OBSOLETE VERSION

These instructions refer to the

labs formerly performed in
Studsvik

L ABORATORY EXERCISE IN REACTOR PHYSICS (IV)

CRITICALITY EXPERIMENT
Instructor: Daniel Westln
08-5537 8792, 070-631 54 08
daniel@neutron.kth.se
INTROD UC TI ON

Reactivity monitoring is required to confirm the safety of certain operations with reactor fuel, such as
reactor loading and approach to criticality. The object of this laboratory exercise is to demonstrate
neutron response measurements used in reactor operations for safely approaching criticality. Neutron
response measurements during reactivity buildup serve as a safety guide. In the laboratory exercise,
control rods are calibrated and the shutdown reactivity worth is determined. The half-life of different
groups of delayed neutron precursors is measured.

N E UTRON M ULTI PLI CAT I ON

The most simply measured reactivity index for monitoring is neutron multiplication or a quantity
roughly proportional to it. It is appropriate to consider the meaning of idealized neutron multiplication.
Let us imagine the introduction of source neutrons into a subcritical reactor. Assuming S source neutrons,
there will be S k neutrons after the first fission generation, S k 2 neutrons after the second generation,
and so on, so that the total number of neutrons per source neutron after many generations approaches a
geometrical series:
(1)

M =

(1 + k + k

+ S

1
1 k

Here, M is called the source multiplication factor. It describes the ratio of the fission source to the
independent source and gives the total number of neutrons appearing in the fissionable material per
source neutron. k is an idealized multiplication constant that describes the multiplication of a source
neutron (this introduces some limitations that will be discussed in more detail below). Please note that
since in the expression above is valid for a subcritical system, k is smaller than one. It is seen that the
reciprocal source multiplication factor, 1 M , decreases as 1-k:
(2)

1
= 1 k
M

This relationship is often employed in bringing a reactor system to criticality, to obtain information on
the degree of subcriticality and to extrapolate to the critical condition. If, for example, the operator starts

pulling control rods, the multiplication constant and thus the source multiplication factor M increases
with increasing rod withdrawal. When the reactor reaches critical condition, the source multiplication
factor becomes infinite. This approach to criticality is monitored by placing a neutron source in the
reactor and measuring the neutron flux for various rod positions. The neutron counting rate is
proportional to M. A plot of the inverse counting rate or 1/M as a function of rod position becomes
equal to zero at the critical condition. If we extrapolate a curve to the zero value then we can predetermine the control rod position at which we reach the critical condition. The curvature is important
for extrapolations to the critical value based on part of the 1/M curve.
In the discussion, it is assumed that the reactor is reasonably close to critical and the neutron population
is in equilibrium. If the reactor is far below critical then eq. (1) is not valid. Eq. (1) only applies to the
multiplication of neutrons in the artificial eigenstate. Eigenstate in this sense, refers to neutrons such
that their energy spectrum and spatial distribution are characteristic of the reactor, i.e. asymptotic
spectrum and so-called fundamental-mode distribution. Source neutrons, in general, do not have these
properties and multiply differently in successive generations compared to eigenstate neutrons. In that
case, we cannot formulate a simple geometric series, based on a single k-value. In close-to-critical
conditions, this effect has negligible influence since source neutrons are small in number in comparison
with the total neutron population. The neutron flux assumes the eigenstate at criticality.
Although measurement of source multiplication is a static technique, there are time-dependent effects
that can be important in the interpretation of results. The asymptotic, true, multiplication following a
reactivity change (e.g. change in control rod position) in a subcritical system is observed only after all
delayed neutrons have attained equilibrium concentration. During a sequence of stepwise addition of
reactivity, as in following 1/M approach to criticality, total asymptotic multiplication will clearly be
observed at each step if one waits sufficiently long after each reactivity addition. Nevertheless, if
reactivity is added too rapidly either stepwise or continuously the observed multiplication will lag the
asymptotic multiplication thus giving an underestimate of the instantaneous multiplication of the system.
The physical reason for this transient effect is clear. The equilibrium condition is identified by a selfsustained constant power level only when all delayed neutron precursors are in equilibrium, meaning that
the formation rate of precursors equals the decay rate.
Under very weak source conditions, the neutron level during the early stages of startup is so low that
statistical fluctuations tend to dominate and the kinetic equations are not applicable. At later stages, the
power level becomes sufficiently high that statistical fluctuations are no longer important. In general,
distorting effects become relatively small near criticality and extrapolated critical points become
progressively more reliable as the critical condition is approached.

R2 -0 R E ACTOR

The laboratory exercise is conducted at the light-water moderated R2-0 experimental reactor. Some
important design parameters are:

The core is submerged in a pool with 9 meters of water.

