (i.e., for an in-line plant or quick turn round in a
nearby centralised plant). It should however be noted (see
Appendix I) that there are serious doubts about heat removal
and separation at this rating.

In the original design the loss of the curtain of lead drops
(e.g. due to pump failures) increased the reactivity due to
replacement of the lead by the salt., But if the proportionate
volumes of lead and salt within the core are made to remain
constant for all confilgurations of the lead curtain by ensuring
that on loss of flow, the lead remains in a pool at the bottom

of the vessel the reactivity was decreased (8k = - 0.003),
This efg%:%k gogether with a negative temperature coefficient of
-6 x 10 —/ C gives a possible means of reactivity control by

the curtain, An analysis of a series of steady state runs at
power levels over the range of interest showed that it was
possible, by altering the inlet and outlet temperature, to keep
the lead flow constant, thus avoiding the effect of reactivity
changes from changes in the geometry of the curtain, and main-
taining adequate circulation and separation.

With this method the salt stays at essentially a constant
temperature and is slightly perturbed by control of the external
lead temperature to control power. Study of short-term and
rapid transient behaviocur could not be done with the effort
available but it seems likely that supplementary control would
be needed for this.

Comparison of Fuel Cycle Performance with other FTast Breeder
Reaector Schemnes

 

In the Table below, the sodium-cooled fast breeder cases
given are considered typical of an oxide-fuelled design and of
possible advanced designs using carbide fuel in the core as
well as blanket. The two gas-cooled fast reactor cases are
taken from recent publications by the Gas Breeder Reactor
Association.
NOT FOR PUBLICATION (COMMERCIATL)

The MAFR doubling times are quoted for 65% Load Factor and
A reprocessing hold-up time of 9 months, with 2% Pu losses to be
comparable Lo the other cases in the preliminary work, doubling
times do not include the plutonium formation for equilibrium
blanket conditions. Although the figures for different systems
are no%t on a strictly comparable basis and the reactor sizes are

different, it is considered the figures can be used to illustrate

the issues adeguately.

TABLE 4.1

Comparison of Performance of Reference
MSFR with other Fast Reactors

 

 

 

Specific
Thermal| Power | System . ) )
Reactor Power Densigy Inventory|Breeding| Doudling
MWt |MWt/m” |Kg Pupsg Gain Time
MWt
1330 MWe TMFRR (Oxide/Carbides| 3100 L84 1.23 0.18 3
Blanket)
1430 MWe LM?3R {(Carbide) 3100 484 1.06 0.33 16
1000 MWe GOFE (¥Vin) 2720 247 1.54 0.45 13
1000 MWe GCFR (Particle) 2840 421 1.31 0.36 17
2500 MWe Tndirectly Cooled 6000 364 1.82% 0.17 4y
MSFR (ClL design 1)
2500 MWe Tndirectly Cooled 6000 364 1.75% 0.29 25
MSFR (0157 Design 1)

 

 

 

 

 

 

¥ Conditions are guoted for the reference design cases with maximun

salt temperatures of 810°C, 20-25% reduction in inventory is

possible if maximum salt temperatures can be increased to 10007C.

/ Use of later data sets with a six batch fuelling cycle gives a
breeding gain of 0.24 and a doubling time of 26 years for the
oxide fuelled IMFBR; this later data has not yet been applied
to the MSHFR cases however.

Tt will he seen that the specific inventory of the molten
salt system is higher than for the other two types, but if Cl-37
is invoked, the improvement in breeding gain means the MSFR can
have a doubling time of about 25 years. Although this is higher
than for the advanced LMFBR and the GCFR with current trends
towards less frequent refuelling the breeding gains of these

reactors would be ilowered and the doubling times would increase.

10

 
v

NOT FOR PUBLICATION (COMMEKCIAL)

. Safety and Operational Aspects

has only been possible to make a very limited aporaiond

of these aspects and the notes below form a preliminary assess-
ment of the problems.

5.1

Normal Operation

 

5.1.1 Startup

The circults and the salt supply will first have
to be suitably preheated. The salt, cither without
plutonium or containing a "safe" concentration of
plutonium, can be introduced into the system from the
dump/storage tanks and the former fully flow tested,
etc. Plutonium can then gradually be added through
the on-line clean up loop, for example, until
criticality at zero power and a nominal Temperature
is reached. Further gradual additions of plutonium
will cause the power to rise and by manipulation
of the secondary circuit conditions, full temnera-
ture at low power reached. Drawing heat from the
secondary circuit will then cause the power to rise
to the desired level. The main operational control
will be by the secondary circuit.

5.1.2 Power Control

The strong negative temperature coefficient can
be used for control by temperature adjustment via the
secondary circuit. For very short term (i.e. spinning
reserve ) demands it may be necessary to incorporate
some additional heat capacity in the secondary system
as well as adjusting a by-pass round the steam
generators. In the event of a pump fault, tn: system
woiild reduce power and it may be possible to adjiust
hack to a higher power if some plutonium addition is
permissible; this would clearly not be done as a
short-term measure.

5.1.% Hotspots

The equivalent in MSFR of the classical
"notspot" of the solid fuel reactor is a reglon of
salt with low circulation rate. A preliminary assess-
ment for the direct cooled case suggests that within
the body of the salt, the considerable turbulence
woiild disperse hot blobs before they reach the wall,
and this argument can probably be applied to the
indirect cooled case. The important thing is to ensure
good control of flow at the core boundary.

5.1.4 Shutdown

It would normally be desirable to use the decay
heat to keep the system at temperature, and indeed if
this were insufficient., to keep it "simmering" at low
power.

11
NOT OB PUBLICATION (COMMERCIAL)

I there was need to reach a sub-critical condition
quickly, it might prove necessary to incorporate a
crude shutdown system but this should not generally
be invoked as the temperature coefficient should

hold the system steady at low power. If necessary,
the salt could be sent to the "ever-safe" dump tanks
with emergency cooling to cope with a rapid drop from
full power.

5.1.5 Small Leakages

 

(a) Leaks between core and blanket would not
have serious effects on the chemistry or
materials as the salts are compatible.
Further investigation would be needed as to
whether it would be better fo arrange
pressures so that core salt leaked to the
blanket to be detected by rise in Pu
concentration, or to allow the (safer)
leakage of blanket salt into the core and
check for change of core mean temperature.

{(b) The coolant circuits will be maintained at
a pressure slightly above the fuel or
blanket salt thus leakage of the less active
helium or lead coolant into the molten salt
will occur. In neither case should tThere
be any serious interactions. Any helium
in-leakage can be dealt with by the salt
cleanup system which also separates the
fission product gases. Traps may be required
at certain points in the circuit to deal with
lead leakage so that its position is determined,
and the effect of remaining small quantities
on parts of the core or blanket not made of
molybdenum must be considered carefully,
especially 1if nickel bearing alloys are used.

(¢) TLeakage of volatile fission products can
only be met by the containment, which would
have to be similar to that of M3RE, with
appropriate detection and leaktight barriers.
It seems reasonable to argue that the pressure
differentials would prevent any significant
leakage of fission products through the
secondary system to the steam circuit.

.2 Fault Conditions

H.2.1 Pump failure will lead to a reduction in
salt flow and the consequent temperature increacse
will reduce power: what subsequent adjustment
would be needed has yet to be examined. The
main gquestions could arise from maldistribution
of flow.

12
NOT FOR PUBLICATION (COMMERCIAL)

H.2.2 Simultaneocus fallure of all pump drives may
be perhaps best avoided by steam turbine drives.
Electrically driven pumps may have advantages for
inclusion in the containment and also are easier
to supply for pre-testing, etc. Inertia would
help pump run-down rates, or some short-term
auxiliary supply might have to be provided until
full dump of fuel has taken place.

5.2.3 Structural damage,if serious, is a further
case where dumping of fuel 1s the ultimate pro-
tection., Fallure of the core-blanket membhrane
leads to no chemical or compatibility problems,
and blanket salt entering the core circuit will
reduce reactivity. The biggest problem arises
from core salt entering the blanket, and although
an initial analysis suggests that temperature
rise could compensate the reactivity increase
involved, this needs further careful study.
Moderate rates of mixing may perhaps be accommodated
by salt temperature rises, and the margins above
operating temperature which could be allowed on
a short term basis are substantial; what the limit
is to the permissible rate of mixing needs to be
established. Failure of the main (blanket) vessel
calls for rapid dumping both from the circuit
and from the catchpots provided. It should be
noted that the dump tanks have been sized to
contain the contents of one secondary coolant
circuit as well as all core/blanket salt. To
protect the primary circuit against any possible
chance of pressurisation from a major steam
generator failure into the secondary system, the
latter will have to be suitably protected by
relief arrangements.

5.2.4 Major circuit fallure and the subsequent fate
of fission products is an important issue in a
fluid fuel system. The molten salt system does
not nave high vapour pressures, and although
some of Ehe primary circuit pressures can reach
3.2 MN/m= (460 psi) due to high pumping losses,
the system with lead secondary coolant is essentially
a hydraulic one in the pressurised regions so
there is not a lot of stored energy. Missile
formation is not therefore very likely and it
should be possible to demonstrate a containment
which will not be breached from this cause.

In a helium cooled system the reactor vessel
could be made to withstand the full helium
pressure which would result from a severe rupture
between the coolant and primary cireuits. Thus
a simultaneous double failure would be involved
to cause a release of activity to the reactor
containment cell.

13
NOT FOR_PUBLICATION (COMMERCIAL)

Tie: cell Itselfl 1s small and can cconomically be
made In the form of a prestressed veasel to wlth-
stand missile damage and to act as a third barrier
to i1ssion product release. A final low pressure
nuilding containment would prevent release to
atmosphere due to small leaks from the earlier
containment stages.

Furthermore there is a continual removal of
gaseous and other volatile fission products from
the salts during operation which will remove a
number of the more obnoxious species. Many of
the remainder will stay with the salt, which can
be withdrawn to the dump tanks. There should not
be anything corresponding to the molten fuel
accident of the sodium cooled and other solid
fuel systems. The major design problem will be
to ensure that (i) any salt escaping into the
containment cannot fall into regions where a
critical mass could form, (ii) any primary
circuit failure cannot lead to critical masses
forming within the circuit or other associated
ones such as blanket and coolant circuits. These
problems will need to be studied in the next
assessment phase.

5.2.5 Emergency cooling takes a different meaning
in an MSFR. The fuel can be dumped to prepared
tanks and the removal of heat from these dump
tanks can be by methods unconstrained by other
requirements as in solid fuel systems. The method
proposed follows the ORNL idea of natural circula-
tion of the salt through coolers in which the heat
is rejected to air drawn past by natural draught
towers. Multiple systems would give a high degree
of integrity and it seems possible that judicious
use of heat capacity would enable it to be claimed
that a forced draught system could be brought into
play to give a more compact and cheapter layout.

6. Engineering Design Philosophy

The major aspects Lo be considered in the overall design
and selection of parameters for a molten salt fast reactor
are discussed below; greater detail is given in the later
sections.

6.1 Choice of Materials

 

The most promising materials suitable for use with
molten salts and possible coolants are listed in Table 6.1
and discussed in more detail in Section 8. The temperature
limits quoted are based on strength considerations, as
corrosion tests have not necessarily been carried out at
these temperatures although lower fTemperature results show
promise.

14
NOT ToR PUBLICATION (COMMERCTIAL)

The effect of fission products must be considered and
investigated before a final choice is made.

 

 

1000-1200 (?)

c o1 _ Approximate
Fluid in contact :
with the material temperaggre limict Possible Materials
Molten chloride 600 Stainless steel
salt 700 Hastelloy N

Molybdenum or TZM

 

 

 

 

Lead 500 Fecralloy or Croloy
1000(%) Molybdenum
Superheated steam 550 Stainless steel or high
nickel alloys
Lower temperature 450 Iron alloys
steam and water
Helium 900-1000 Nickel alloys
1000 Molybdenum or TZM(?)

 

In view of its strength at high temperatures, good irradia-
tion resistance and high thermal conductivity, molybdenum
or its alloys are promising materials for all high temper-
ature components in contact with fuel and blanket salt
including the core/blanket membrane, intermediate heat
exchanger tubing, duct and vessel linings. Development
work on the fabrication of light molybdenum sections
including welding and heat treatment will be required and
is already proceeding for other applications. The high
specific cost of molybdenum is acceptable as long as the
heavier outer vessel and ducts, ete, are run at lower
temperatures in order to allow the use of more easily
fabricated and cheaper materials.

The use of molybdenum, or 1fs alloys such as TZM¥,
leads to the poss%bility of high maximum fuel salt tempera-
tures (above 1000°C) which will assist in reducing the
fuel salt inventory and allow the possibility of gas
turbine cycles and/or process heat applications using helium
as an intermediate and final coolant.

The choice of materials for the remainder of the inter-
mediate circuit of an indirectly cooled design will depend
on the choice of coolant and power cycle. As the amounts
of materials involved are greater than those in the primary
circuit the choice is more cost sensitive and except for
the high temperature helium case mentioned there is
incentive to run at lower temperatures allowing cheaper
materials to be used.
¥ Titanium, zirconium alloy of molybdenum.

15
HOT OR PUBLICATION (COMMERCG LAT)

iy thilo cane, hhowever, a halance hetween cost, corrogion,
thernal stress and avoldance of saltl freezling on the
internediated heat exchanger tubes must be preserved, the
latter of which can be restrictive if liquid metal inter-
mediate coonlant is employed.

