2055-8

Joint ICTP/IAEA School on Physics and Technology of Fast Reactors

Systems

9 - 20 November 2009

Fast Reactor Core Design

Module 2 : Core Neutronics

P. Puthiyavinayagam
IGCAR
Kalpakkam
India
IAEA/ICTP School on Physics and Technology of
Fast Reactor Systems

Lectures on Fast Reactor Core Design

Module 2 : Core Neutronics

Compiled & Delivered by

P,Puthiyavinayagam
IGCAR, Kalpakkam, India

The Abdus Salam

International Centre for Theoretical Physics
Trieste, Italy

Nov 9-20, 2009

2.0 INTRODUCTION

In a fast reactor, the neutrons have high energy in the core centre. But outside the
core in the shield regions they get slowed down to low energies. Thus knowledge of basic
neutron cross-sections accurately in the entire energy range of neutrons is imperative.
Compared to thermal neutron reactors it is essential to use much larger number of energy
groups in the core design of FBR as the processes of neutron slowing down and capture in
the fast and resonance energy regions need to be modelled by international collaboration
between different countries, measurements and evaluation of huge nuclear data bases have
been done through IAEA co-ordination.

The objective of the core design is safe operation, economic power generation and
self sufficient fissile material production. This requires the design of core with highest
acceptable linear pin power, core burnup capability, suitable fuel pin diameter and the most
suitable safety related parameters. Selection of fuel pin diameter affects the fissile inventory,
fuel cycle cost and breeding gain. The core height is to be decided based on the coolant
pressure drop requirement. To achieve better breeding ratio, the reactor can be designed
with radial and axial blankets.

Once the volume fractions of the fuel, coolant and the structural materials are
determined from the above considerations, the core physics design starts with the
determination of the fuel enrichment necessary to keep the reactor critical at full power with
sufficient excess reactivity for burnup compensation. Distribution of neutron flux and power
throughout the core, breeding ratio achievable for the fuel/blanket configuration, refuelling
scheme for the core and blankets to ensure maximum availability of the reactor and highest
possible burnup for each of the fuel and blanket subassembly, safety related parameters like
material worth, Doppler reactivity, delayed neutron fractions etc. are the outputs of the
physics design studies. Optimization of power production in core is essential for economic
power generation. Precise balance need to be maintained between power produced in
different locations of core. This needs to be achieved with every refuelling. This is possible
only if the fuel isotopic compositions in different fuel and blanket subassemblies are reliably
estimated as a function of burnup which is a challenging task. The core design also ensures
that the core temperature and power coefficients of reactivity are negative so that any off
normal increase in temperature or power leads to a reduction in reactivity and power. The
fuel management scheme is decided based on the targeted peak burnup for which the fuel
pins are designed. Another important part of the physics design is the determination of the
number and layout of the absorber rods (AR) for reactivity control during normal operation as
well as in accident situations. Adequate worth has to be provided for absorber rods for safe
shut down of reactor in all eventualities. Their locations and B-10 enrichment need to be
optimized. For achieving reliability targets two or more independent shutdown systems are
normally deployed in power reactors. The interaction effect between the absorber rods of
the shutdown systems needs to be studied carefully.

The basic physics concepts covered in this lecture are: (i) breeding (ii) doubling time
(iii) reactivity effects (iv) reactor control and (v) core shielding aspects. Detailed coverage of
reactor kinetics aspects is beyond the scope of this lecture and it is covered in other
lectures.
2.1 PHYSICS OF BREEDING

Breeding is the process of converting a fertile nuclide into a fissile nuclide. A neutron
capture in fertile nuclide U238 followed by two β - emissions produces the fissile nuclide Pu239.
The breeding ratio (BR) is defined as the ratio of fissile material produced (FP) to the fissile
material destroyed (FD) in a cycle either through fission or through capture.

2
 FP
BR =
FD

Fuel cycle is the time between periodic refuelling. When BR is less than unity, this ratio is
known as conversion ratio (CR).

Breeding Ratio

The expression for the breeding ratio can be derived as

⎡
( ) ( ⎤
)
~

⎢⎣ν − 1 + α − A − L + F ν ' −1 ⎥
BR = ⎦ (1)
1+ α
where

ν = No. of neutrons emitted per fission in a fissile nuclide

ν' = No. of neutrons emitted per fission in a fertile nuclide
α = Capture to fission ratio of a fissile nuclide
F = No. of fertile nuclides fissioned per fission of a fissile nuclide
A = No. of neutrons absorbed in SS and sodium per fission in a fissile nuclide
~
L = No. of neutrons leaked from the blanket per fission of a fissile nuclide.

In Eq. 1, numerator gives the excess of neutrons that are available for capture in U-238 to
produce Pu-239.

To derive the expression, the following notations used:

™ Subscripts ‘c’ and ‘f refer capture and fission.

™ ‘fi’ means fissile ; ‘fe’ means fertile and ‘nf means neither fissile nor fertile.
™ Nfi and Nfe are the number densities of fissile and fertile nuclides respectively.

Let 'P' be the total rate of neutrons produced per unit volume. It can be written as
(
P = ν N fi σ ffi + ν ' N feσ ffe φ ) (2)
In order to produce these neutrons, we must loose neutrons at the rate of L1 per unit
volume,

( )
L1 = N fiσ ffi + N fe σ ffe φ (3)
Neutrons are lost by parasitic capture at the rate of L2 per unit volume as

( )
L2 = N nf σ cnf + N fi σ cfi φ (4)
~
If L is the leakage rate from the blanket per fission of a fissile nuclide, total neutron leakage
from the blanket is
( )
~
L3 = L N fi σ ffi φ (5)

So, the net production of neutrons, P, per unit volume is

P = P – L1 – L 2 – L 3
It is to be noted that P neutrons are available for breeding. In other words, P is the rate at
which new fissile nuclides are produced.