Core support structure is suspended at a depth of 8 m, with room for 100 fuel assemblies in a
quadratic geometry, see figure 1

Five control rods (boron carbide) are used to control reactivity.

Alpha-beryllium neutron source is located in the core.

Two low power neutron detectors (fission counters)

38 fuel assemblies containing 200-250 g of U-235, see figure 2

Neutron detectors surround the core and signals from these are converted to thermal power, period, etc.
which are displayed on instrumentation on the control board. These signals are also used by the core
protection system to signal automatic shutdown.

Figure 1. A) Sketch of the core B) Load plan for R2-0 (The core was redesigned some time
ago. The present core has five control rods as in B.)
PROC ED UR ES

Approach to criticality
The object of this laboratory exercise is to demonstrate neutron multiplication measurements as used for
safely approaching criticality. The basic requirements for this type of measurement are a neutron source,
either inherent in the fuel or added to it, and a neutron detector with an adequate response to neutrons
from fission. If these requirements are satisfied, the detector response rate under equilibrium conditions
is proportional to neutron multiplication, and its reciprocal approaches zero as the fuel assembly is made
to approach the critical state. The criticality approach curve (e.g. reciprocal response rate vs. the control
rod position) is then sufficiently predictable for extrapolation to criticality. It is possible to predetermine
the control-rod position at which the reactor becomes critical. It is desirable to monitor the approach,
and analyze the data, to ensure that the approach is orderly and that one is far from or close to critical.
At the start of the exercise the control rods are fully inserted into the core. Read the water temperature
in the pool. The control rods are then withdrawn in successive steps. Following each step the neutron
count rate is registered. Experimentally the neutron count rate is measured by two neutron detectors
(fission counters) located at different positions in the reactor.
The neutron level following each addition of reactivity should be allowed to stabilize or settle out in
order to obtain an accurate indication of asymptotic multiplication before proceeding with the next
reactivity addition. This settling out time becomes progressively larger as criticality is approached.
The reciprocal count rate vs. axial control rod position is plotted graphically. Successive extrapolations
are carried out in order to get an updated estimate of criticality. Note that the extrapolation to the critical
dimension must use the shape of the 1/M dependence. Tentative extrapolations of the curve to
criticality (zero reciprocal rate) give conservative estimates provided a nose-diving shape (negative
curvature) can be avoided. In such cases, the shape of the reciprocal count curve must be anticipated in
order to choose safe steps of control rod withdrawal. A proven technique is to combine data from both
detectors and use the curve that tails out while the other dives.
As we have observed, the curve of reciprocal neutron-count rate is linearly proportional to the
multiplication constant.

1
1

1 k
n M
n = neutron count rate
In practice, however, the desirable linear approach curve is frequently difficult to attain, for example, due
to the variation of control rod effectiveness with position, rod shadowing etc. Source and detector

location may also distort the linear approach curve. The reciprocal neutron count rate is linearly
proportional to the control rod position only if control rod effectiveness (reactivity worth) is constant
with respect to position.

Z ss only if
= constant
n
Z ss
Z SS = control rod position
Control rod calibration
Reactivity calibrations are performed in the neighborhood of criticality. The most common technique is
by period measurement in the slightly super critical region. Approximate reactivity calibration of one of
the boron carbide rods (S1-S5) is obtained by:
The reactor is operating at constant power of approximately 50 W. Power should be kept low to avoid
thermal feedback, but high enough to ensure stable reactor operation. Control rods that are not to be
calibrated are positioned at equal axial height.
1) The control rod to be calibrated is withdrawn 50 units, or so. Aim at a doubling time in between
40 50 seconds. The reactivity increase is obtained by measuring the doubling time and using
the reactivity vs. doubling time curve in figure 2.
2) The other control rods are adjusted to resume stationary power at 50 W.
3) Tasks 1 and 2 are repeated for the full control rod length.
4) The differential reactivity worth,

, is plotted. The integrated control rod reactivity

Z ss Z ss

worth should also be calculated and plotted.

Doubling time as function of reactivity

90

Doubling time (s)

80
70
60
50
40
30
20
10

60

80

100

120

140

160

180

Reactivity (pcm)

Figure 2.

Doubling time vs. reactivity

Power increase
When control rod calibration is done, the power is raised to approximately 50 kW. The reactor control is
switched to automatic operation. Observe the Cherenkov radiation. Register control rod position and
the water temperature.