With a direct lead cooled design materials must be
compatible with salt, lead and steam necessitating duplex
materials at least in the steam generators.

~.2 Fuel Inventory

Tt is important to reduce the fuel salt inventory in
order to reduce fuel costs, especially in the case of high
plutoninm values, and also to increase the rate of instal-
lation of plant for a given plutonium supply.

In the directly cooled scheme a considerable core
volume is predicted as necessary to allow time for
coarse separation of the fuel salt from the lead.
I'urthermore external separators, involving considerable
inventory in themselves, would be required to prevent
any salt, carried over with the lead, freezing on the steam
generator surfaces and also to reduce the inventory of such
salt in the externa% circuits. As the net lead flow to the
boilers is about 10~ cubic metres per day the quantities of
salt in the lead passing to the boilers must be no greater
than a few parts per million, or alternatively, very frequent
melting off would be required. The cycling times would be
impractically short for typical cyclone separator performance
of between 0.1 and 1% separation.

In view of these apparent limitations the directly
cooled design is no longer preferred. The features of
such a design as it evolved during the early study are
given in Appendix 1 for the record. Performance parameters
are given under the heading of Design 4, see Section 11.

In the indirect system employing external heat
exchangers (IHX) the two basic choices are {a) an intermediate
coolant which passes heat to steam generators or (b) a gaseous
intermediate coolant (helium) passing direct to gas turbines.

With an indirectly cooled version there is no signifi-
cant limitation of thermal rating that can be achieved in
the core, but a lower limit in core size 1s imposed if the
external/internal inventory is not to be too great for
nuclear stability and room 1s to be provided for connecting
duct work.

Temperature limitations necessitate a certain fuel salt
flow rate to the intermediate heat exchangers and although
the inventory connected with the heat exchangers themselves
can be reasonably low that in the ductwork predominates in
a practical design. Consliderable thought has been given to
various ways of reducing duct length but unconventional
methods tend to be less reliable or less easily maintained.

16
NOT FOR PUBLICATION (COMMERCIAL)

 

Duct cross sectlional areas can be reduced by employing
high velocities but vibration effects may be limiting.

A balance between the cost of pumping power and fuel
inventory must also be achieved. Choice of intermediate
coolant is clearly a further factor as it influences the
heat exchanger arrangement and temperature conditions.
This is discussed in the following section.

A more fundamental way of reducing external inventory
1s by increasing the temperature rise of the fuel salt
flowing through the core in order to reduce volumetric flow
rates. The core inlet temperature must be sufficiently
above the fuel salt freezing point to prevent excessive salt
freezing on the cool end of the heat exchangers; therfore as
high an outlet temperature as possible is desirable within
the limits imposed by materials and thermal stress considera-
tions.

6.3 Choice of Intermediate Coolant

Liquid coolants are considered only in conjunction with
a steam power plant as it seems doubtful whether anything is
to be gained by transferring the heat in a further set of
heat exchangers to drive gas turbines.

Of the liquid metals sodium suffers from the disadvantage
that fuel leakage into the coolant would cause an exothermic
reaction and removal by processing would be difficult, so that
even if it 1s assumed that the general problems of sodium
technology will be overcome by LMFBR development, there is
a basic objection.

Lead therefore appears to be more attractive chemically
as it does not react significantly with either salit or
steam and should not enlarge any leaks. Pumped by-pass
systems are required on the lead intermediate circuits to
give the required operating temperatures to avoid high thermal
stresses in the steam generators due to high temperature
differences, and to avoid salt freezing on the intermediate
heat exchanger tubing; although it may be arguable that a
thin layer of frozen salt could be allowed (see 7.2).

A molten salt such as NaF:NaBFu or Na Al 01M 1s best on

engineering and thermal hydraulic considerations as it gives
well matched heat transfer with low fuel inventory and reduces
freezing and thermal stress problems. Considerable experience
of pumping and general use of molten salts has been accumulated.
However leakage of steam into the coolant salt could produce
locally high corrosion, and processing difficulties would arise
if the coolant leaked into the fuel salt. It was therefore

concluding that molten salts probably were not a satisfactory
choice.

17
NOT FOR PUBLICATION (COMMERCIAL)

Although the cost of lead is high this is not an over-
whelming objection to its use and lead has been chosen as
the preferred liquid coolant. Development work ou opumps and
components as well as on its corrosive effects will be
required and high strength construction will be needed due
fo the inertia effects resulfing from the high density.
Special measures will be required to prevent lead freezing
at low power conditions and at start-up. Low alloy steels
can be used for the cooler parts of the lead circuit but
other measures may be required where temperatures are in
the region of 500°C or over. Duplex tubing will be needed
in the superheater as nickel bearing alloys are attacked
by lead.

Of the gaseous coolants C0O2 and other oxidising gases
will probably react with the fuel salt, in case of leakage,
to form UO2 which could overload the oxide removal section
of the salt clean-up plant. Helium is therefore preferred
on grounds of compatibility with fuel salt, for its high
temperature applications, for reduction in thermal stress
problems, and as considerable knowledge and experience
will be available by the time molten salt fast reactors are
built. It is also likely to be well suited for use with
the molybdenum heat exchanger tubing although this is not
yet fully endorsed. It will require a higher fuel salt
inventory than for liquld coolants for the same temperature
conditions, and require consideration of the safety issues
arising from a pressurised coolant.

6.4 Choice of Power Plant

The designs of MSFR considered in this report have been
linked to a conventional subcritical steam cycle to give a
cost comparison of the lead cooled case with a sodium cooled
fast reactor employing similar steam conditions. A super-
critical cycle could be employed or, to make fuller use of
the inherent high temperatures, process heat or topping
cycle applications could be considered.

Either lead or helium intermediate coolants may be
used with an HTR plant.

The use of a helium gas turbine cycle is attractive as
it does make use of the high temperature potential of the
molten salt system as well as being suitable for dry cooling
heat rejection where the situation demands. It also forms
an engineering entity as the gas turbine plant is similar
to that of the reactor in terms of size and scantlings;
the working pressures are lower than for a steam plant and
the possibility of prefabrication with low installation
times could lead to savings on construction and interest
charges. Plant costs may be reduced by the elimination of
the further heat exchangers required for steam generation
although the extent of cost savings will depend on the
overall efficiency selected for a gas turbine cycle.

18
NOT FOR PUBLICATION (COMMERCIAIL)

It a rapid MSFR system development is required to meet
expanding electricity demands, plutonium inventory must be
kept low requiring n hiphefficiency cycle with fairly
contly recuperantors, but it rapid expansion ol Lthe syatom
is not o important, and especially i! plutonium is of low
value, the cost of the recuperators could be reduced or
eliminated.

Only one preliminary design of core and intermediate
heat exchangers for helium cooling is given in this report.
Alternative helium cooled designs matched to the power
plant requirements and using concrete pressure vessels
will be given in subsequent reports.

6.5 Blanket Inventory and Cooling

 

There is not so much need to reduce the blanket
inventory as for the core fuel as the U-238 value is low,
however if Cl-3%7 is used to enhance breeding its cost is
such that excessive blanket volume should be avoided.
Pumping and cooling will be similar to that of the core
with arrangements so that the blanket power output varying
from 2 to 15% over the fuel cycle can be accommodated as
a contribution to the tofal intermediate coolant flow.

6.6 Reliability and Maintenance

 

There are no intricate mechanical features in a
molten salt reactor, thus the basic reliability of the
system should be high and be similar to that of chemical
or hydraulic plant. To achieve this state parameters and
materials must be chosen to be within the limits endorsed
by long term testing. This particularly applies to the
effects of corrcsion, thermal expansion, and high salt
velocity. If these features are thoroughly understood tThe
reliability of pumps, heat exchangers and containing
materials can be ensured. As there are no complex refuel-
ling mechanisms and no close tolerances as in solid fuelled
reactors and no mechanical conftrol devices in the core or
blanket a number of major factors affecting reliability
are eliminated.

It is important to show that similar principles of
simplicity, reliability and maintenance are applicable to
the auxiliary plant required for clean-up and reprocessing
and this is an essential part of any further study programme.

The ORNL Molten Salt Reactor Experiment has demonstrated
that maintenance on an active fluid fuel system can be
carried out without undue difficulty, although it is
important to design in features to assist in this respect.
Such features should include ease of pump replacement and
means of heat exchanger repair, by replacement or by in
situ repair or plugging of tubes. As an example, Design
2 illustrates a U-tube heat exchanger design giving freedom
of expansion of major items and with facility for plugging
any leaking tubes from the less active coolant circuit
side.

19
NOT FOR PUBLICATION (COMMERCIAL)

Thiese features are obtained at the cost of some extra fuel
inventory.

6.7 Engineering Development

The most important development work required is that
on the corrosion aspects of the fuel and blanket molten
salts with the complete spectrum of fission products and
likely impurities. It is important to determine the upper
1imit of temperature for long term operation under the
appropriate neutron flux levels. The principal areas of
component development appear to be:- P

(a) Work on pumps will be required to develop high
temperature pumps of considerable power for fuel
and blanket salt with geometry to suit the low
inventory requirements, and to deal with the
high density of the lead coolant if selected.

There is experience in low powermolten salt
pumps and more general experience of similar
problems of sealing, bearings and high heads in
liquid metal and HP water pumps but considerable
testing will be required under the proposed
operating conditions.

(b) On heat exchangers with small bore thin-walled
tubes closely pitched in a similar manner to
fuel clusters which will require development
and testing to resist the thermal and vibrational
strains over a long working 1ife and to be
suitable for manufacture from molybdenum.

(c) For the core and blanket hydraulic model work
is needed (i) to prove that there is good flow
distribution with no large stagnant zones and
(ii) to determine the effects of velocity and
density on stresses and vibrational effects to
establish designs with adequate limits.
Fabrication and expansion considerations will
also need study.

(d) Full scale and fully instrumented work with hot
molten salt will eventually be required as
proving tests in all these areas of development.

(e) The fate of the fission products will require
special attention as the order of magnitude of
non-gaseous and non-volatile products which do
not form soluble chlorides is about 0.8 tonne
per year; this could be serious if uneven
deposition occurs. The experience of MSRE may
give some indications of distribution although
the power was low in comparison with the MSFR
proposals.

20
NOT FOH PUBLTCATION (COMMISRCIAL)
outd i Sonbenm Desiens and Auxiliosry Plant

7.1 General Asnoeets

 

A summary of the features of Tour initial designs as
illustraced in Figures 1 to 4 is given in Table 7.1. Figures
la and ib {Design 1) show typical arrangements of reactor
and hecat cxchangers for indirect lead cooling for minimum
fuel inventory, the latter figuresshowling the larger heat
exchangers needed if the lower and probably more realistic
salt conductivity value is used. Figure 2 (Design 2) shows
an indirect lead cooled design layed out for freedom of
expansion of the major components, for cooler outer regions
of both core and blanket and with U tube heat exchangers in
which tubes could be plugged from the less active lead cooled
side. Figure % (Design 3) illustrates the dominance of the heat
exchangers on the layout of a helium Sooled indirect design
for a maximun salt temperature of 850°C. Figure 4 (Design 4)
is a tentative layout of a direct lead cooled design using
concepts which emerged as the result of a theoretical study
of direct cooling with only limited and very small scale
experimental backing.

Flow diagrams are given in Figures 9, 10 and 11 and
detailed parameters in Section 11. Details of the directly
cooled design are given in Appendix I and preliminary ideas
for an internally cooled MSFR are covered in Appendix II.

As compatibility problems in case of leakage have led
to the rejection of molten salt as an intermediate coolant
no details are given in the main text, although parameters
have been calculated.

Figures 5, 6 and 7 showing the overall plant layout
have been based on Design 1, but the other designs employing
a steam cycle wlll be similar.

A1l designs employ a multiplicity of cooling circuits
to reduce the effect of pump fallure. In case of circuit
rupture the whole core and blanket may be dumped into tanks
provided with natural circulation emergency cooling. A
valuable feature of the MSFR is that the fuel may be drained
and the whole reactor vessel or individual heat exchanger
units replaced. No item of plant is more than 6.5 m diameter
and shop fabrication should be possible. The compact layout
is indicated in the filgures and an overall size comparison
with a typical LMFBR and the lower powered MSBR 1s shown on
Figure 8.

21
NOT FOR PUBLICATION (COMMERCIAL)

TABLE 7.1 - 2500 MW(e) NET MSFR:

SUMMARY OF FEATURES

 

 

 

 

Type Indirectly Cooled Directly Cooled
Coolant Lead Lead Helium Lead Curtain
Design 1e 4 z/‘ 4°
Fuel salt temp in/out % 650/810 660/810 670/970 800 Max
Coolant temp in/out % 480/600 570/670 400/850 370/630
Outer Blanket Temp  °C 650 650 650 650
Core Shape and Size “Spherical® 3.2m dia | "Spherical" 3.2m dia "Spherical" 3.2m dia Vertical cylindrical
3.2m dia x 5m high
Flow in core Downwards Upwards in centre, cool| Upwards in centre, cool Poloidal with lead
return down in outer return down in outer curtain
annular region annual region
Heat exchangers 8. Vertical Tube in 8. U tubes in annulus 10, J tube in shell, -
8hell, salt flow surrounding pump aszlt Helium on shell side,
thro' tubes. on shell side, cool Counter flow.
Counterflow. Tube return in outermoat Removable complete.
bundle removable. annulus - counterflow,
pluggable,
Fuel Salt Pumps 8 centrifugal 8 semi-axial two stage | 10 semi-axial two Lead jet pump vanes inj
stage corporated into top of]
core inducing poloidal
salt flow
Pump Operating ° 650 810 670 -
Temperature C
Pumping Power (all
pumps) o
(a) Core salt MW 580 94 55 -
(b) Blanket salt MW 23 21 38 23
(c) Coolant core/ o
blanket MW 78/12 84/16 134/29 (IHX & pipework
only) 120-240/12
Total W 161° 215 256 160-280
Fuiel salt volumes o 172/21,5° 16/28 16/22 37/15-18
core/ext e ; ool
Total fuel inventory
Pu239 equivalent Te 9.5 10.8 9.b 9.2
Breeding gain (C137) 0.29 0.3 0.3 0.35
Linear Doubling Time
Years 25 28 24 21

Other features

Noten:

 

Veszel runs at upper
blanket temperature
in places. Bellows
required,

 

Pumps run hot.