3
The destruction rate of fissile fuel per unit volume is
(
N fi σ ffi + σ cfi φ ) (7)

So, the breeding ratio can be written as

(
BR = P / N fi σ ffi + σ cfi φ ) (8)

⎡
( ) ( ) ⎤
~

⎢⎣ν N fi σ fi
f +ν '
N fe σ fe
f − N fi σ fi
f + N σ
fe f
fe
− N σ
nf c
nf
+ N σ
fi c
fi
− L N fiσ ffi ⎥⎦ φ
BR = (9)
(
N fi σ ffi + σ cfi φ )
Dividing numerator and denominator by N fiσ ffi , we get

⎡
( ) ⎤
( )
~

⎢⎣ν − 1 + α − A − L + F ν ' −1 ⎥
BR = ⎦ (10)
1+ α
where

σ cfi N nf σ cnf N feσ ffe νσ ffi

α= ; A= ; F= ; η=
σ ffi N fiσ ffi N fiσ ffi σ ffi + σ cfi

Eq. 10 can be written as

⎡
( ) ⎤
~

⎢⎣ F ν '
− 1 − A − L ⎥⎦
BR = (η −1) + (11)
1+ α

It is to be noted that the nuclear constants, ν ,ν ' , α, A and F depends on the energy
~
spectrum. Leakage factor L depends on both the geometry and the materials present in
core and blanket. The contribution to BR from the fast fission of fertile nuclide is given by
the term F(ν’ -1). Among the fertile nuclides, Pu-240 provides a much greater fast fission
bonus than U-238 due to its low fission threshold and high value of σf . The fast fission
bonus from Th-232 is very small.

A reasonable approximation of Eq. 11 is to ignore the last term. Thus,

BR = η - 1 (12)

™ Of the η neutrons produced, η-1 neutrons are available for breeding.

™ BR > 1 implies that η - 1 should be > 2 or in other terms η > 2.
™ Taking parasitic losses into account (L), η > 2 + L. This is the condition for breeding.

Breeding gain (G) is defined as G = BR-1

FEOC − FBOC FG
G= = ,
FD FD

4
where FG is the fissile material gained per cycle. FBOC is the fissile inventory in the reactor
at the beginning of a cycle (i.e. directly after refuelling) and FEOC at the end of the cycle (i.e.
when the reactor is shut down for refuelling).

Reasons for high BR in fast reactors

BR depends on the core design (fuel type, neutron spectrum, pin size etc.). In fast
reactors, average energy of neutrons is high. All nuclear parameters favour high value of BR
in a fast neutron spectrum as follows:
a) v increases with increasing neutron energy.
b) α for fissile nuclides decreases with increasing energy (Fig.2.1). Pu-239 has low value of
α compared to U-235
c) η increases in fast spectrum (see Fig.2.1).
d) The values of F, the number of fertile atoms fissioned is higher in fast neutron spectrum
(U-238 fission cross section is higher than Th-232).
e) The value of A, the number of neutrons lost by parasitic capture in structure and coolant
per fission of fissile nuclide, is low in fast spectrum.
f) Though neutron leakage from the core will be very high (30-40 % in small cores), the
leakage from the reactor is made very small by surrounding it with a thick blanket.

Energy variation of α Energy variation of η

Nuclide α
Flux Spectrum Pu-239 U-235 U-233
U-235 0.292
LWR 2.04 2.06 2.26
Pu-239 0.248
SFR (oxide) 2.45 2.1 2.31
Pu-241 0.157

Average value of α in fast spectrum Spectrum averaged η

Fig 2.1 :Energy variation of α and η and their average values for fissile isotopes

5
 The following conclusion can also be derived from the above;
™ Although η is slightly greater than 2, breeding may not be possible after accounting
loss from leakage. Hence, thermal breeders are not possible with Pu-239 or U-235.
™ It is possible to construct thermal breeder with U-233.
™ Pu-239 is a better fuel for fast reactors than U-235 because of its breeding potential

2.2 BREEDING POTENTIAL OF VARIOUS FUELS

The maximum breeding possible in fast reactor with oxide, carbide and metal fuels is
given in table 2.1.

Table 2.1 : Breeding Potential of Oxide, Carbide and metal fuelled FBRs
Fuel Breeding Ratio Potential
Oxide 1.15 - 1.30
Carbide 1.30 - 1.40
Metal 1.50 - 1.65

2.3 EXTERNAL VERSUS INTERNAL BREEDING

In the external breeding concept, all fertile material is placed outside the core in the
blanket region. In the internal breeding concept, the core would consist of both fissile and
fertile materials. Most of the current fast reactors have adopted internal breeding concept.
The core with this concept can either be homogeneous or heterogeneous depending on the
desirable breeding. Heterogeneous concept gives high breeding ratio and less sodium void
reactivity but requires very high fissile fuel inventory. The external and internal breeding
concepts are compared in the table given below.