Determine delayed neutron precursor half-lives

Reactor shutdown is carried out at the end of the exercise. All control rods are dropped into the reactor.
Immediately following shutdown, the neutron count rate is registered from one of the detectors.
Measurements of the neutron count rate, using short intervals, last for approximately 4 minutes. The
count rate is plotted on a linear-logarithmic scale as a function of time. Using the plot, the half-life of
groups of delayed neutron precursors is determined (try to find as many groups as possible). Data is
stored in the control room computer. Bring a 3.5 inch floppy to retrieve the data.

Determination of the shutdown margin

The shutdown margin is defined by the margin to criticality under the condition that the most reactive
control rod remains in the top position. According to safety regulations for the R2-0 reactor, the
shutdown margin must be at least 1.5%. Besides the reactivity worth of the control rods, the shutdown
margin depends on poisoning and the water temperature, according to eq. (3).
actual
i

M =
dz
dT
df

T
f
i inside z
cold
non poisoned
poisoned

outside

(3)

i = index for all control rods except the most reactive

f = poisoning

= reactivity
T = temperature

i
is obtained from the control rod calibration. Assume that all control rods have the same reactivity
z
worth. It can be assumed that the water temperature does not fall below 20 C. The water temperature
coefficient has been measured to 12 pcm per degree C. Assume that the reactor is not poisoned.

Determine the excess reactivity

All reactors have a certain amount of built-in excess reactivity, to compensate for burnup and poisoning.
We may estimate this excess reactivity for the R2-0 reactor using eq. (4):
(4)

R=
i

i
dz
z
criticality
top

i = index for all control rods

The integral is calculated using measured data on control rod reactivity worth. The calculated value is an
indication of the burnup potential.

R E ACTOR SA F ET Y

Reactivity insertion
Experiments conducted in the U.S SPERT reactors have demonstrated that a sudden reactivity increase
of 2% in an initially critical reactor results in a strong power excursion and partial core disruption. Even
a seemingly small reactivity insertion of 1% results in a serious power increase. The total reactivity worth
of all control rods in R2-0 is at most 2.4 %. The control rods can be withdrawn at a rate of 1 mm/s,
which means that it takes approximately 10 minutes to raise the rods from bottom to top position. It can
be shown that the automatic shutdown system is sufficient to protect against reactivity insertion rates of
this kind (5 pcm per second).

In the course of measurements of approach to criticality and reactivity calibration, there is ample
opportunity for successively improved checking of the proper functioning of instrumentation and the
stability of the reactor. A series of consistency checks during buildup of reactivity is good nuclear safety
practice. Anomalies that appear at any given step should be understood.

Important
It is VERY important to pay attention to safety instructions. Even if the reactor is at low power, there is
a possibility that automatic shutdown is activated in case too much reactivity is inserted. This should be
avoided since it requires a long startup procedure. In manual operation, it is important to observe
reactivity changes. Safety considerations impose some limits:

The doubling time1 must not fall below 22 seconds

Reactivity insertion must not exceed 160 pcm

A doubling-time meter is of considerable value as an operator aid; the doubling-time scram-trip, is set
for a rather short period to avoid spurious scrams. Observe the doubling time during control rod
withdrawal.

PR EPAR ATORY EXERCI S ES

What time-dependent parameters influence the state of a critical reactor?

What is Cherenkov radiation and how does it arise?

What is the reactor period?

Why do delayed neutrons occur and what fraction of the total neutron population do they
amount?

How can we see the effects of thermal feedbacks in the core?

R EPORT

The laboratory exercise should be documented in a complete report in English or in Swedish. The
required contents are the following:

Abstract The abstract should summarize your results and methods. It should be short
but precise.

Introduction Give an introduction to the experiment making it possible for a fellow

student who has not performed this laboratory exercise to understand the rest of the
report and to relate the result to what she already knew. The last part of the
introduction should be written like as a brief introduction to the following parts of the
report.

Theoretical background Describe the theory behind the experiment. Present the
mathematical framework.

1 At Studsvik, the expression doubling time is commonly used when discussing changes in the reactor power. The expression
period is commonly used though. The doubling time is the time needed to double reactor power. The period is the time
needed to increase power a factor e.

Description of the experimental set-up Explain how the experiment is set up. Go into
details where it is necessary, avoid it where it is not.

Description of the experimental procedure and performed calculations First of all

measured data from the experiment should be presented. Give the raw data as it is
before you perform any calculations. As an example the time for a messurement should
be given as 2003-12-04 16.34.50 and 2003-12-04 16.35.10, not as 20 s. Then perform
the calculations presenting intermediate results where necessary. Here the 20 s would be
presented. Use the layout to present your data and calculations nice and clear. Use
tables.

Results Present your results.

Conclusions and discussion Conclude and discuss your results, why are they
important? Are they correct?

Preparatory exercises Present the solutions of your preparatory exercises, preferably

as an appendix to your report.