Outer vessels run at
cool salt return
temperature.

Freedom for expansion
allowed, generally,

 

Bellows required,

 

 

* Assuming highest (ionised) value of thermal conductivity of fuel salt and 70:30 salt NP, 557°C, see section 7.3

# Assuming wmean of ionised and un-ionised theoretical values of thermal conductivity and 60:40 salt MP, 577°C.

® Normalised to 6000 MW{th) maximum core rating and 900 MW(th) maximum blanket rating from detailed parameters
quoted for 5700 MW{th) and 300 MW(th) respectively.

e2
NOT FOR PUBLICATION (COMMERCIAL)

Tt is ceen that the overall building and civil work
are reduced for MSFR compared with a LMFBR largely due to
the absence of fuel handling facilities. If reprocessing
by an in-line pyrochemical process proves practicable and
economic the space required to house the plant and attendant
control and maintenance areas can be incorporated in the
building without increasing its size significantly. A
saving in space could be made if high pressure storage
of fission gases is feasible.

With the available effort it was necessary to
concentrate the design work on the reactor vessel and
primary heat removal systems, nominal values only being
obtained for the remaining systems associated with the
reactor, e.y. steam generators, dump system with emergency
cooling, ofl-gas system with delay beds, fission product
removal nnd c¢lean up. It has become apparent that lhece
tatter items have a significant influence on the plant and
greater attention would in future have to be given to the
principles involved and the detailed deslgn in order to
establish gquantitative costs.

7.2 The Indirectly Cooled Concept

The reference lead cooled designs have been based on
a total reactor output of 6000 MW(thermal) over 98% of
which is initially produced in the core, but allowing for
what was considered to be an upper limit of 15% power to
pe produced in the blanket at the end of the fuel cycle.

The core is roughly spherical of 3.2m diameter and 16 to
17 mo volume. Small varlations of core size would have

little effect on costs or external fuel inventory. An
outer blanket zone of 1m nominal thickness is also cooled
by external heat exchangers of similar type to those
cooling the core salt. Cool return flow is arranged in
the blanket to keep the reactor vessel below 650°C and
allow nickel alloys or stainless steel to be used for the
outer vessel. This principle of keeping down temperatures
in outer vessels or pipework is followed as much as possible
in the layout. The core is separated from the blanket by
a single, or double, thin molybdenum membrane of lcm total
thickness which is suitably stiffened to withstand the
static heads and differential pumping pressures. Pump
inlet pressure is only slightly above atmospheric, thus
the outer vessel is only required to contain pressure due
to pumping at between 2.6 and 3.6 MN/m? (380 and 520 psi)
depending upon the design. Pumping pressures between core
and blanket must be approximately balanced.

The vessel and heat exchangers would be mounted in a
hot box fog pre-heating above the fuel/blanket salt melting
point (580°C). Lower temperature hot boxes or locally
applied heating will surround the secondary piping and
steam gengrators in the lead cooled case (lead melting
point 327 °C).

23
NOT FOR PUBLICATION(COMMERCIAL)

The general reactor arrangement is governed by (a)
the need for a compact system to primarily reduce the fuel
and secondarily the blanket inventory costs; (b) the max-
imum temperature allowable in the core region from corrosion
and thermal stress considerations in the core(or blanket)
and intermediate heat exchangers or steam generators, and
(¢) the minimum salt temperature on the heat exchanger
surfaces if freezing is to be avoided.

Maximum fuel temperatures have been limited in the
reference designs to the lowest values judged to give an
acceptable compromige for the above conditions for early o
MSFRs. This is 810°C for the lead cooled designs and 970°C
for helium cooled systems. Corrosion tests at higher
temperatures, i.e. above 1000 °C are required to indicate
the potential of higher limits which would allow lower fuel
inventory and_ pumping costs. Steady state. al. stresses
with the temperature differentials applying in the heat =
exchanger tubing and boundaries between the hot salt outlet
and cool return or at the core blanket membrane are accept-
able, but thermal sleeves may be required at discontinuities.
Consideration has yet to be given to the effects of sudden
change of load on thermal stressing, although preliminary
control studies indicate that the change of the fuel salt
mean temperature is modest. High salt flow rates are
required to give values of the salt return temperature
which, in allowing for the boudnary layer temperature drop
gives wall temperatures high enough to prevent freezing or
plating out of fission products on the IHX surfaces.
Similarly, the lowest lead temperature must be kept high
due to the good heat transfer on the lead side. This in
turn necessitates high pumping losses in the intermediate
circuit. If freezing can be allowed without causing
excessive pressure drop (due to a "slushy" mixture?) or
without causing excessive plating out of fission products
then flow rates could be considerably reduced with consequent
savings in inventory and pumping power. The freezing
problem is not so pronounced with helium cooling as the IHX
tubing surface runs at a high temperature due to the lower
heat transfer on the helium side.

To reduce the pumping powers and the temperature
differences between the lead and the water in the evaporators,
which could cause high thermal stresses, it is necessary to
pass lower lead flows through the steam generators and mix
with the flow from the IHX to give more suitable temperatures
as shown on the Flow diagram, Figure 9. The parameter list
shows the quantities.

7.3 Design Variations of the Indirectly Cooled Concept

A summary of the main distinguishing features of the
various designs is given in Table 7.1l.

2l
NOT FOR_PUBLICATION (COMMERCIAL)

As meanurement of the salt thermal conductivity io
not. yet complete, Desligns 2 and 3 are hased on wha% i
consldered o he a realistic value of 0.5 wom b KT which
i the mean of the lonlsed and un-ionised values. Design
1L used the original, more optimistic value of 2.0 w.m K‘i;

In all designs the pumps are arranged with vertical
shafts as successfully tested at ORNL.

Further work is confinuing on the helium cooled
variations which will show an improved layout and tackle
the problem of a high pressure intermediate coolant circuit.

It should be noted that in helium cooled designs the
high pressure will cause any small leaks to be from the
helium into the salt, thus preventing the spread of activity.
The 1limiting leakage will depend on the capacity of the
salt off-gas clean-up plant.

In order to give a true comparison of the lead and
nelium cooled designs, it will be necessary to carry out
a reasonably accurate comparison of overall costs, including
the cost of introducing a new lead coolant technology and
the costs of safeguards against helium circuit rupture.

7.4 Special Auxiliary Plant

It should be noted that only a small amount of
engineering design work has been done on the auxiliary
plant, most ?f which is similar to that proposed by ORNL
for the MSER 2) except where the chemical composition of
the chloride salt in MSFR causes differences. Nominal
parameters are given in Section 8, the Flow diagrams,
Figures 9 and 10, and building layout, Figures 6 and 7, show
the disposition of the main items.

7.5 Filling, Draining and Dump Systems

The arrangements are shown in outline on the flow
diagram, Figure 9. Two tanks are mounted below the reactor
vessel serving as filling and drain tanks respectively.

They are each sized to contain the whole salt inventory
associated either with the core or the blanket plus the

lead inventory contained in one of the respective coolant
circuits which, being individual loops, can be regarded

as independent. The two tanks also serve as dump Tanks

for quick dumping of the core and/or blanket as described

in Section 7.6, or for receiving salt which leaks from the
reactor vessel via the catchpot below the reactor. A

third tank of equal volume to one of the two tanks described
is provided as a standby and to serve as a delay tank for
salt removal for processing. If separate processing of core
and blanket is required this tank will be divided into two.
The emergency cooling system would also be used to remove
the decay heat on normal draining.

25
NOT FOR PUBLICATION (COMMERCIAL)

Smaller delay tanks are provided for batch entry and
discharge Lo glve a mirimum delay period for fuel or
hlanket salt passing to the processing plant. Murther
tanks are provided for enrichment trimming. Mechanical
pumps or gas pressure can be used for salt transfer. For
initial filling the fuel or blanket salt will be brought
to site in solid form and heated under inert gas in small
non-critical batches to run into the respective non-critical
shaped tank from which i1t can be pumped into the reactor.

7.6 Emergency Cooling System

The system provides a reliable and completely separate
means of removing heat after reactor shutdown by draining
the fuel and/or blanket salt into the dump tanks. Cooling
of the dump tanks is reliably achieved by natural circula-
tion of a low meling point salt or Na-K through U-tubes,
occupying most of the drain tank volume, to boiling water
heat exchangers. The steam formed is condensed in air
cooled condensers situated in a normal or forced draught
stack. Alternatively, the heat may be transferred directly
to a large pool which is allowed to boil, thus accommodating
decay heat for a protracted period without make-up.
Condensers or make-up water would be provided for continuous
operation.

The heat removal capability of the emergency cooling
system described in this report was based on decay heat
removal of 2%% from each tank on the assumption that all
fuel salt could drain into one tank only. When the effect
of remaining neutron power after complete loss of main
cooling is added considerably higher heat production (up
to 6% of full power) could occur in a rapid dump. Temper-
ature transients will be studied in relation to dump times
to determine a more realistic assessment of the heat removal
required allowing for the heat capacity of the emergency
coolant.

7.7 Salt Cleanup and Gaseous Fission Product Removal

These terms refer to the on-line treatment of fuel
and blanket salt to remove oxides, to reduce UCly to UC13
and to remove gaseous fission products. A tentative flow-
sheet is shown in Figure 10.

In order to avoid build-up of gases which would
affect circulation and cause pressure to build up and also
to reduce the amount of gaseous fission products that may
be released in the event of circuit rupture, the off-gases
are continuously removed. A similar system to that proposed
for MSBR by ORNL could be used. In this the gases are
separated in a recirculating loop across the main fuel (and
blanket) salt pumps by injecting helium bubbles to cause
nucleation. The fission products removed consist mainly
of xenon, krypton, tritium and small particles (fog) of
noble metals.

26
NOT FOR PUBLICATION (COMMERCTIAL)

The gases, together with those purged from the pump
bowls and {rom the highest points of the salt circuits,
are passed through a cooler with a small quantity of salt
from the main pumps, to the drain tank. Due to the size
of the tank there is considerable holdup and reduction in
decay heat (which passes to the tank coolers). Some noble
metals will probably deposit. The gases are then passed
through particle traps to charcoal beds for absorption.
In the MSBR system 80% of the gas flow is passed to a
chemical trap, and after compression into a surge tank it
drives the bubble generator and returns to the primary
circuit. The remaining 20% of off-gas flow may be passed
to a long delay system in which all isotopes except the
10-year Kr-85 will decay to an insignificant level. The
gases are purified by removing Kr-85, stable isotopes of
XKr Xe, water,etc, before returning to the primary circuit
via a surge tank and compressors.

It is likely that the noble metals, except those
escaping as fog, will deposit in various parts of the
primary or FP removal circuits as already referred fo in

6.7.

The remaining fission products, ie, 50-60% of those
formed, will be dealt with in the external processing plant
unless on-line treatment processes can be developed.

It would appear that the main design work required
tfor gaseous fission product removal is in sizing the plant
to cope with the absorption required in the delay beds and
the heat removal requirements to avoid excessive temperature
rise in the charcoal. The MSFR building layout shown
provides for an outer annular ring for delay bheds based on
this method and sized pro rata with power to the MSBR
proposals. The form of the beds can either be as in
DRAGON, in which the cooling water is contained in cylindrical
tanks or, as in the MSBR case, in a swimming pool type
trough with low pressure steam formed above the pool and
passed o condenser units.

In view of the large size of the delay bed system
which dwarfs the reactor, alternative methods of storing
fission product gases should be considered. There appear
to be design principles which could permit the gases to be
stored safely and reliably at high pressure.

In order to clean up the salt chemically, flow from
the main pumps is passed through:-

(a) an UCly reduction bed formed by inserting uranium
metal into the stream in sufficient quantity to
satisfy the reaction - the form of uranium,
whether rod or gauze or turnings must be determined;

27
NOT FOR PUBLICATION (COMMERCIAL)

(b) oxide removal bed. This may take the form of a
continuous process of inJection of liquid Na-Al-
Cly and flltration or cyclone separation of the
solid alumina formed. Alternatively a batch
process could be used and alumina precipitated
could be removed alternately from the on and off
line plants.

The tentative flow rates and reaction requirements
are given in the parameter list - these must be confirmed
before worthwhile plant design and costing can proceed.

7.8 Processing Plant

The term "processing" refers to separation of a large
proportion of the remaining fission products from the core
and blanket salt, separation of the plutonium from the
blanket salt, and enrichment of the core salt to replace
fuel burnt up.