Table 2.2 : Comparison of external and internal breeding

External breeding (hard spectrum) Internal breeding (soft spectrum)

Few fast fissions in 238U Contribution of fast fissions in 238U bigger
Delayed neutron fraction β is small β larger (contribution of 238U)
Small Doppler effect on reactivity Great Doppler effect
Frequent fuel change Long burnup; less frequent reloading
Low rating of fissile High rating of fissile material
Small burnup High burnup
Low total power High total power possible
Blanket thick Blanket thin
Stronger absorbers in core possible Strong absorbers in core are disturbing
Low Pu inventory High Pu inventory
Higher breeding gain Lower breeding gain
High enrichment Low enrichment

2.4 DOUBLING TIME

Another important parameter used in the physics of fast reactors is the doubling time.
It is the time required by a breeder reactor to produce enough fissile material in excess of its

6
own fissile inventory to fuel an identical reactor. In other words, it is the time required to
produce double the initial inventory of fissile fuel. It is an index of merit which is used to
compare different reactor designs, fuels and fuel cycle systems involving several breeder
reactors. It can be defined in many different ways. Though conceptually it is simple, it
involves several aspects in making the definition quite varied from others.

Wyckoff and Greebler attempted to standardize the doubling time concepts (Ref:
Nuclear Technology, 21, 1974, pp. 158-164) as given below.

RDT = Reactor Doubling Time

= time required for a specific reactor to produce enough fissile material in excess of
its own fissile inventory to fuel a new, identical reactor.

SDT = System Doubling Time

= time required for a specific reactor to produce enough fissile material in excess of
its own fissile inventory to supply its requirement of fissile material external to the
reactor and to fuel a new, identical reactor.

CSDT = Compound System Doubling Time

= time required for a system of identical breeder reactors to double the fissile
material in the system, assuming that the number of reactors is increasing at a rate
such that all of the fissile material is being utilized.

Reactor Doubling Time (RDT)

This is the time required for a particular breeder reactor to produce enough fissile
material in excess of its own fissile inventory to fuel an identical reactor. Hence it is the time
necessary to double the initial load of fissile fuel.
If M0 (kg) – Initial fissile inventory

Mg (kg/y) – Fissile material gained during one year

Then RDT = (M0/Mg)

Exm: If Mg = 0.1 M0 and every year this 0.1 M0 is set aside (since only M0 is needed in the
reactor), then after 10 years there would be 2 M0 of fissile material – M0 still in the reactor
and M0 set aside.

Although the equation for RDT is simple, an accurate calculation of Mg is not. If in a

system G = Breeding gain, P = Power (in MW) and f = fraction of time the reactor is at rated
power

Mg = G.(fissile mass destroyed/year)

≅ G.(1+α).(fissile mass fissioned/year)

( )( )(
⎡ P × 10 6 2.93 × 1010 fissions / W.s × 3.15 × 10 7 s / y × f × 239 kg / kg − mol
M g ≅ G.(1 + α) ⎢
)⎤⎥
⎢⎣ 6.02 × 10 26 atoms / kg − mol ⎥⎦
G Pf (1 + α)
≅
2.7

7
 2.7 M 0
Hence, RDT ≅ in years
G Pf (1 + α)

It can be seen that RDT is proportional to the fissile specific inventory (M0/P) and is inversely
proportional to the breeding gain. For example an increase in BR from 1.2 to 1.4 results in a
factor of two reduction in doubling time.

FBOC
We can define RDT in other words as, RDT = where C is the fuel cycles per year.
(FG ).C
Where, FBOC - fissile material in core and blanket at the beginning of the fuel cycle
FG - fissile material gained per cycle
C -number of cycles per year (1/tc – tc in year)

However, doubling time depends on the (i) the relative amount of fissile material in the
reactor and the other parts of the fuel cycle, e.g., storage, reprocessing and fabrication (ii)
the manner in which the fissile material bred in one reactor is used to generate more fissile
material. This gives rise to other definitions of doubling time.

System Doubling Time (SDT)

RDT does not take into account for the processes in the fuel cycle external to the
reactor. These processes are part of the breeder reactor system which are as follows.
(i) time required for external processing also affects the doubling time.
241
(ii) losses – (a) loss of material in fabrication and reprocessing (b) loss of Pu from β-
decay (in U-Pu cycle)

Out-of-reactor fissile inventory is normally in the range of 60 % of the in-reactor

inventory which include;

(i) fuel that is being fabricated

(ii) being shipped to the plant
(iii) in storage at the plant awaiting loading or in cool-down period
(iv) in shipment for reprocessing
(v) being reprocessed

By denoting the fissile inventory in these processes as fissile external (FE), the effect
of this inventory is accounted for through a factor (the ex-reactor factor) defined as

⎛ FBOC + FE ⎞
EF = ⎜ ⎟ ≈ 1.6
⎝ FBOC ⎠
The processing losses FPL, are usually based on processing loss fraction, PLF which is the
fraction of material lost in fabrication and reprocessing operations; this fraction may range
from 1% to a few percent. With refueling fraction RF, being defined as the fraction of core
fuel assemblies replaced at each refueling, the processing losses per cycle are,
1
FPL = (FBOC + FEOC)(RF)(PLF)
2
Where FEOC - fissile material in core and blanket at the end of the fuel cycle

The 241Pu decay external to the reactor during one fuel cycle is

8
 241
Pu ED =
1
2
( 241
) (
Pu BOC+ 241 Pu EOC (EF − 1) 1 − e − λ 4
tc
)
Define all losses external to the reactor per cycle as

FL = FPL+ 241 Pu ED

SDT =
(FBOC)(EF)
(FG − FL)(C)
Compound System Doubling Time

A compound or exponential doubling time assumes that the fissile material produced
in a given cycle is continuously re-invested in an identical breeder. It is meaningful only for a
power reactor network containing a number of breeder reactors with similar characteristics.

dM
= λM M = fissile mass in the system as a function of time (kg)
dt
λ = fractional increase in fissile mass in the system per unit time (y-1)

λ=
(FG − FL)C ≡ ⎛ 1 ⎞
⎜ ⎟
(FBOC)(EF) ⎝ SDT ⎠
M0 = initial fissile mass in the system of reactors
⎛ M ⎞
⎜⎜ ⎟⎟ = e t SDT

⎝ M0 ⎠
[M0 could also be thought of as the number of reactors in the system at time zero × (FBOC)
× EF]

The compound system doubling time is the time required for (M/M0) to become equal to 2.

CSDT = SDT ln2 = 0.693 (SDT)

Summary of a study to estimate the breeding and the doubling time for oxide, carbide, metal
fuelled FBRs with various fuel cycles is given below in Table 2.3.

Table 2.3 : BR and DT for various fuels

233
Cycle Pu-U Pu fueled-Th U/238U-Th 233
U-Th
233
Core Fissile/Fertile LWR Pu/Dep. U LWR Pu/Th U/238U 233
U/Th
Blanket Fertile Dep. U Th Th Th
Oxide RDT (y) 16 29 23 112
BR 1.28 1.20 1.16 1.041
Carbide RDT (y) 9 20 15 91
BR 1.42 1.23 1.23 1.044
Metal RDT (y) 6 12 12 43
BR 1.63 1.38 1.30 1.11
Capacity factor – 75 % assumed

9
2.5 REACTIVITY EFFECTS
In nuclear reactors the variation of power affects the temperature of the fuel,
structural materials and coolant, which in turn affects the reactivity and hence power. This
again leads to change in temperature. Thus there is a feedback loop in the reactor. Since
reactivity is a measure of neutron multiplication and hence determines the power, the
feedback effects of temperature changes are best studied by the reactivity coefficients.
The reasons for reactivity changes in a fast reactor are similar to those in thermal
reactors. Major reactivity additions in the fast reactors are caused by the following.
(i) temperature changes
(ii) control rod ejection
(iii) core compaction
(iv) coolant voiding
Temperature variations in a reactor core will affect core multiplication, both because
of the resulting density changes in core components (due to expansion of fuel, clad, coolant,
absorber rod drive, grid plate and vessel) that change macroscopic cross sections and
because of a change in the thermal motion of core nuclei, which changes microscopic cross
sections (i.e., Doppler effect).
In fast reactors, the neutron energy spectrum includes the resonance regions of both
the fissionable (235U, 233U, 239Pu, 241Pu) and the non-fissionable (232Th, 238U, 240Pu) fuel
isotopes. The Doppler effect in fast reactors is due almost entirely to resonances below
about 25 keV. An increase in fuel temperature will produce an increase in both fission and
absorption cross-sections, and resulting changes can be positive or negative, depending on
the exact composition. Detailed design calculations, using methods benchmarked against
critical experiments, indicate that in large reactors with high fertile to fissile ratio, the Doppler
coefficient is sufficiently negative to provide a prompt shutdown mechanism in the event of
excess fission heating of the fuel.
In a sodium cooled fast reactor, reactivity change that occurs when sodium is voided
is a predominant phenomenon. The reactivity change that occurs when sodium is voided
from fast reactor can be separated into leakage, absorption and spectral components. The
neutron capture component is positive but small in magnitude. Thus the net sodium reactivity
coefficient is dependent on leakage (which is negative) and the spectral shift component
(which is positive). The spectral and capture components are normally largest in the center
of the core, where the neutron flux and importance function are largest and the leakage
component is normally largest in the outer part of the core, where the flux gradient is the
largest. One therefore anticipates that the sign of the sodium reactivity coefficient for a
specific reactor design will change from negative to positive as the reactor size is increased.
Even this is true when the core composition is changed from MOX to metal. However the
challenge of FBR core design is to reduce the sodium void reactivity to zero or negative
value.
Other reactivity effects are due to the expansions of absorber rod drive and vessel.
While the former is a negative reactivity effect the latter gives positive reactivity feedback.
There is no large fission product poison effect (like xenon poison) in a fast reactor, due to
their low absorption cross sections at high neutron energy. Each reactivity effect is described
below.

2.5.1 Temperature Effects

Temperature affects the reactivity as follows :

10
 ™ Nuclear Doppler Effect
™ Expansion (fuel, coolant and structure)

Nuclear Doppler Effect

The atomic vibration of a target nucleus increases when the temperature of the
system is raised. This results in change of relative velocity between target nucleus and the
neutron and the effective interaction cross section of neutron is changed. The peak value of
the cross section reduces and the cross section in the wings are increased (Fig.2.2). But,
the total area under the resonance is conserved. This phenomenon is called the Doppler
broadening of a resonance. The broadening of a resonance leads to higher average flux in
the resonance region (due to energy self-shielding), whereas the flux variation is very less in
the wings. The net result of this effect is that the reaction rate is increased. It is to be noted
that the contribution from the wings only leads to higher reaction rate.

The broadening of capture resonances causes negative reactivity effects in a reactor.