For present design purposes it has been assumed that
processing will be by aqueous methods and will be carried
out remote from the reactor, possibly at some central
national site or a plant serving several reactors and no
plant is described here. If the economics of on-site
nyroehemical processing turn out to be more favourable than
is predicted at present, the plant can easlly be accommodated
within the reactor bullding due to its small size.

It may be possible to direct the blanket salt into
the core, and only process the salt taken from the core.

7.9 Remaining Plant

 

7.9.1 Containment

The aim will be to construct the primary and
intermediate circults to standards to give negligible
leakage rates (as achieved on similar size components
on the DRAGON reactor).

To deal with any leakage and to provide an 1nert
atmosphere round the hot parts of the plant, if this
is necessary, a secondary containment is formed by a
low pressure leak-tight membrane around the walls,
floors and ceiling of the plant cells. This membrane
can also form part of the ducting for the gas circula-
tion required to cool the concrete structure and
shielding. Heat losses with insulataon restricting
the conoretg temperature to below 70 C would be about
200 watts/m= with 40 em of insulation. This would
equate to about 3MW total heat removal by water or air
cooling.

A tertiary containment can be incorporated in the

main building, the volume of which must be sufficient
to contain the stored energy of any gases present.

28
NOT FOR PUBLICATION (COMMERCIAL)

These will be confined to cover gas volumes at low Lo
moderate pressure in the lead cooled design but for

the helium cooled version a larger outer containment
volume and/or intermediate prestressed concrete contain-
ment may be required.

7.9.2 Preheating

Preheating to about 600°C for the salt circuits
and 330°C for lead will be required. In some cases
it will be more economic and convenient to heat whole
spaces such as the reactor vault and processing plant,
but where longer lengths of plpework are involved as
in the lead pipes to the steam generators, trace
heating would be preferred. No particular problems
have been encountered with preheating on the ORNL MSRE.

7.9.5 Shielding

No calculations have been made. A nominal thick-
ness of 75 metres has been allowed on the drawings at
the critical positions.

7.9.4 Pumps

The single stage centrifugal salt pumps shown in
Design 1 can be similar in principle to those success-
fully tested in the MSRE and in loops up to 16,000
nours running at 650°C. For the two stage semi-axial
design basic development work may be¢ required; in both
cases development and testing will be required to
successfully increase sizes to the 5-12 MW range.

There is no direct experience of high power,
continuously running lead pumps. The design principles
would probably follow the sodium pump design developed
for sodium fast reactors but with higher power and
strength requirements.

7.9.5 Steam Plant

Standard _steam conditions of 16.3 MN/ 2 (2350 psig)
at 833K (1050°F) with reheat to 811K (1000°F) are
proposed using once-through boilers. For costing and
layout purposes a once-through U tube design or
evaporator, superheater and reheater has been used.
Jnits are duplicated in each of the main circuits to
reduce the tube plate size.

Two 1350 MW({e) gross output T/A sets are shown,
although a greater number of lower power sets could
be egqually conveniently situated. Station load would
be about 200 MW including approximately 160 MW pumping
power, giving a net efficiency of 41.7%.
NOT FOR PUBLICATION (COMMERCIAL)
8. Materials

The problems of selecting structural materials for the
MSFR family of reactors are difficult but not insuperable.
Even at this early stage it is possible to make recommendations
which appear feasible and the problem areas which have been
identified are probably not intractable. Fach design will be
considered separately.

a. Design 1. (Indirectly cooled with lead - first version)

The choice of materials for the core components, in
contact with fuel salt and blanket salt is limited to high
nickel alloys or refractory metals. 1In practical terms
the choice is probably between Hastelloy-N (or more
advanced developments of that alloy) and molybdenum alloys
(particularly TZM). The superior high temperature strength
of molybdenum alloys and their greater resistance to radia-
tion embrittlement at high temperature make them the more
attractive for the higher temperature components (eg the
core/blanket membrane). Hastelloy-N is a possibility for
the less severe conditions of the reactor vessel and the
salt ducts returning from the heat exchanger. If molybdenum
were used for these components it would presumably have to
take the form of duplex materials, eg a stainless steel
vessel internally clad with molybdenum; if massive moly-
bdenum were used steps would have to be taken To prevent
its external oxidation, for example by containing the
reactor vessel and other sensitive components in a hot
box purged with an inert or slightly reducing atmosphere.
There is some incentive for using molybdenum alloys for all
the components, including the pumps, which come in contact
with the salt. Firstly, having decided fo use molybdenum
for the core/blanket membrane, a change to another material
elsewhere in the salt circuit would introduce the possibility
(difficult to assess at present) of dissimilar metal effects.,
Secondly, nickel alloys resist molten lead very poorly
compared with molybdenum and hence would probably be
subject to attack when the lead leaked into the salt.

The heat exchanger tubes are subjected to both the
fuel salt and molten lead and hence would preferably be
fabricated of molybdenum alloys. Such alloys are normally
highly resistant to lead but there is conflicting evidence
about the effects of small amounts of oxygen in the lead,
as would result from a steam generator leakage. For this
reason it may be advisable to minimise the oxygen level
in the lead by means of a getter (magnesium or calcium)
dissolved in the lead. Alternatively the molybdenum may
have to be clad externally with an alloy resistant to
"oxygenated" lead. A ferritic steel containing aluminium
(Fecraloy) which should form a protective alumina film
under these conditions is a candidate cladding material.
The possibility of making the heat exchanger tubes of
Hastelloy-N has only slight attractions and in any case
this material would have to be clad externally either with
Fecraloy or molybdenum to make it resistant to lead.

50
NOT FOR PUBLICATION (COMMERCIAL)

 

The ducts carrying the lead to and from the steam generators
and zlso the steam generator shells could probably be
molybdenum (protected from external oxidation either by a
stainless steel cladding or by means of an inert atmosphere)
or less probably of Fecraloy (externally clad with stain-
less steel for strength considerations).

The steam generator tubes pose another problem associated
with the probable steam leaks into the lead, because it is
likely that the "lead side" of the tubes will locally be
in contact with steam in quantities much too high to be
removed by an possible getter. Hence it may be difficult
at present to recommend molybdenum for use in the steam
generator tubes. An alternative, as yet untried, would be
to use TFecraloy in the hope that it is resistant both to
steam and fto oxygenated lead. Its poor high temperature
strength could be overcome by using it in the form of a
Fecraloy-stainless steel duplex and the selection of stain-
less steels to resist the water/steam side of the steam
generator should be conventional.

b. Design 2. (Indirectly cooled with lead - second version).

Considerations of materials selection are very similar
to those of Design 1. The use of cool salt from the heat
exchanger to reduce the temperature of the core/blanket
membrane and of the reactor vessel may permit the adoption
of Hastelloy-N for these components. However other considera-
tions may still demand the use of molybdenum, in which case
the lower temperature may actually be a disadvantage in
that radiation embrittlement of molybdenum may be more
severe.,

Another major difference between Designs 1 and 2 1is
the higher salt pump temperature in the latter and this may
increase material selection and engineering problems.

c. Design 3 (Helium cooled).

Many of the features of the reactor itself are similar
to those of Design 2 except that the temperatures are
signifigantly higher (the core temperature is 970°C instead
of $10°¢C). This would certainly necessitate the use of
molybdenum alloys, especially on mechanical grounds, and
although there is no special reason to doubt their corrosion
resistance at this temperature, any supporting evidence is
completely lacking. Design 3 shows one significant relaxa-
tion of conditions compared with Design 2, viz thg reduction
ofrthg salt pump operating temperature from 8007C to

670°C, and this will doubtless ease the engineering and
the selection of materials.

The adoption of helium cooling would probably go even
further to ease the materials selection problems but it
should be made clear that much less effort has been devoted
to materials selection for this design than for the lead
cooled designs. Hence there could easily be problem areas

pa
NOT FOR _PUBLICATION (COMMERCIAL)

as yet unidentified but equally challenging as found with
the lead cooled systems.

The assumption is made that a helium cooled system
could make extensive use of the work already done with
respect to HTRs, and specifically in relation to gas-turbine
versions, and that the MSFR version could ride on the
future HTR programme. In connection with HTRs it is well
recognised that, in practice, helium is not an inert coolant
since traces of air, moisture and other oxidising gases
can cause corrosion and traces of carbonaceous gas can bring
about carburisation of candidate structural materials. A
great deal of work still has to be carried out before a
detailed specification of materials could be made. However
it does seem likely that (a) it will be possible to specify
the impurity levels acceptable in the helium, methods of
controlling them and what intentional additions, if any,
should be made, for example, to render the helium reducing,
and (b) selected nickel alloys could be used for the turbines
and other circuit components. This would restrict the need
for molybdenum in the coolant circuit to heat exchanger
tubes and similar applications.

d. Design 4 (Directly cooled with lead)

The directly cooled system is rather less attractive
from the point of view of materials selection than the
indirectly cooled system for several reasons:

1. There are more situations where components have
to resist a number of aggresive corrodents
simultaneously (eg lead, fuel salt and air);
hence Hastelloy N is less attractive for core
components which may come in contact with
entrained lead.

2. The lead will contain fission products, trans-
mutation products, and, very likely, entrained
salt.

3. There is less scope for adjusting the detailed

composition of the lead without the constraints
imposed by interactions with the fuel salt.

Although it does not seem that any of these aspects
impose insuperable problems they do increase the difficulties
of specifying materials for the system. Nevertheless it
still seems probable that suitable materials can be selected
from the range already considered, eg, molybdenum, Hastelloy-N,
stainless steel and Fecraloy.

Costs
9.1 Capital Costs

The indirect lead cooled MSFR has been compared with
a pool version of a sodium-cooled fast breeder system of

32
NOT FOR PUBLICATION (COMMERCIAL)

circa 1990, as bheing the reactor type nearest in general
form to the MS8FKR. The difficulties of doing this even in
broad terms will be recognised but it is possible to
identify certain areas where the MSFR has potential for
cost savings:-

(a) the MSFR system avoids elaborate control systems
and complex fuel handling equipment. There
should therefore be gains on compactness and
simplicity grounds, but there will be extra
costs for certain ancillary systems which must
be allowed for.

(b) savings on contractors' and customers' costs
should accrue because the size of The components
indicates that shop fabrication is possible in
nearly all cases; further, simplified erection,
testing and commissioning procedures should be
possible with a plant of chemical type. By the
same arguments there should be reductions in
construction time with consequent savings in
interest and overhead charges.

The net result of this semi-quantitative comparison
is an estimated saving on capital cost relative to a 1990
LMFBR of £12-20/kW(e)*, 1972 values.

Although it may be argued that by the time MSFRs could
be in operation LMFBRs would have advanced beyond the 1990
version, 1t should be noted that the comparison referred
to above does not credit the MSFR with two areas of possible
further potential:-

(a) 1if full advantage could be taken of the higher
temperature potential of the system (see Section
7.2) a possibly optimistic but not unrealistic
estimate of further savings on plant costs leads
to forecasts of gains of about £25/kW, for the
lead secondary coolant/steam plant system.

(b) should the further studies of helium as secondary
coolant show promise of meeting the safety argu-
ments, the possible temperatures permit considera-
tion of gas-turbines, which with a layout integrated
in a pre-stressed concrete pressure container
vessel, is clearly a line worth pursuing for
further savings.

9.2 Fuel Costs

For comparison with other system fuel costs, the costing
is based onl.l.72 prices, 25 years operation, 75% load
factor. The values given below have been assessed for the
Indirect System (Design 1) using off-site aqueous solvent
extraction methods for heavy metal processing and removal
of residual fission products. An on-line plant is assumed
for removal of volatile fission products and salt clean-up.

* £/kW quoted as present worth.

33
NOT FOR PUBLICATION (COMMERCIAL)

The effect of this continuous partial removal of
fission products is discussed further below. The cost
ffigures allow for inventories both in the reactor and in
the reprocessing cycle, where a hold up time of six months
is takne, composed of three months cooling-off delay period
and three months processing and transport time. 1In
addition, an allowance of one month's stock of fresh fuel
at the reactor site is made. Estimates were made also of
the cost of separating Cl1-37 and its effect on reprocessing
procedures so as to be able to make a comparison with the
natural chlorine case which would allow inter alia for the
inventory aspects introduced by the higher-valued fuel salt
containg Cl-37.

As has been mentioned earlier, the on-line pyrochemical
process was examined in the hope that it would contribute
to reduced inventory and thereby effect improvements over
the aqueous processing route, but while it appeared
technically feasible, the initial estimates were discouraging
due to indications of high costs arising from batch processing
of small "non-critical" quantities and the remote operation
and maintenance involved. This is an area which requires
further study.

Tt is possible that the combination of an on-line
salt clean-up plant with plate-out of noble metals will
result in an effect equivalent to 50% of the fission
products being removed from the core region, and a survey
analysis was carried out assuming this value to assess
the desirable rates of throughput for the external reprocess-
ing plant. The rate has been expressed as a cycle time,
ie, the time to process a volume of salt equal to the
primary and blanket system inventories. The survey indicated
that system inventory and doubling time are relatively
insensitive to cycle time even up to values of 15 years.
However, for cycle times of over 10 years the heat generated
in the blanket exceeds 10% of the total heat output and it
is judged that excessive capital cost would then be involved
in blanket heat removal plant which would be running below
capacity for over one-third of the reactor life. IT was
therefore decided to use an equivalent cycle time of 7.5
years as a compromise. This figure is a judgment, and
further study is desirable to derive a better understanding
of the effects of the various parameters.