The effect from a fission resonance is to provide the positive reactivity. Doppler feedback is
prompt and it is one of the inherent safety
parameters. The magnitude of Doppler
feedback in a reactor depends upon the
following :
™ Temperature change
™ Atom densities of fissile and fertile
nuclides.
™ Neutron spectrum (Depends on fuel
type and core size)

The net Doppler feedback from the

fissile nuclides (Pu-239, U-235) is a balance
between the contributions from capture and
fission resonances. The net effect from Pu- Fig.2.2 : Doppler Broadening of a Resonance
239 is found to be negligible (slightly positive).
Fertile nuclides (U-238, Pu-240) do not have fission resonances but they have capture
resonances. Main contribution of negative Doppler feedback in a reactor is from the fertile
nuclides. Fast reactor core can be designed such that Doppler feedback can counter act the
positive sodium void effect. Following are the important observations :

™ For pure fissile materials, net Doppler feed back could be slightly positive
™ For large core with MOX fuel, Doppler feedback can be more negative.
™ For small reactors, Doppler feedback is negligible due to hard spectrum, but it will be
significant for higher size reactors.

Doppler Coefficient

It is the change in reactivity due to change in

temperature and it has following dependence for oxide
fuel :
dk K D
=
dT T

Where KD is the Doppler constant and T is the

temperature in Kelvin. KD is equal to Fig.2.3 : Change in Keff with temp

11
 dk
K D =T
dT

The reactivity loss for a temperature change from T1 to T2 K is (Fig 2.3)

T2
dT
ρD = KD ∫
T1
T

⎛ T2 ⎞
= K D ln ⎜⎜ ⎟⎟
⎝ T1 ⎠

For various reactor spectrums, dk / dT has the following dependence as evidenced by using
the resonance theory:

dk 1
Metal fueled reactor (hard spectrum ) : ∝ 3/ 2
dT T

⎛ dk ⎞ 1
Oxide fuelled FBRs (medium spectrum) : ⎜ ⎟∝
⎝ dT ⎠ T

dk 1
LWR (softer spectrum) : ∝ 1/ 2
dT T

It shows that the spectrum of the oxide fuelled fast reactor lies in between the hard
spectrum and the soft spectrum. A prompt negative feedback is particularly important for fast
reactors because of two mechanisms, fuel compaction and sodium loss, which have the
potential to make the reactor super prompt critical. The Doppler effect provides this prompt
negative feedback with ceramic fuel. It is important to note that, feedback response is
prompt only when the fissile and fertile materials are mixed intimately together. This is
warranted because most of the fission energy comes from the kinetic energy of the fission
products and hence fertile isotopes, if they are very close to fissile isotopes, cause loss in
the kinetic energy which in turn leads to increase in temperature and thermal motion of the
fertile nuclei thereby causing large Doppler effect.

Further, Doppler effect comes predominantly from the capture of low energy
neutrons. Energy self-shielding is required for an increase in the temperature of the fertile
material to increase the effective cross-section of that material. At high energies there is little
self-shielding. Ceramic-fueled reactors, due to the presence of oxygen or carbon in the fuel,
have a soft enough neutron spectrum to have a large Doppler effect.

Dimensional Changes Effects

Core might undergo dimensional changes either during normal operation or during
transients. When temperature increases, fuel will expand axially. The structures supporting
the core will expand radially as the temperature increases. During fuel melting in a severe
accident, fuel might slump down causing core compaction or it may be expelled causing fuel
expansion. All these changes would induce reactivity changes which have to be understood.

Detailed analysis taking into all factors including leakage etc show that, when core is
getting compacted, there will be an increase in the reactivity. Similarly, when the fuel
temperature increases, the fuel expands axially. The radial fuel expansion will have little
impact on the core radial expansion, as it is primarily governed by the structures of the core

12
restraint system. Hence, the primary core expansion resulting from temperature increase
would be in the axial direction. The analysis has shown that axial leakage does not change
much with fuel axial expansion, radial leakage increases thereby producing a negative
reactivity effect. The role of axial expansion of fuel in the normal solid fuel pin geometry is to
provide a prompt negative reactivity feedback at the start of a power transient. This
mechanism is the principal prompt negative feedback available in a metal fuelled fast
reactor.

In a metal fuelled fast reactor where the spectrum is hard, the axial expansion
provide the much wanted negative reactivity feedback. Whereas, in the oxide fuelled fast
reactors, the axial expansion effect is very less due to the fact that the fuel cracks and due to
lack of structural integrity, axial expansion mechanism is somewhat absent or unreliable and
hence importance is given to Doppler effect in oxide fuelled reactors in order to get prompt
negative feedback of reactivity.

2.5.2 Isothermal Temperature Coefficient

Δρ
Isothermal temperature coefficient of reactivity, ρ T = , considers change in
ΔT
reactivity when the core temperature is raised or lowered uniformly (T is same throughout
the core). It has contributions from (i) Doppler effects and (ii) Expansion effects

It may be noted that for small reactors, there will be negligible contribution from
Doppler and major contribution will be from expansion effects. It is also to be noted that
there will be less Doppler feedback for metal fuelled reactors (spectrum hard) but large
negative contribution from expansion effects.

2.5.3 Power Coefficient of Reactivity

Power coefficient is defined as the change in reactivity due to power change.

Δρ
ρP =
ΔP
This coefficient is defined to take into account the non-uniform temperature rise
(
across the core due to power generation. Power generation results in a ΔT = Toutlet − Tinlet )
across the core. Fuel temperature (Tf) also increases due to power change.

2.5.4 Sodium Void Reactivity Effect

Sodium voiding due to its boiling or loss will result in large change of reactivity due to
the following effects:

™ Spectral hardening
™ Less capture
™ Increased leakage

Spectral Hardening: Because of the reduced elastic

and in-elastic scattering by sodium, the neutron
spectrum becomes harder (Fig.2.4)

This effect gives always positive reactivity due to

Fig.2.4 : Hardening of neutron spectrum
™ An increase in fission in fertile nuclides like Th- due to sodium voiding
232, U-238 and Pu-240.