The investigation showed that while the use of Cl-37
gave a notable improvement in nuclear performance in
reducing doubling time (see Section 4), the additional
inventory charges resulting from the increased salt costs
made the fuel costs higher for the Cl-37 case than for
natural chlorine salt.

Taking plutonium as having zero value so as to bring
out the "fabrication and reprocessing" aspects showed that
compared with the oxide fuelled LMFBR of Table 4.1# the
potential savings for the MSFR, expressed as £/kW dis-
counted present worth, were £16/kW for the Cl-37 case and

34
NOT FOR PUBLICATION (COMMERCIAL)

£20/kW for the natural chlorine case. It was sald in the
Introduction that the one of the hoped-for gains would be

a considerable saving on fuel "fabrication and reprocessing"
charges and these figures demonstrate that this is possible.
But if plutonium is given a typical value of £5/gm,
inventory effects reduce the potential gains to £7/kW with
C1-37 and £8/kW with natural chlorine.

If however the ma§imum fuel salt temperatures can be
increaged from the 800°C of the reference lead-cooled case
to 920°C, the inventory requirements reduce by 25% which
with plutonium at £5/gm gives a further reduction of about
£5/kW for both C1-37 and natural chlorine MSFRs. The
inventory then approaches that for the oxide-fuelled LMFBR
and the natural chlorine version doubling time would fall
to a value similar to that of the LMFBR.

Q9,3 Wwverall Generating Costs

The indications from the above estimates are that the
total generating costs for the reference case lead-cooled
MSFR (assuming similar operating costs) would be in the
range £20-30/kW less than for a 1990 oxide fuelled LMFBR
with plutonium at £5/gm, on a discounted present worth
pasis and referred to 1.1.72 prices. With plutonium at
zero value this margin increases by about £8/kW. Alternatively
if the higher temperature MSFR version can be established,
the gain with plutonium at £5/gm would be increased by £10/kW
or more. It is of interest to note that on the present
analysis, the gains claimed accrue somewhat more from
capital cost savings than from fuel cost savings and this
puts emphasis on the importance of further investigation
of plant design, and maintenance problems, and the associated
safety issues, to ascertain better the validity of the claim.
Tt should be noted also that the fuel costs have been
derived on the basis of agueous processing for heavy metal
where the route is reasonably well established, and that
it has only been necessary to invoke an on-line high Temp-
erature process for the salt clean-up.

Although therefore it can be expected that there may
be improvements in LMFBRs compared with the 1990 verslion,
the MSFR appears to have potential justifying its [urther
study, bearing in mind also that no claim has been made
for benefits arising from avoiding the losses in operating
time arising from shutdowns for refuelling in the LMFBR.

10. Conclusions

The conclusions which can be derived from the studies
summarised in this report are:-

(i) of the alternative forms of MSFR examined, the
indirect-cooled versions (see Figures lb or 2) appear
to offer most potential for further investigation.
There are however uncertainties in a number of physical
properties (notably thermal conductivity) which reflect

35
(i1)

(iii)

(iv)

NOT FOR PUBLICATION (COMMERCIAL)

noticeably on choice of design conditions, and these,
coupled with some uncertainties on nuclear data,

make it difficult to achieve any substantial degree
of optimisation in design at this stage.. Choice of
secondary coolant is an important feature, and of the
liquids, lead appears to be the most suitable taking
into account compatibility with the fuel salt. Since,
however, helium technology will be developed for HTR's
and it should pose less materials problems, study of
the possible use of helium in an MSFR, which will
require careful consideration of the safely issues

of a pressurised coolant, should be undertaken.

the chemistry of the chloride salts proposed for fuel
and blanket shows, on the limited work to date,
encouraging signs of being controllable to the required
degree; but both in this area, where the effects of
potential precipitating species (eg, due to oxygen and
nitrogen from adventitious air ingress, sulphur from
Cl-3%5, fission products) and the possible deposition

of noble metal fission products need study. In the
materials and corrosion fields, much more work is
clearly needed. The extent of the materials work will
pe affected by the choice of secondary coclant, but

it is evident that considerable reliance will have to
be placed on the demonstration of good corrosion
performance and development in the technique of fabrica-
tion of molybdenum or its alloys for the key components
(eg, core vessel and heat exchangers) of the core and
blanket circuits. Suitable design should enable more
conventional materials to be used for the main vessels
and other components.

there appear to be good prospects of being able to
operate an MSFR without conventional nuclear control
systems. This together with the avoidance of
refuelling equipment, offers the possibility of a
compact system with more the character of a chemical
plant. It is from these possibilities that the
potential for capital cost savings compared with an
LMFBR can be discerned, but there are the imporftant
areas of containment, emergency dumping and ancillary
systems requiring further investigation to assess
any off-setting effects.

the reference lead-cooled indirect design of this
report is inferior to a 1990 oxide-fuelled ILMFBR in
fuel inventory but current evidence suggests that an
increase in primary salt top temperature could be
contemplated to a level where the inventories are
comparable. The doubling times with natural chlorine
in the salt would then also be comparable, and if
separated Cl-37 were used doubling times in the range
20-25 years seem possible.

36
NOT FOR PUBLICATION (COMMERCIAL)

(v) the preliminary costing (which is all it has been
possible to carry out), 1lndicates that the reference
MSFR has potential for savings relative to a 1990
LMFBKR both in the capital cost (about £12-20/kWe)
and fuel cost areas (£8/kW). The increased top
temperature scheme referred to in (iv) shows further
gains of £5/kW for capital and £5/kW for fuel costs,
the latter reflecting the reduced inventory charges
with plutonium taken at £5/gm. These costs are for
natural chlorine salt; Cl-37 although giving the
gains in doubling time already mentioned, leads to a
slightly higher fuel cost because of increased inven-
tory charges, so its use must be judged (apart from
any chemistry gains in reducing the amount of sulphur
formed from Cl-35) on its value in an overall generating
system where increased installation rates can have
discounted worth.

(vi) the fuel costings have taken aqueous solvent extraction
as the process of for heavy metal handling and show
favourable results. It is not therefore necessary to
invoke a close-coupled pyrochemical method, and indeed
the nreliminary investigations made, while indicating
technical feasibility, raised doubts about the costs
o this method with the small scale batching technique
considered; this 1s an area requiring further examina-
tion.

(vii) the investigations reported here have indicated the
value of the Intrinsic self-regulatory characteristics
of the MEFR with its favourable temperature coefficient,
and the potential for fuel dumping in the event of
faults but have also brought out the importance of
further careful study of fault conditions, including
cross-leakage with a fluid fuel, and of containment
requirements.

Parameters for Indirect and Direct Cooled Versions of a
Molten Salt Fast [Reactor - Summary

Reactor power {(total) MW(th) 6000
Gross electrical output MW(e) 2 x 1350
Nett electrical output MW(e) 2 x 1250
Steam conditions - TS MN/m2 16.73

PSig 2350

°c 565
Overall plant thermal % 41.8

efficiency '

Fuel salt NaCl:UClszuClj 60/37/3 mol %
Blanket salt NaCl:UCl3 60/4%0 mol %
Melting point 577°C (850 K)

37
NOT FOR _PUBLICATION (COMMERCIAL)

 

 

 

 

Indirect Dircet
beairn No 1 2 3 1
Goolant Lead Lead Helium lead
l'uel salt properties used (2) (1) (1) (2)
Volume of core m 17 16 16 37
Fuel salt volume external 3
TO core m* 21.5 28 22 15-18
Total breeding gain 0.29 0.3 0.3 0.53
Doubling time (net) years 25 28 24 21
Core inventory (including
external volume)
- Pu-239 equivalent te 9.5 10.8 9.4 9.2
- U-238 te 63 72 62 60
- C1 te 55 48 54 49
- Na te 11 13 11 10
Volume of blanket/external 46/5 48/17 53/22 83/5
volume blanket salt
Rlanket inventory - U-238 te 84 107 124 165
- Cl te 68 87 100 116
- Na te 14 18 21 2
. -1 -1
(1) Thermal conductivity taken as K = 0.5 wm =~ k .
(2) 70:30 fuel salt properties used with MP 557°C and
K = 2.0 w m™Ll k¥~1 for heat transfer calculations.
Neutrons Fluxes
Approximately applicable to all designs
2
(a) Max. n/cm s 2 X 1016
1
(b) At core/blanket Membrane n/cmgs 5 x 10 i
(¢) At Outer Vessel n/cmes 3 X 10%
1
(d) At heat exchangers n/cm S 3 x 10 J
Core Conditions Design 1 2 3 Y
Power developed in core MW(th) 5700 6000 6000 5700
Mean tTemp
Core inlet temperature °¢c 65C 660 670 680
Core outlet temperature °c 810 810 970 800
NOT FOR PUBLICATION (COMMERCIAL)
2

 

 

Max pressure in core vessel KN/m 1630 970 3140 L000/2700
psi 236 140 454 580,/%90

specific power rating:-

(a) based on net core volume MW/m3 340 375 375 154

(b) based on total volume MW/m3 154 138 158 110

including external fuel
salt volume

Core Heat Removal

Coolant Lead Lead Helium Lead
Curtain

Number of coolant circuits 8 8 10 8

Fuel salt flow (each) mj/s 2.45 3.14 1.22 Internal

Coolant flow (each) mj/s 3.6 4.6 260 kg/s 1.7

Coolant inlet/outlet temp °c 480/600 570/670 L400/850 370/630

Core Salt e

Heat exchangers (same number

as coolant circuits)

LMTD °c 190 112 190 -

lean heat flux w/cm2 372 106 107 -

Max heat flux w/cm2 410 147 155 -

Tube ID/0D mm 9/12 9/10 6/8

Pitch of tubes mm 19.20 11.5A 14.4 O

Number of tubes 3600 12,080 6164

Heat transfer length/ m 1.5/2.9 2.0/2.5 3.6/4.1

overall length of tubes

RBlanket Conditions

Power developed MW 300 900 300 300
(assumed for design purposes -

up to 15% ie 900 MW(th) could

be developed affer a 10 year

fuel cycle)

Blanket inlet temperature C 600 660 630 600
Blanket outlet temperature °c 650 760 700 650
Max pressure kw/m2 1630 1000 3300 1630
Effective blanket volume m” ke 48 53 83
Total blanket salt volume mj 51 65 75 88

including cooling circuits

39
NOT FOR PUBLICATION (COMMERCIAL)

Blanket Heat Removal
Coolant as for core
Number of coolant circuits

Rlanket salt flow (each
circuit)

Coolant flow (each)
Coolant inlet/outlet temp

Blanket Salt
lleat Exchangers

LMTD

Mean heat flux
Tubes ID/OD
Tube pitch

No off

Heat transfer length/
overall length

Pressure Drop

 

Core & ducts
Heat exchgrs

Fuel salt side:

Coolant (IHX
circuit)

: Heat exchgrs
Pipework

m’/s & 4 5 4
mB/s 0.8 1.43 1.7 0.8
mj/s 0.7 2.1 116 kg/s 0.7

°c  480/550 570/636 400/600 480/550

Coolant (steam generator circuit)

Steam generators and pipework

 

 

Pumps
Fuel Salt Pumps
Duty (each pump) Flow
Ap
At 70% overall effi- Power
ciency
Power normalised to
6000 MW produced in
core
Blanket salt pumps
Flow
ADp

190 100 155 190
572 95 107 372
mm 9/12 9/10 6/8 9/12
mm 19.2w1 11.5oa 11.0em 19.2 =
565 5500 5900 565
1.5/2.0 1.5/2.01.3/1.8 1.5/2.0
KN/mg 183 970 970 ) Internal
KN/m< 1450 1670 2170 )
KN/mg 812 1050 148 ) 5000/
KN/m< 184 300 70 ) 10000
(nominal)
KN/m° 1290 700 - 700
(if 670°C
upper temp
is acceptable)
No off 8 a3 10 -
mj/s 2.45 3,14 1.22 -
KN/m° 1630 2640 3140 -
MW 5.7 11.8 5.5 -
MW 6.0 11.8 5.5 -
No off h h 5 Y
m’/s 0.8 1.43 1.7 0.8
KN/m2 1630 2640 3140 -

40
NOT FOR PUBLICATION (COMMERCIAL)

 

At 70% overall effi- Power MW 1.9 5.3 7.5 1.9
clency
Power normalised for MW 5.7 5¢9 7.5 ST
Q00 MW(th) produced in
hlanket
(Cooliant Pumps
(a) Core THX only No off 4 i 5 8
Flow mj/s 3.6 4.6 260 kg/s 1.7
Ap KN/m2 996 1350 218 5000-10, 000
Power MW 5.1 8.8 135.4 12-24
(b)) Core THX + No off 4 4 5 4 (blanket
blanket HX HX only)
Flow mj/s 5.7% 6.7 376 2.1%
Ap KN/m2 996 1350 218 996
Power MW 8.1*% 13.0% 19.5% 3.0%

* For 900 MW{(th) blanket power

7{

 

 

Total Design Pumping Power Design 1 2 3 4
(excluding steam side)
Fuel salt MW 48 94 55 -
Blanket salt MW 23 21 30 23
‘oolant (a) THX ecircuit, MW 52 87 136 108-200
inecl by-pass
(b) Steam generator MW 34 14 - 14
circuit
157 216 oo1™ 145-257

¥ Combined installed pumping power for 6000 MW{th) from core
and 900 MW(th) from blanket.
In operation total power from core and blanket would be 6000 MW
as designed, therefore power consumed by pump would be reduced to
B7% of the design total capacity assuming perfect distribution,etec.

Includes nominal power for loss of Ap in coolant ducts to turbine but
no losses in turbine plant, intercooler, precooler or recuperator.
All coolant pressure drop will be made good by the turbine driven
compressors and all coolant pumping power is included in the overall
rlant efficiency, ie about 41% with reasonable recuperation.

Steam Generator Conditions

The conditions gquoted are applicable to Design 1. Little or no bypass
flow would be required for Design 4. Probably greater bypass flow would
- be required for Design 2, depending upon the maximum temperature difference
;hat can be tolerated in the evaporator section to limit lowest lead
temperature, if the superheater or reheat sections impose limits of upper

lead temperature a second bypass may be required. The helium cooled

41
NOT FOR PUBLICATION (COMMERCIAL)

Design 3 is assumed to be used with a gas turbine plant.

Coolant flow through each IHX and temp rise : 3.6 mj/s 480-600°C
Lead flow to each of 8 steam generators and 3 5
temp fall : 2.2 m/s 600-400"C
Steam flow from each of 8 generators : 10 Kg/hr
Evaporator duty (2 per generator), each MW ¢ 200 MW
Superheater duty (2 per generator), each MW : 136 MW
Reheat duty (2 per generator), each MW : 46 MW
'eed water 1nlet temperature, oC : 238
Outlet steam conditions, pressure MN/m2 : 16.3
superheat  °C . 565
temp ’
0
e e

Pumps for steam generator circuit

(lead side) Design No : 1 4

No off : 8

Flow mj/s : 2.2 1.7

Ap KN/m2 : 1290 700
At 70% overall efficiency

Power MW : 4.2 1.7

The following parameters refer to Design 1

(1)

Filling, Drain and Dump Tanks

 

Fuel salt (core) filling,
drain and dump tank

Net volume m3 130
Maximum heat m 18
Design pressure MN/m2 1.2
Maximum temperature o
(during fuel dump) C 750
provisional figure

Blanket sa%t filling, drain &

dump tank(2)
Net volume m3 130
Maximum head m 18
Design pressure MN/m2 1.28

o

Max temperature C 750
NOT FOR PUBLICATION (COMMERCIAL)

(1 and 2 are combined during a fast dump so that they will
contain, between them, all the fuel salt and one of the eight

steam generator circuits lead inventory of total volume 37 +
51 + 160, 1%* 247 m> + allowance for expansion tank capacities
total 260 m”).

Cooling for each tank, based on assumption that all the
core and blanket arrives in one tank 2%% of full power assumed,
subject to further investigation.

MW 150

Combined 1lst stage delay tank and core or blanket auxiliary
drain tank.

5

Net volume m 100
(Divided into two

equal parts if core

and blanket is to be

processed separately)

Maximum head m 18
Design pressure (to

give interchange- MN/m2 1.2
ability)

Maximum temperature oC 750
Cooling provided Not yet assessed

2nd Stage delay tank

 

Net volume m 20
(Divided into two equal

parts if required as

above)

Head, design pressure
and maxXx temperature
as for 3

Cooling provided Not yet assessed
Of f-gas system delay beds

(No detailed calculations carried

out, Volume pro rata to MSBR

power). Each tank contains
water cooled charcoal filled

U tubes
No units 16
Volume each (h = 2 x 8m dia = 5.5m) 38 m3
Total volume 610 m>

43
10.6

11.

12.

NOT FOR PUBLICATION (COMMERCIAL)

Cooling requirements
(Pro rata MSBR: 47 hr hold up system
Long delay system

Lead Filling and Drain Tanks

No of units

Volume each m
(each tank will hold
contents of one
intermediate circuit)

Maximum head (of lead) m

Design pressure MN/m

Design temperature C

Off-gpas system

Notional values based on MSBR

atm /Q/s
mj/s

Helium purge flow rate

Nominal working pressure,bar
of system (equal to
that of primary pump psig
suction)

Fission product flow
rate in helium
stream is approx-
imately 0.1%

Cleanup system
The highest flow rate is that required

for UCLy reduction, based on SG/MSFR
(71)P13 flow rate required:

 

fl/s
mj/s
GPM
Component weights
Matl.Vol
m>
Reactor vessel
including nozzles, flanges 9

by

Not yet assessed

7 .5MW

O0.7MW)

16

20 (10m long x
1.6 m dia)

18
2.1

500

14
1.4 x 10

15

2.3
2.% x 107~

31

Matl Wt
Te

68.8
NOT FOR PUBLICATION (COMMERCIAL)

Molybdenum core shell 1.4 14.3
including outer liner, tube
liners to HX's

Main heat exchanger shell 2.52 19.2
only - (1 off)
HX tubes and tubefil%tes 0.93% 5.3 (tubes)
(0.53m”) and (O.4m 4,1 (tubeplates)
HX connecting pipework 0.19 1.45
(1 set)
Blanket cooler HX shell 0.4 3.06

Only - (1 off)

Blanket cooler HX tubes 0.17 1.3
and tubeplgtes (0.08m>)
and (0.09m~)

Cooler connecting pipework 0.21 1.55
(1 set)

 

Alloy Steel

Total weight (te) of Mb and Steel
Reactor vessel - 68 .8
Core shell (Mb) 14.3 -
HX shells (8 off) - 153.6
HX tubes (8 sets) 42.4 -
HX tubeplates (8 sets) - 32.8
HX connecting pipework (8 sets) - 11.6
Blanket cooler shell (8 off) - 24.5
Blanket cooler tube and tubeplates - 10.4
(8 off)
BC connecting pipework (8 sets) - 12.4
Totals 56.7 314.2
REFERENCES

1. EIR-Bericht Nr 215 Molten Chlorates Fast Breeder Reactor.
Problems and Possibilities. M Taube, J Ligou.

2. ORNL 4541. Conceptual Design Study of a Single Fluid

Molten Salt Breeder Reactor. Compiled and Edited by
Roy C Robertson.

45
NOT FOR PUBLICATION (COMMERCIAL)
ACKNOWLEDGEMENTS

This report summarises the results of work at Culham,
Harwell, Risley and Winfrith. Acknowledgement is due to
the following:-

Culham

Mr
Mr
Mr
Dr
Mr
Mr
Dr

Adlam
Dullforce

S Peckover
Reynolds
Whipple
Jukes
Roberts

=SS g3

<t 3@

Risley

Mr R H Allardice
Mr J E Shallcross

Winfrith

Mr D M Harris
Mr W C May

Mr J E C Mills
Dr D L Reed

Mr G L Shires
Mr W E Simmons
Mr R J Symes
Harwell

Dr R C Asher
Dr J K Dawson
Mr B L Eyre

Mr B R Harder
Mr F Hudswell
Dr C F Knights
Mr A E Little
Dr G Long

Dr H P C McKay
Mr S F Pugh
Mr J C Ralph
Mr D Scargill

46
NOT [FOR PUBLICATION (COMMERCIAL)
APPENDIX I
FEATURES OF THE DIRECT SYSTEM

Considerable effort has been devoted to preliminary
consideration of the mechanics of fluids and heat transfer in
a MSFR which is directly cooled by a curtain of lead droplets
failling through the fuel salt in the outer regions of the
core. The lead would then flow to steam generators as in the
indirect case.

The main problems in a directly cooled MSFR are:-

(a) to achieve a sufficiently high core rating to give
a reasonable fuel inventory

(b) to disentrain the salt from the lead flow so that as
much as possible of the salt remains in the core

(¢) to prevent lead droplets recirculating within the
core and causing an excessive build up in the central
region.

Very little experimental evidence is yet availlable, but
from theoretical considerations it appears that core rating is
limited to 70 MW/m” for free gravity fall of the lead curtain,
and to meet egonomic fuel cycle costs a rating of not less
than 150 MW/m” is required. This could in principle be
achieved by inducing poloidal flow of the core salt with lead
flow injected helically at 30 C to the horizontal (to increase
the path length for heat transfer and separation) at velocities
in the range of 20 to 30 m/s to give mean salt velocities of
about 10 m/s.

Theory indicates that to achieve the required rating a
curtain of lead droplets of 3 mm mean diameter and a packing
fraction of 0.35/0.4 is necessary; Jet pumps would be required
at the lead injection point to accelerate the salt to give
low velocity slip relative to the high lead velocity required
for the necessary throuput.

A core height/diameter ratio in the region of 1.5 is
probably required to give sufficiently long paths for heat
transfer and separation of the lead from the salt in the
lower core regions.

Preliminary experiments have indicated that heat transfer
is likely to be adequate. A theoretical model was established
to combine the heat transfer with the separation but the
performance of the latter aspect is very doubtful as little
experimental work has been done at velocities and densities
approaching the real case. A further limit on performance
arises from the hydraulic losses in the core as lead jet
velocities in excess of 20 m/s would require pumping powers
greater than that of the indirect system and local losses 1n
the jet pump region would be high. Another area of uncertainty
is how, following the downward helical flow in the outer regions,

47
NOT FOR PUBLICATION(COMMERCIAL)

the salt can be induced to flow smoothly upwards in the central
region of the core and enter the jet pump inlets without undue
losses or the need for elaborate guidance systems. A large
amount of experimental and theoretical work would be required
to investigate these problems and it appears that alternative
coolant patterns would have similar uncertainties.

Figure 4b is an illustration of an attempt to combine
the features discussed above and shows a radial arrangement
of jet pumps with the lower part of the core vessel shaped to
give coarse separation within the core and further separation
by means of curved ducts at the core exit.

A5 the quantity of lead flowing is 106 mj/day, very effi-
cient separation to ppm levels is required to prevent excessive
salt hold up in the external lead circuit itself, where 1n
addition salt freezing could occur on the cooler surfaces of
the steam generators. In the latter case daily or even more
frequent thermal cycling would be required to reduce the total
amount building up.

In the absence of information on performance or cyclone
separators under these conditions, it has been assumed that
two stages of cyclonic separation will be adequate and a tent-
ative reactor layout as shown in Figure 4a has been drawn up.
The parameters given in this Appendix, which also apply to
Design 4 referred to in the Main Report, have been drawn up
on this basis.

The use of jet pumps and lead separators external to the
core has been shown to involve too great an inventory.

Steam generators would be similar to that proposed for
the indirect system. There is no restriction on the lead
return temperature to the reactor as long as it is above 1ts
melting point but the upper limit of reactor outlet lead
temperature will depend on thermal stresses in the steam
generators. Two sets of lead pumps are arranged in series
with a steam generator by-pass. This arrangement allows flows
meeting the core hydraulics and heat transfer requirements
and the steam generator requirements to be adjusted independently
since it cannot be assumed that these would match the pumping
pressure in the reactor.

To maintain the correct flow patftern and separation
conditions in the core almost constant flow seems desirable.
Power control would be achieved by changing the coolant
temperature range using the steam generator by-pass or possibly
by changing the depth of lead pool in the bottom of the reactor,
thus varying the amount of fissile salt in the core.

In order to avoid any difficulty with increase of
reactivity in the immediate period following a cessation of
lead flow, a U-leg is provided in the lead entry lines, which
ensures that salt will be displaced from the core region by a
lead pool building up at the bottom of the core vessel.

48
NOT FOR PUBLICATION (COMMERCIAL)

 

Bianket cooling wlll be indirect using external heat exchangers
as 1t would he difficult to provide the correct geometry for direct
cooling (for which the basic requirement does not exist as ratings
are comparatively low and inventory considerations are not so
restrictive).

TABLE A.1
MSFR_DIRECT LEAD CURTAIN COOLED SYSTEM - PARAMETERS

NB A1l values quoted are nominal and based on preliminary
information which is subject to change.

Core Model - Helical poloidal flow with jet pumps

Nominal rating 150 Mw/m3

Core thermal power MWt 5700

Blanket thermal power (5% of MWt 300
total)

Core height m 5

Core diameter m 3.2

Giross core volume m 40
Specific power in core MW/m3 142

Lead Coolant

 

Total lead flow rate to core  Kg/s 142,000
m /s 13.5

Temperature at core inlet °c 370
Temperature at core outlet °c 630
Mean core salt temperature °c 680
Maximum core salt temperature °¢c 800
Tvypical Curtain Conditions
ingle of descent of lead O jOO

(1) (2)
Curtain width mm 600 680
T,ead fraction in curtain 0.4 0.35

49
NOT FOR PIBLICATION (COMMERCIAL)

Curtaln inner diameter m 2.0 1.84
Horizontal c¢.s.a. within m2 3.14 2.65
d = 2m

Horizontal curtain c.s.a. m2 4,86 5.35
Salt flow rate (approx) mj/s 20 25
Mean lead velocity in curtain m/s 10.5 9
Salt veloecity in curtain {approx) m/s 10.0 9.95

Jet Pump Conditions

NB It is assuned that the salt enters vertically

Angle of jets and vanes to O 300
horizontal (1) (2)
Lead injection velocity m/s 30 20
Salt velocity in throat m/s 29 19
No of jet pumps (not optimised) 60 60
Pitch 6° 6°
Length of slot in each unit mm 600 600
Mean width of slot mm 13.2 16.5
Thickness of slot sides (each) mm 1.4 1.9
Overall mean width of injection  mm 16 20.3

slot at outlet

Mean width of salt passages mm 10 16
either side of slots

Mean width of blading between mm 34 15
jet pumps at throat

Ratio outlet area/throat area 1.9 1.28
(Note: Further deceleration must occur in curtain zone)

 

 

Lead Circuit Pressure Drop v D Possible
(based on jet velocity of 5 extra

20 m/s) m/s KN/m KN /m2
Top plenum > 50

Inlet to jets 20 500 +500

50
NOT FOR PUBLICATION (COMMERCIAL)

Klf loss at Jet

PE due to height of core
Outlet duct (150° bend)
Friction

Cyclone separators (1 set)

External lead circuit inc. steam

generators

 

20 1500 +500
550
10 500
50

6.7 1180 +1180 (if
second set)

900

5040 +2180

= 1220

Thus,the overall pressure drog of the lead lies between

5000 kW/m® (725 psi) and 3000 kW/m

velocities of 20 m/s.