13
 ™ Decrease in capture to fission rate for fissile nuclides (Pu-239).
™ Decrease in capture in fertile and structural nuclides.

Less Capture: Reduction of capture in sodium will result in a positive reactivity. This is
generally a small effect since sodium capture cross section is very small.

Increased Leakage: With sodium removal from the core,

the leakage probability of neutrons increases and the
reactivity thus decreases. Major contribution to this
effect will come from the sodium near the core boundary.
Leakage effect can be increased by

™ Increasing the surface-to-volume ratio of the core.

This can be achieved by choosing a core with a
smaller core height to diameter ratio (H/D), a so-
called ‘pancake’ core.
™ Heterogeneous core

It is to be noted that the magnitude of leakage

effect decreases as the core size increases. Sodium
loss reactivity effect is space dependent (Fig.2.5).
Sodium loss from the centre of the core yields a highly Fig.2.5 : Spatial dependence of
positive reactivity effect, whereas its loss at the edge various effects due to sodium voiding
gives negative effect.

The net effect of all three effects will be negative if leakage effect predominates; for
example in small reactors (like in FBTR). Large cores like PFBR has positive sodium void
coefficient. For PFBR, complete voiding of sodium results in a reactivity change of +620
pcm. However, during normal operation of a reactor, reactivity due to sodium density
changes is very small.

2.5.5 Long term Reactivity Loss from Burnup

An operating reactor loses reactivity due to the burnup of fuel and accumulation of
fission products. In fast reactors, fission products do not behave as neutron poisons to the
extent as they do in thermal reactors. The major loss in reactivity comes from the loss of
fissile elements. This is partly compensated by the internal breeding which takes place in
the core. There is a small reactivity gain due to the build-up of fissile material in the
blankets.

2.5.6 Reactivity Effects Due to Core Compaction

It is to be noted that thermal reactors are designed in a most reactive configuration.

It means that any disturbance to the core configuration lead to negative reactivity. It is not
the case for fast reactors. Core compaction leads to positive reactivity. During core
compaction, core size reduces and fuel density increases. Net effect is to reduce the
leakage rate and hence the core reactivity increases.

2.5.7 Fuel Expansion and Bowing Effects

Due to radial power distribution across the core, the parts of fuel element closer to
the reactor centre will be slightly hotter than the parts away from the centre. If the fuel
elements are cantilevered at the bottom (top free), the differential expansion of the fuel
element will cause a net movement of fuel outward from the core centre (bowing) as the

14
power increases giving a negative reactivity. In other words, cantilevered from bottom of a
fuel sub-assembly causes flowering of core giving negative reactivity. In case of contact at
two pad levels, the fuel element sometimes may move inwards, giving positive reactivity. In
practise, however, the fuel elements cannot move freely very far before they touch each
other, and the reactor may be designed so that they cannot move freely at all. For this
reason, the effect of bowing on reactivity is likely to be very small.

2.6 REACTOR CONTROL

The power produced in a nuclear reactor core is proportional to the rate of fission
which is proportional to the product of the fissile atom concentration and the neutron
population, N, in the core. The neutron population level arises from a dynamic equilibrium
between the generation of neutrons by the fission of fuel atoms and the loss of neutrons by
absorption in fuel and other atoms and by leakage from the core.

A fundamental parameter is the multiplication factor, k, which is the ratio of the rate of
generation of neutrons to the rate of loss of neutrons. The fission rate and reactor power will
hence continuously increase, decrease or be constant accordingly as k is greater than, less
than or equal to unity and the reactor is correspondingly said to be super critical, subcritical
or critical. In an operating reactor, k can be changed with limits by movement of the control
rods which changes the rate of absorption of neutrons. An alternate parameter to k is the
reactivity ρ defined as ρ = (k -1)/k. Study of the normal and abnormal changes to the core
reactivity is important for control and safety.

Prompt neutron lifetime and delayed neutron fraction

The prompt neutron lifetime, l is the average time from the `birth’ of a neutron in the
fission of a fuel atom to the `death’ of the neutron by absorption or leakage. The life time in a
thermal reactor consists of two parts; (i) slowing down time (ii) thermal neutron life time or
diffusion time. As the slowing down time is much less than the diffusion time (slowing down
time vary depending on the moderator material used), the prompt neutron life time can be
equated to diffusion time. In a fast reactor, l is of the order of 10-7 sec as compared to of the
order of 10-4 to 10-3 sec for thermal reactors. In this connection, it must be noted that over
95% of the reactions in a fast reactor are scattering reactions which reduce the neutron
energy and prolong the neutron life. The average rate of loss of neutrons in N/l and the rate
of production is kN/l. Thus the growth of neutron population, for constant k, will be
exponential with a time constant T = l/(k-1) called the reactor period.

Normal control movements make (k-1) of the order of 10-4 to 10-3 so that the period
would be of the order of 10-4 to 10-3 for fast reactors and of the order of a second for thermal
reactors. These are quite rapid response times and would have made reactor control
problematic, especially in the case of fast reactors. However, as explained below, the
presence of delayed neutrons increases the response time and makes it the same for both
fast or thermal reactors.