(1160 psi) for jet pump

The pressure drop might increase to 10,000 kW/m2 (1450 psi)
for 30 m/s jet velocity (i.e. an additional 2000 kW/m<).

Total pumping power including 20 MW(e) for blanket would be
117 MW(e) minimum based on 20 m/s jet velocity and one set of
separators or possibly double this figure if 30 m/s jet velocity
and two sets of separators are required.

Volumes anc¢ fuel salt inventory
(very approximate)
Core

Blanket (5 m high x 3.2 m ID
x 5.2 m OD)

Lead in reactor vessel when
running

Lead in 1 set of 16 separators
and connecting pipework

amount of lead containg
salt

(If second set of separators
required =

For 5% salt content in lead
(With two sets separators

Volume of salt in return lines

m” 37
m 83
m3 30
m3 70
m3 100
m 172)
mj 5
m> 8.6)
m 5.6

51
NOT FOR PUBLICATION (COMMERCIAL)

LLead in steam generator m3 1000
circuits (8)

Salt content if 0.5% m3 5

Possible t%tal salt inventory outside core = ~_ 15 mj

(or 18.6 m” if two sets of separators)
Dump Volumes for all core + lead 1200 m3 (excluding
(of 260 m> for indirect system blanket )

with one lead circuit dumped)

Pressures (20 m/s jet velocity)

KNgmg Psi

At rest (no circulation) at 727 105
bottom of separators

When running (1) at bottom 2700 390
of core (due to back pressure
from separators)

(2) Internal pressure in jet 3500 500
slots between 2800 and 4200
KN/m2 depending on
situation of losses

(3) Differential pressure between 1500 220
salt and 1ead in slots

he
NOT FOR PUBLICATION (COMMERCIAL)
APPENDIX IT
INTERNALLY COOLED MSFR
I INTRODUCT 1 ON

The possibility of internally cooling the core of a MSFR
using a tubed heat exchanger situated in the outer core region
and blanket salt or lead cooling was considered in the early
stapges of the recent MSFR work. Within the ligits of salt
temperature considered at that time (i.e., 800°C) the performance
was poor and gave very low breeding due to the absorption effects
of' the material in the core.

Taube(l) has since proposed a system in which the whole core
is virtually occupied by cooling tubes through which blanket salt
is passed. This has the merit of removing the heat at source and
although an upper salt temperature of 998°C is quoted, the max%mum
tube wall temperature under normal operating conditions is 854-C.
Some comments on this concept are given below.

2. DISCUSSION

Examination of ETR Bericht Nr 215¢1) indicated that the
reactor physics results quoted had been based on Bondorenko's
data and from previous experience 1t was known that this data
set would produce a much harder spectrum (and therefore more
probably a more favourable performance) than with the FD4 set used
for the MSFR studies. To study this further, some calculations
with FD4 on a simplified model of the Taube concept were carried
out and these indeed revealed that FD4 would give much lower levels
of breeding gain, so much so that it required the replacement of
natural chlor%ag by Cl-37 to bring values back to those quoted by
Taube. Ligou has repeated the calculations for the Taube concept
using data from the GGC3 file. This gave a somewhat higher
critical mass than in the initial report or in the FD4 calculations,
but a breeding gain close to that of EIR-215, i.e., substantially
higher than FD4. The latter difference appears to be mainly due
to the much lower average chlorine cross-section of the data Ligou
used. Clearly there is scope for further comparative work and
improvement of data.

Another potential source of difference which could depress
the performance of the concept proposed by Taube was that the
thermal property data used in the reference MSFR cases were
materially less favourable than those used in EIR-215 and this
could have a substantial effect on a scheme in which good heat
transport and heat transfer within the reactor core is highly
important if too much neutron absorbing material for the heat
transfer surfaces is to be avoided.

It was therefore decided to carry out an outline study of
the concept using data as used for the MSFR studies, and for a 6000
MW(t) size, so as to be able to make a comparison with the indirectly
cooled version.

55
NOT FOR PUBLICATION (COMMERCIAL)
3, OUTLINE SCHEME

1t was assumed that the blanket salt, from which with this
scheme both the core heat and that generated within the blanket
itself has to be removed, is circulated to external coolers similar
to the intermediate heat exchangers of the indirectly-cooled design
but readjusted to take advantage of the less costly inventory
ef'fects of blanket salt compared with core salt. It was further
assumed that a secondary coolant would be needed for possible
leakage and contalnment reasons.

The effect of the poorer physical property data is substantial.
In the Taube arrangement the cooling tubes extend across the whole
core region and a low core salt velocity of about 2 m/sec is all
that is needed. With the MSFR data, core salt velocities have to
be much higher and even at 5 m/sec which was taken as a typical
permissible uppgr limit, the maximum salt temperaturgs are in the
range 1100/1160°C with tube wall temperatures of 900 C for the same
secondary coolant conditions as in MSFR (indirectly-cooled).
Furthermore the considerable increase in circulation rate of the
core salt necessitates the use of large pumps and the core has
been rearranged to put heat exchange surfaces on as large a
diameter as possible to give reasonably low velocities across
the tube banks. Figure 12 shows a schematic arrangement in which
a clear central zone is created to reduce the cross flow velocities
and forms a continuation of the pump diffusers so as to minimise
inventory. Blanket salt vélocities must be kept up if tube-wall
temperatures are not to become excessive and this leads to high
pumping powers and very high pressure differentials.

Physics calculations were made using a simplified cylindrical
approximation to the arrangement of Figure 12 and they immediately
brought out the importance of minimising tube wall thickness and
pitech to reduce fuel inventory as illustrated in Table 1.

The parameter list of Table 2 is based on tubes of 3 m
length, 5.0 mm.i.d., and 5.6 mm.o.d., on a pitch/diameter ratio
of 1.%5. This tubing is very thin walled and may have exceeded
the practicable lower limit. It would need thorough testing and
demonstration.

b, PERFORMANCE

The preliminary investigations of the internally cooled MSFR
described indicate good nuclear performance. The Pu-239 inventory
of between 6 and 7 te for 6000 MW(th) is low, comparable to Taube's
proposal, and less than for the indirectly-cooled systems described
although for similar maximum salt temperatures the differences
would be small.

The estimates indicate a reduction of doubling time from 25
yvears for an indirect system to 20 years for the internally cooled
Cl-3%7 system on the basis of external reprocessing (by aqueous
methods) where the external hold up represents a significant
fraction of the system inventory. Clearly as potential reactor
performance improves, the incentive to develop close-coupled
reprocessing increases. This reduction in doubling time is made

54
NOT FOR PUBLICATION (COMMERCIAL)

at the expense of high temperatures in the plant and with heat
exchanger tubing subject to high neutron fluxes, although damage
may anneal out. Also there is the requirements for an additional
set of full capacity heat exchangers external to the reactor.

A first estimate suggests that the capital and running costs
due to the extra heat exchangers and high total pumping power with
the extra cost of Cl-37 in the external heat exchangers cooling
the blankst almost offset the saving in fuel inventory; thus
there is little direct economic gain.

5. CONCULSIONS

In principle an internally-cooled system should offer the
prospect of low fuel inventory and although this preliminary study
has not shown the margins compared with the indirectly cooled
system that initially seemed possible, the concept is sufficiently
interesting to Jjustify further investigation to optimise the
nuclear and thermo-hydraulic performance. If results are
encouraging further investigation will be required to determine
the limits of corrosion ang strength of Molybdenum or its alloys
at temperatures up to 1200°C. The design problems associated
with high velocities and high pumping pressures will need far
more consideration than has been given in this outline scheme.

55
96

TABLE

1

NUCLEAR PERFORMANCE OF A PRELIMINARY UKAEA DESIGN OF AN

INTERNALLY COOLED 6000 MW(t ) MSFR USING CL3%7 SALT

 

 

 

 

. . . Without
Heattegghanger Effecglve criticall Total With 9 months processing hold-up hold-up
pitch?d;gieter Vgiime Mass Breeding
ratio (1) (Puz29) | Gain system | SRS | poubling |Doubling
Inventory y Time Time
Inventory
M3 te Kg/Pu239 Years Years
MW(t
1.3 12.5 2.31 .225 5.94 0.99 19.9 11.7
1.7 19.8 3.85 .2b3 7.26 1.20 19.1 14.0

 

 

 

 

 

 

 

 

 

(1) The effective core volume including some of the salt in the connections from
the pumps to the core heat exchanger tubing.

 

{TYVIDHANWNOD) NOILYDITdNd HO4 ION
NOT FOR PUBLICATION (COMMERCIAL)
TABLE 2
MSFR Internally Coocled Design Parameters

Maximum core power (at start fuel cycle) MWth &, 000
Height of core (selected) m 3.0

Core-Blanket Heat Exchanger (in core)

Number of tubes (in annular form) 60, 000
Tube internal/external diameter mm 5/5.6
Pitch/outer diameter ratio 1.3
Pitch geometry Square
Dimensions

Upper central uncoocled zone core diameter d1 m .72
Lower central uncooled zone core diameter do m 1.5
Cuter lower core overall diameter dz m 2.5
Pump suction diameter (for 4 pumps) each d} m 0.63
Volumes

Fuel salt in core zone (incl HE) m% 8
Fuel salt external 1n pumps and connections m 5

*low rates

Fuel salt flow rate m5/s 12.4
fuel salt velocity in HE m/s 7
Blanket salt flow rate mj/s 17.6
Blanket salt flow velocity in HE m/s 15

Heat Transfer coefficients

 

Fuel salt/wall W/em©-C 1.1
Wall (O.%mm thick molybdenum) w/cmggc 35.0
Wall/blanket salt W/CmEOC 1.9
Overall | W/cm©-C 0.6
Heat Flux

Mean heat flux w/cm2 190
Temperatures

Log mean temp difference required OC 317

; . 0

Unccoled core zone temp rise oC 124
LTD/GTD C 290/390
Max fuel salt temp 1100
Blanket salt inlet/outlet temp to HE °c 590/810
Pressure drop (coreheat exchanger only)

Blanket salt Ap MN/mg 6.5 (
Fuel salt Ap MN/m 1.2 (

57
NOT FOR PUBLICATION (COMMERCIAL)

Pumping power (core heat exchanger only)
Fuel/Blanket salt MW(e) 21/164

58
CORE
BLANKET

FUEL SALT PUMP—8 off
BLANKET COOLER— 4off

ISOLATION VALVE

LEAD INLET

LEAD OUTLET

GRAPHITE

HIGH TEMPERATURE
RESISTANT LINING

12 DRAIN LINE

13 FLOW DISTRIBUTOR

14 SUPPORT FLANGE

LD P g T B o

= O

BLANKET COOLER PUMP-4off

INTERMEDIATE HEAT EXCHANGER-8off

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

T
-~
N ®
L »
[ A
! —
| 4
"
TE' -]
. -] ;
\ :
‘\ y ’
DESIGN |

INDIRECTLY COOLED MS.F.R.
( LEAD COOLED)

6000 MW Th

FIG. I
TOTAL CORE SALT INVENTORY
SCHEME " A — 42.8) M3 \

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SCHEME "B — 40.09M?
W
' {
- i: | | ,'_ j.‘.__..
- Vol Vo
SCHEME ' B’ | SCHEME A’
N |
--
O I 2 aMm
boaastygnd 0 b )
SCALE M.SFR.—ALTERNATIVE DESIGNS OF INTERMEDIATE

HE AT EXCHANGERS FOR LOW CONDUCTIVITY FUEL SALTS FIG. Ib
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

CORE
BLANKET
INTERMEDIATE HEAT ENCHANGER - 8 off
FUEL SALT PUMP — 8§ off

BLANKET COOLER — 4 off

BLANKET COOLER PUMP - 4 off
INTERMEDIATE COOLANT MEADER
INTERMEDHATE COOLANT tNLET
WNTERMEDIATE COOL ANT OUTLET

IO DRAIN LH9E & VALVE

LD B g o s W -

 

 

 

 

 

 

jfl

 

 

 

 

 

SCRAP DETAIL SHOWANG
CHANGE N TUBE SECTION
AT _LPPER ENDS OF HEAT
EXCHANGER TUBE BUNDLES

DESIGN 2

INDIRECTLY COOLED MSFR.
( LEAD COOLED )

WITH ALTERNATIVE CORE & INTERMEDIATE
HEAT EXCHANGER LAYOUT
6000 MW Th.

FIG. 2
 

 

  

 

 

    
   
  

 
 

CORE

BLANKEY

INTERMEDIATE HEAT EXCHANGER - 10 off
FUEL SALT PUMP - O off

BLANMET COOLER - 5 off

BLAMKET COOLER PUMP — 5 off
ISOLATION VALVE

INTERMEDIATE COOLANT INLET
INTERMEDIATE COOLANT QUTLET

S oy N e Ry

-

 

10 GRAPHITE
It DRAIN LINE & YALVE E]— 1 1 3 - _jm DESIGN 3
iz SueeoRT INDIRECTLY COOLED MSFR

PRELIMINARY DESIGN OF
HELIUM COOLED VERSION
6000 MW Th.