A small fraction β (delayed fraction) of the neutrons produced in the fission of a fuel
atom are, in fact, emitted subsequent to the fission by decay of certain radioactive fission
products (there are six delayed neutron precursors) with mean lives ranging from 0.3 sec to
80 sec. β is 0.00204 and the average mean life τ is 14.6 sec for Pu239 and the values differ
for different nuclides. When delayed neutrons are included, the reactor period can be
obtained, when k is near unity, by replacing the prompt neutron lifetime l by a weighted
mean neutron lifetime l* = (1 - β )l + βτ . The term l* is of the order of 0.05 sec and hence
for (k-1) of the order of 10-4 to 10-3 sec, a period of the order of 50 to 500 sec will result. On
the other hand, if delayed neutrons are not included in the calculation, the reactor period

15
would be about 0.45 sec which is very much less than the value obtained (14.6 sec) with
delayed neutrons. The reason is that because of the delay in the release of a fraction of the
neutrons, the average effective generation time is greater than the prompt neutron
generation time by certain quantity, which depends on the extent to which the delayed
neutrons contribute to the fission chain
reaction.
When ρ is greater than 0 by an
amount β , then the neutron population
grows at a rate governed by the extremely
short prompt neutron lifetime l and not by
the weighted mean neutron lifetime l*. The
reactor under such conditions is said to be
super prompt critical and very rapid rates of
power increase become possible. An
important safety consideration is to protect
against the possibility of a reactor becoming
super prompt critical. On account of the
importance of the delayed neutron fraction
on the kinetic behaviour of the chain
reaction, ρ is often measured in terms of β.
An amount of reactivity equal to β is called
a `dollar of reactivity’ which makes the
reactor super prompt critical. All the above
hypothesis hold good only if the reactor
system is not far from critical and reactivity
ρ < β.

A plot of reactor period vs reactivity

is shown in Fig 2.6. It can be observed that,
for a reactivity of less than about $0.9, the
reactor period is unaffected by neutron Fig.2.6 : Reactor period variation with reactivity ($)
lifetime i.e fast and thermal reactors behave for Pu-239 fuelled reactor
identically. Detailed reactor kinetics is
covered in other lectures and beyond the scope of this lecture.

2.7 OTHER ASPECTS OF PHYSICS DESIGN

The other aspects for completing the physics design are fuel enrichment and fuel SA
worth, absorber worth and shutdown margin. These are only briefly described and detailed
description is beyond the scope of this lecture.

The fuel enrichment defined by the volume fraction of the fissile in the fuel should be
sufficient to ensure a critical core during an entire cycle of operation. At the end of the cycle,
it should have a minimum reactivity margin normally around a few hundred pcm. It is
required to overcome the reactivity losses during the operational cycle accounting for
temperature and power effects and burnup reactivity loss including uncertainties. Fuel SA
worth is determined by the safety criteria on the Keff adopted that individual SA when
immersed in water should satisfy. Similarly absorber worth is also decided by the number of
shutdown systems provided, various safety criteria to be satisfied such as (n-1) rods to be
considered for a system and worst possible core configuration etc. and minimum shutdown
margin that is to be provided. Accordingly, the absorber rods position and the number are
determined by means of detailed calculations.

16
2.8 SHIELDING

Purpose and principle

The main purpose of the shield systems is to
(i) protect operating personnel from radiations and
(ii) reduce the radiation exposure of structural components.
Radiations entering the shield from the reactor can produce internal heating and
damage to shield materials. Shielding should not only take care of primary radiations
(neutrons and gamma rays from the core) but also secondary radiations produced outside
the core as a result of interaction of the primary radiations, mainly neutron with nuclei of the
fertile, coolant and structural materials.
Therefore the shielding of nuclear reactor systems is concerned with the study of
radiation distribution in the shield towards reducing the escaping radiation to acceptable
levels and of limiting the adverse physical effects on the shield materials.

Distribution of energy in a single fission is given below.

KE of fission products ≈ 168 MeV

Instantaneous γ energy 7 MeV
KE of fission neutron 5 MeV
β particles from fission products 7 MeV
γ from fission products 6 MeV
Neutrinos 10 MeV
TOTAL 200 MeV

In principle the radiations, which might escape from a reactor, include alpha and beta
particles, gamma rays, neutrons of various energies, fission fragments and even protons
resulting from (n,p) reactions. In practice, however, only gamma rays and neutrons need to
be considered as they are the most penetrating. Any material, which attenuates these
radiations to sufficient extent, will automatically reduce all the other to negligible proportions.
For purposes of shield design, the neutrons and gamma rays are considered from the stand
point of view of their place of origin. The primary radiations are defined as those, which
originate in the core; the secondary radiations are produced outside the core as a result of
the interaction of primary radiations (mainly neutrons) with nuclei in structural, coolant and
shield materials.

Sources and problems of radiation in a fast reactor

a) Cores are small, power densities higher

This implies that leakage of fast neutron into the shield is high. Thermal reactor
shields do not have to handle so many fast neutrons. These fast neutrons leaking out
have energies greater that 10-15 KeV on an average. Neutrons with energies greater
than 3MeV are also quite considerable. Therefore neutron shields should contain a
large amount of moderator material to slow down the neutrons before capture. The
capture rates are much higher at lower energies
b) Sodium Activation

Na23 (n,γ) → Na24 ⎯15

⎯→ h
Mg24 ; 1.38 MeV γ and 2.76 MeV γ

⎯ Ne22 ; 1.28 MeV γ and 0.44 MeV γ

Na23 (n,2n) → Na22 ⎯⎯→
2.6 y

17
 Shielding against these activation gammas is one of the important problems in fast
reactors.
c) Corrosion products

The constituent elements of stainless steel get activated in the reactor

280 d
⎯→ Mn54 0.84 MeV γ
Fe54 (n,p) ⎯⎯

Ni58 (n,p) ⎯71

⎯→d
Co58 0.81 MeV γ
Fe58 (n, γ) ⎯ 4⎯ → Fe 1.3 MeV γ
59
⎯
.9 d

Co59 (n, γ) ⎯ 5⎯ ⎯ → Co 1.33 & 1.17 MeV γ

. 27 y 60

For maintenance of primary pumps and IHXs , these radiations pose problems.
d) Argon gas activity

Argon gas is used as inert gas blanket over sodium in the reactor to maintain inert
gas atmosphere. This also get activated
Ar40 → Ar41 ⎯1⎯
.8 h
→ 1.29 MeV γ
e) Clad rupture
If clad rupture happen, fission gases (isotopes Kr and Xe which are radioactive)
escape into cover gas. Because have and (e) adequate shielding has to be provided
for cover gas pipes.