FIG. 3

~. . -~ -
 

 

 

 

 

 

 

 

 

 

 

STEAM GENERATOR
CIRCWNT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I CORE

JET PUMPS - 6001 ?

CORE CODLANT LEAD PUMPS— B ot
LEAD BOOSTER PUMPS - Bor
CYCLONE SEPARATORS - $Soft
ADOIMIONAL CYCL . SERARATORS - 10f T req,
SEPARATED SALT RING MAIN
SEPARATED SALT PUMP-) oft

9 BLANKET

10 BLANKET COOLANT FLOW DETRBUTOR
i1 BLANKET COOLERS-4 ot

12 END BLANKET COOLING RING MAINS

13 SAMPUING CONNECTION

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

© 9 & A WK

 

 

 

 

 

 

—=~-& SALT-CORE
———a LEAD
—— SALT—BLANKET

 

 

DESIGN 4
DIRECT COOLED MSFR.

WITH JET PAUMP ASSISTED CIRCLL ATION

 

 

( LEAD COOLED)
5000 MW Th

FIG. 4q

 
e ( (ORE SALT

e | E A0

 

 

 

P

 

 

 

SIMPLIFIED DIAGRAM OF DIRECT SYSTEM
SHOWING LEAD CURTAIN AND POLOIDAL FLOW

- BLANKET

(Cooled by
external heat
exchanger )

| CURTAIN

Salt'squeezed
 out’of lead in
this region.

Further sait

] separation zone

 

SEPARATED

‘ .\\\ SALT

LEAD

ouT-10
SEPARATORS

FIG. 4b
 

)
"/

mimm/

,.'.js\\ .

R

i

TN

TN
AT W

NCTTT 0
oL W
Ll T AL M
L L e
,-'.Iifl..'i h\. y
) \ ]

/a

/

 

 

 

 

 

CENERAL VIEW OF 2500 MWe MSFR (INOIRECT SYSTEM— BASED ON DESIGN 1)

FIG. 5
r—"fi#;m_ 38 M ________,‘,,,fif..i;.,fi,.._____ M

     
 
   
 
 
  
  
  
 
   

 

 

 

 

 

 

 

 

TURBINE HALL
STEAM MAIN

DECAY HEAT REMOVAL EXCHANGERS

LEAD DRAIN / FILLING TANKS

REACTOR VESSEL STORAGE / PLANT MANTENANCE CHAMBER
HP CYLINDER

P CYLINDER

L. P CYLINDERS

ALTERNATOR 1350 MW (E) GROSS

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O oM O™

 

SCALE

 

 

 

t—

 

 

 

 

 

MSFR.
g BUILDING LAYOUT

(INDRECT SYSTEM-BASED
ON DESIGN |

2O M

FIG. 6
1O REACTOR VESSEL
Il BLANKET COOLER
12 INTERMEDIATE HEAT EXCHANGER iH.X.
13 DELAY BEDS
i4 STEAM GENFRATORS
15 CLEAN-UP PLANT ROOM
16 BLANKET COOLING PUMPS
17 FUEL SALT PUMPS
18 1HX LEAD CIRCULATION PUMPS
19 STEAM GENERATOR LEAD CIRCULATION PUMPS
20 FUEL STORE FOR POWER TRIMMING
2t CORE FILL /ORAIN & PART MAIN DUMP SYSTEM
22 (AUX CORE & BLANKET DRAIN TaNK
{PRIMARY DECAY TANK
23 BLANKET FILL /DRAIN & PART MAIN DUMP SYSTEM
24 SECONDARY DECAY TANK
25 TEMP INSULATING PARTITION WALLS & FLOORING
26 TEMPERATURE INSULATION
27 ACCESS CORRIDOR
28 CONTROL ROOM

 

 

 

 

 

 

 

 

 

 

 

SECTION 8:8B SECTION C:C

SEE FIG4h FOR ITEMS -9

TO NATURAL OR FORCED DRAUGHT

- COCLERS
oM IOMm 20M
L b
SCALE

 

SECTION A:A FlG 7

M.S.F.R. BUILDING LAYOUT (INDIRECT SYSTEM—~BASED ON DESIGNT)
 

 

 

 

 

 

 

 

 

 

 

 

NN NN NANNRKRE

 

 

 

 

 

 

 

b— - ——

SG

 

 

 

|

 

 

 

e
——————

 

 

 

 

 

 

SG - STEAM GENERATING PLANT

 

DRG No 4W S0362

SIZE COMPARISON OF ™MSER. WITH

SCALE S0 METRLS

\
looo

 

 

 

 

 

 

 

 

 

 

 

 

CFER. & MSBR

     

O

FIG. 8

_ _ |
T A
' i 415
37 -
NN s ‘L 1} <% 1
28 "i % ‘ ]
A
C.F.R,
2% 1300 MWe B
!
!
4
i
— 160

|
B

 

 
NATURAL DRAUGHT

HEAT REMOVAL
(Or dies¢l=electric

driven forced
draught cooling

towers)

L

tSteom

SALT / WATER
HEAT EXCHGR

Natural circ.
salt cooling
to remove
decay heat
from drain
tanks

-

1|

 

!

~

N
SN

N \\ \‘\\

< N
\ ™

NSNS

r" Z J/ “ //

 

 

 

=

 

 

 

 

 

 

 

 

 

 

s

1

BLANKET |
COOLERS | 4
40FF |
'

1

'

 

 

 

Connections t
cleanup and ofcf)'

  
    
  

)

/
BLANKET

e 308 systems

  

 

Connechons fo
cleanup a

gas systems >3

   
    

 

 

 

 

 

 

 

 

 

 

 

 

 
 

CATCHPOT DRAIN
( Normally closed )

 

 

 

8 OFF

D DUMP
YA

RAIN &
LVE

S

 

 

 

 

 

 

|

-~

. T/A SETS-2 OFF /J’—}A
btk o o | :

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 
 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

——n AT e — o 3
% Vf HP P
I p ’ . i e ---=-+-—-4-—i -T_—m_ P
~F30 4 e L
n % g \1 FEED
s cooun | [corcmseel T4
r CIRCUIT PUMPS 1 # HEADER (M
b ’ TANK L
| 4 FROM SECOND E— [
" | EVAPORATOR b
L . L]
: A i {
' iy :
: 4 POSSIBLE ——
; TO ADDTN T2
; K fi, FEED | 3 3
! /] HEATER
' g
SUPERHEATER
7 TO SECOND
STEAM GENERATOR IT
CIRCUIT PUMPS A PORATOR
’ L
\Direcr enriched
fuel addition EHEATER
DRAIN
TANKS
SOLID
FUEL

 

 

 

 

 

 

| |

 

 

 

 

 

 

 

 

TO
PROCESSING

~——

 

 

 

\/

 

 

SECONDARY DECAY
TANK(S)

tank & part main
system.

\Blcnkef fling sdrain
dump

)]

 

 

 

Division-if
Erocessmq 2 .
lanket salt s required

      
 

Auxilliary core ¢
blanket drain tank.
Primary decay tank

separgte

 

 

 

U

 

et e sl

|
3]

 

 

 

 

 

 

 

 

 

U

FUEL STORE FOR
POWER TRIMMING

 

 

Core filling ¢ drain tank
¢ part main dump system

EHEATER

CORE FUEL FILLING POINT
(Blankef filling similar )

FIG.9

FLOW DIAGRAM—INDIRECT SYSTEM

I —INERT GAS SYSTEM

SHOWING

REMOYAL SYSTEMS.

PRIMARY & iNTERMEDIATE
STEAM GENERATION ALSCO FILLING, DRAINING,DUMP & DETAY

CIRCUITS, BLANKET COOLING &

HEAT
PURGE FLOW TO‘
OTHER PUMPS

PRIMARY SYSTEM @ —-
(FUEL € BLANKET)

 

ACCUMULATOR

 

 

 

 

 

 

 

 

 
 

 

 

     
  
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

+——gp» OFF GAS SYSTEM

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 SURGE
| TANK
FROM OTHER - COMPRESSOR |
SEPARATORS :
_—-M» i
— -
UCI, REDUCTION f O o f
_BEDS ) ENTRAINMENT |
(Addmon of Uranium ) SEPARATOR i
el ;
’ MAIN DELAY
FUEL SALT “;‘;85 LONG DELAY
cm%hgmo | Xa HOLDUP
| 2c.{m.
GAS .
LIQUID SEPARATOR - 47 Hr Xe Hefm.
Na Al Cla = %=1 HOLDUP SYSTEM
OXIDE |
REMOVAL BUBBLE | PARTICLE 9cfm.
REDS GENERATOR n TRAP
FUEL SALT | (If required) :
DRAIN TANK | E”,.EQ'NT#C?{
I
CYCLONE SEPARATION !
OF ALUMINA
ACCUMULATOR
| SURGE
i TANK
. COMPRESSOR

FLOW DIAGRAM —FUEL & BLANKET SALT CLEANUP &

 

TO OTHER BUBBLE
GENERATORS

SALT CLEANUP SYSTEM

OFFGA

OFF GAS SYSTEM

(with acknowledgements to O RNL)

SSING

FIG IO

 

 

 
 

Y
FROM
CLEAN VPt

 

CORE HEADER

 

 

 

 

LEAD PUMPs (8 orF)

 

 

 

 

 

T/A SETS (2-0FF)

 

 

P

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

O

. TANK
BLANKET ‘ FRoM CLEAN UP
HEADER p—% PRIM
TANK T Oom RIMARY
T BLANKEY
K f
- ” RADIAL
Y 1 8L ANKET

 

 

BLANHET
COOLERS

(4 orfF)

 

!

 

 

 

 

lLANK’E! ;

SALT

LEAD

  
 

LEAD CURTAIN
|

 

 

SEPARATED

 

 

 

 

 

 

 

 

  

/SALT RING MAIN

A

 

%‘Q-M

 

To 8L ANKET LEAD
CLEAN UP / TO CLEAN UP -
‘ - -y
BoTToM SEPARATORS l"‘ 1
BLANKET

!

(’6 "‘F.\) \x 7
/ v
Possi8LE ADPDITIONAL }

separaToRs (16 a“)

 

 

FLOW DIAGRAM

FIG. 11

 

 

STEAM

GENERATOR

QIRCUIT

LEAD POMPS
(8 otr)

 

   
  
  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

P | —
HP
]
\ L\l
CONDE SER
FLED TRAK
LEAD
MEADER [ l
' TANK
l FROM 56COND
1 EVAPORATOR
NP
SUPERHEATER
To §KCOND
Rfn voiT,
—
EVAPORATOR
DRAIN ;
REREATE R ;o
i
STEAM GENERATORS (8 UNITS)
EACH WiTH 2 SiTS OF EVAPORATORS
SUPCRIHEARTERY L& 1REHEATERS
s o P —e—
- e D

— DIRECT SYSTEM. PRIMARY & INTERMEDIATE

BLANKET COOLING & STEAM GENERATION.
 

 
 

CORE SALT PUMPS-«\
(4 off ) N
N
BLANKET SALT TO
EXTERNAL COOLERS
/ TERNA

 

 

 

 

 

 

 

 

 

A > (8off )
./
R
P};
S
NOTE. REFLECTOR NOT SHOWN FlG |2
INTERNALLY COOLEL
?“ 1 Il I ZL i { 1 L} i ?M M S F R
SCALE .F.

Note: Dimensions relate to tube pitch/diameter & 14 6000 MW Th
NOT FOR PUBLICATION (COMMERCIAL)

DISTRIBITION

WINFRTITH

Mr H Cartwright A32
Dr J E R Holmes A32
Mr J Smith (12) A%2
Mr C P Gratton A32
Mr D M Harris B40O
Mr W C May A32
Mr J E C Mills A32
Mr G L Shires BA4O
Mr W E Simmons (8) A32
Mr R J Symes Az2
CULHAM

Dr R & Pease

Mr J H Adlam

Mr T Dullforce

Mr R 3 Peckover

Dr J A Reynolds

Mr R T Whiopple

Mr J D Jukes

Dr K V Roberts

DOUNREAY

Dr R H Allardice

HARWELL

Dr R C Asher
PDr J K Dawson
Mr B L Eyre

Mr B R Harder

Mr F Hudswell

Dr C F Knights

Dr B A S Lister

Mr A E Little

Dr G Long

Dr H P C McKay

r S F Pugh

Mr J C Ralph

Mr D Scargill

Dr P M S Jones PAU 156
Mr H Killingback PAU 156
Mr J Richards PAU 156
RISLEY

Dr T N Marsham
Mr C Beveridge
NOT FOR PUBLICATION (COMMERCIAL)

 

RISLEY (Continued)

Mr
Mr
Mr
Mr
Dr

o=z O

A Farmer

B TIlirfe

H MclLaren
stephenson
D Smith

RISLEY

BNEFL

Mr il
Mr D

A Hughes
S Briggs

LONDON OFFICE

 

Mr R
Mr I,
Mr S

 

I. R Nicholson
G Brookes
T C C Hood

Gibbs
K Wright
Jones

Preston

D Vaughan

CEGB Cheltenham
CEGB Courtenay House
Berkley Laboratory

Inverlair Avenue