Point Kernel Methods

In the point kernel method, the fundamental assumption is that the extended sources
of radiation can be regarded as consisting of differential isotropic point sources and that the
effect of the radiation from the whole source at the point of interest can be obtained by the
summation of the individual sources that comprise the entire source region. In most practical
shielding applications the assumption that the gamma rays are emitted isotropically is
physically valid when considering effects on macroscopic scale. The effect of radiation at a
particular point is usually identified with the response of an idealized detector situated there.
The point kernel technique is then applied to various important types of extended source,
disc, source and volume source to obtain doses with shield material present in between
source and the point at which the dose estimation is made. In most practical cases the dose
at the point of interest consists of contributions not only from uncollided radiation but also
from secondary radiation such as scattered radiation. Contribution of secondary radiation is
taken care through the concept of build up factor, which will be discussed later.

Build up factor

In practical cases, the dose at the point of interest consists of contributions not only
from uncollided radiation but also from scattered radiation in shields. Contributions from both
the scattered and the uncollided radiations are taken care of through build up factors which
E0

total dose ∫R C ( E )Φ ( E )dE

is defined as, B = = 0

dose due to unscattered radiation RC ( E0 )Φ ( E0 )

18
where E0 is the incident gamma ray energy, RC (E) is the dose conversion coefficient and
photon flux φ(E) at the position P where dose rate is being calculated.

Build up factors have the following important properties:

(i) Depends on the distance of penetration through the attenuating medium in

terms of mean free paths μt.
(ii) The type of medium, expressed in terms of the atomic charge number Z of
material.
(iii) The energy of the source photons.

Neutron attenuation

Neutron shielding analysis is complex. It involves not only attenuation of primary

(source) but also production and attenuation of secondary particles such as photons.
Neutron shielding analysis might also require consideration of energy deposition in the
shielding material, and the resulting thermal and radiation damage effects. Neutron
attenuation calculations are usually performed using large computers with extensive cross
section libraries using transport calculations codes. The use of neutron build up factors has
not been practical because

1) For fast neutrons, the probability of scattering is much higher than absorption in
most materials. As a result neutron build up factor can assume very large values
2) Rapid variation of cross sections with energy makes the build up factor too
sensitive to initial energy.
3) Angular distribution of the scattered neutrons may also very rapidly with neutron
energy.

Therefore the correction for the uncollided flux densities for scattered fast neutrons is
generally difficult. An approximate technique is to replace the actual total cross section by an
effective removal cross section in the uncollided flux density description. This concept is
found useful (particularly) when relatively thin slabs of shield materials are immersed in an
infinite hydrogeneous medium such as water.

Shielding materials used

For slowing down as well as thermal shields (these are shields introduced to protect reactor
vessel from excessive radiation damage): - stainless steel:- very good inelastic scatterer,
can bring down the energy of neutron even by 1 MeV
Slow neutron capture: - compounds of Boron (boron is a good absorber) and carbon (which
moderates) special mention may be made of B4C and Borated Graphite
Sodium itself moderates if neutrons are made to go large distances through Na (as in the
pool type of reactors)
Gamma attenuation: - Any high Z material with high density. Lead is the best example.

Shielding arrangement within the reactor vessel

Radial shields: Alternating layers of various thickness of shielding materials SS, B4C

19
Axial shields: Alternating layers of SS and B4C to reduce embrittlement of structural
materials.

Typical Shield Design Criteria

A typical set of criteria for shield design is given below. First is the basic set of
criteria. Second set is the criteria which might vary from reactor to reactor. However, for pool
reactors, most of the criteria would apply.
(1) The occupational dose limit is 20 mSv per year average over a period of 5 years
and 30 mSv in any calender year.
(2) All exposure shall be kept below 1μSv/h in all fully occupied regions such as
control room.

Other typical shielding design criteria that would be adopted other than ICRP limits
for the operating personnel and the site are as follows:
(i) dpa and helium production at the grid plate and the core cover plate
(ii) top shield placed above the pool of sodium
(iii) flasks handling intermediate heat exchangers and primary sodium pump and all other
flasks used for irradiated material transport
(iv) dose from secondary sodium pumps, dose on contact for components in sodium
when taken out for maintenance.

A typical core layout of a pool reactor consisting of shielding SA in the periphery and
the axial shield arrangement provided in the SA are shown in Fig 2.7 and 2.8. The radiation
streaming paths in a pool type reactor is also shown in fig 2.9.

References:

1. Fast Breeder Reactors, Alan E. Waltar & Albert B. Reynolds, Pergamon Press 1982
2. Nuclear Reactor Engineering, Samuel Glasstone & Alexander Sesonske
3. Internal lecture notes of IGCAR, Kalpakkam.

20
 Fig 2.7: Core layout showing shielding SA

Fig 2.8 : Fuel SA with axial shield

Fig 2.9 : Typical pool type reactor assembly
– Radiation streaming paths in top shield

21