-2- TDAI -244

2.0 PHYSICAL DESCRIPTION OF THE CANDU-600

2.1 The Pressure Tube COnCS-Dt

A CANDU reactor is a heavy water moderated, natural

uranium fueiled reactor utilizing the pressure tube concept. This
consists of an array of pressure tubes, containing the reactor fuel and
coolant, passing through a large horizontal cylindrical vessel (the
calandria) containing the heavy water moderator and reflector. The
overall arrangement is shown in Figure 2.1.1. Pressurized heavy water
coolant is pumped through the pressure :ubes, cooling fuel and conveying
heat from the fuel to the outlet header and from there to the steam
generators. Each pressure tube is isolated and insulated from the heavy
water moderator by a concentric calandria tube (see Figure 2.1.2). The
annular space between the pressure and calandria tubes is filled with
gas. This configuration results in the moderator system being operated
independently of the high temperature, high pressure coolant in the
pressure tube. The heat generation in .the moderator is very low thus
obviating the need for a high-strength pressure vessel.

Due to the physical separation of coolant and moderator

the latter operates at the relatively cool iemperature of approximately
7o”c. This means that the cool moderator can act as the heat sink under
certain accident condi:ions. Also, it means that the reactivity and
control devices which are positioned interstitially between the pressure
tubes operate in a low pressure low temperature environment.

Experimental evidence indicates tha: pressure tubes will

leak before they break, since their thickness is less than the critical
crack length. Should pressure tube leaks develop they can readily be i
detected by monitoring the moisture content in the gas space between the
pressure tube and the calandria tube. This is normally done on a continu-
ous basis. Also the pressure tube concep: makes it possible to detect
release of fission products from the fuel in an individual fuel channel
due to cladding defects. This can be done while the reactor is operating.
 -3- TDAI-244

The pressure tube concept also permits the flexibility to

subdivide the primary heat transport system into more than one circuit
should the process of optimizing the design of the shutdown systems to
cope with loss-of-coolant akidents, the design of the emergency coolant
injection system, and the design of the primary heat transport system
components indicate this is desirable.

2.2 Reactor Calandria

The calandria is-a horizontal cylindrical shell the

primary purpose of which is to support the fuel channels assemblies and
to contain the heavy water nnderator and reflector. The calandria also
supports guide tubes for reactor devices and in-core instrumentation.
These pass between the calandria tubes and are :herefore situated in a
low-pressure environment.

The calandria is provided with pressure relief valves as

part of a cover gas system which regulates pressure of the moderator
system under normal operation. Rupture discs located at the end of the
four pressure relief pipes (see Figure 2.1.1) ~limit the pressure rise in
the calandria that would occur in the event of an accidental rupture of
a pressure and calandria tube, although the probability of this actually
happening is very small.

The calandria assembly is embedded within the light water

filled carbon-steel lined concrete vault (see Figure 2.2-l). At each
end of the caiandria shell is an end shield containing biological shielding
material in the form of carbon steel.balls and light water. The end
shield and calandria at one end is attached to the calandria vault to
limit the seismic response of the calandria assembly.

2.3 Reactivity Devices

The primary method used to control the reactivity of

CANDU reactors is through on-line refuelling which occurs on a daily
basis. In the 600 HWe CANCU PHW there are six means of changing the
 -4- TDAI -241?

reactivity state of the core besides refuelling. Four of these are used
for norm+1 control functions including controlled shutdown and two are
used by special safety systems for rapid shutdown during accident
conditions.

For control purposes the following are used:

(a) 14 liquid ozone controllers (Hz0 filled compartments)
(b) 21 adjuster rods
(c) 4 mechanical control absorbers

(d) moderator poison

For the special shutdown systems the following are used:

(a) 28 cadmium shutoff rods in one shutdown system
(b) 6 nozzles which permit rapid injection of gadolinium
solution into the moderator which comprise a second,
completely independent shutdown system.

Table 2.3-l gives typical reactivity worths and maximum

rates of change of reactivity of these devices.
 -5- TOAI-244

TABLE 2.3-I

Control And Safety Systems Devices

Typical Reactivity Uorths And Maximum Rates

Total Maximum
Reactivity, Reactivity
Function Device Worth (mk) Rate (mk/s)

control 14 Zone Controllers 7 + 0.14

Control 21 Adjusters 15 + 0.1

c0nrr01 4 Mechanical Control i 0.0075 (driving)

Absorbers 10 - 3.5 (dropping)

control Moderator Poison - +O.Ol (extrac:ing)

Safety 28 Shutoff Units 30 - 50

Safety Poison Injection >300 - 50

Nozzles

f
1 mk is a% value of 0.001 or 0.1%
 - 6 - TDAI -244

All reactivity devices are located or introduced into

guide tubes permanently positioned in the low-pressure moderator environ-
ment. These guide tubes are located interstitially between rows of
calandria tubes as shown in Figure 2.1-2. There exists no mechanism for
rapidly ejecting any of these rods, nor can they drop out of the core.
This is a distinctive safety feature of the pressure tube reactor design.
The maximum reactivity rates achievable by driving all control devices
together is about .35 mk* per second, which is well within the design
capability of the shutdown systems.

The locations of these devices are shown schematically in

Figures 2.3-l (Plan View), 2.3-2 (Side Elevation) and 2.3-3 (End
Elevation).

2.4 Core Design Details

The use of natural uranium fuel and heavy water as moderator

and coolan: combined with capability to.refuel the reactor on power
leads to a design characterized by good neutron economy, since the
fraction of all neutrons produced which are absorbed in the fuel is high
throughout most of the life of the reactor.

The fuel channels are arranged on a square lattice with a

286 millimetre pitch (see Figure 2.1-2). This is a near optimum geometry
from a reactivity standpoint. Figure 2.4-l shows the reactivity change
of a uniform lattice as a function of lattice pitch. A consequence of
the particular lattice geometry used in the CANDU PHW is that the neutron
energy spectrum is very well thermalized. The associated long migration
length {or neutrons and the long neutron lifetime have an important
bearingion methods used in the reactor physics analysis and on the
requirements for the shutdown systems from the neutronic point-of-vie%.

* I mk is a F value of 0.001 or 0.14.

 -7- TDAI -244

2.4.1 Liauid Zone Controllers

The purpose of the liquid zone control system is to

provide the continuous fine control of the reactivity and hence reactor
power level. It is needed because fuelling is not truly continuous but
done in small increments (usually at least 8 bundles at one time). It
also compensates for other mi’nor perturbations in parameters such as
temperatures which cause small reactivity changes. This system is also
designed to accomplish spatial control of :he power distribution which
prevents xenon induced power oscillations in the power distribution from
developing.

The system is contained in six vertically oriented tubes

running interstitially between the fuel channels from the top io the
bottom of the core in the positions shown in Figure 2.3-l. The two
central tubes are divided into three compartments each by appropriately
placed bulkheads and the four outer tubes are divided into two compart-
ments to give a total of 14 individually controllable compartments in
the reactor. H20 is fed to these compartments through small dian;cter
tubing and the plumbing of this tubing <s arranged such that the level
of H20 in the comparcmen: can be controlled by varying the relative
value of the in-flow and the out-flow rates.

The reactor regulating system adjusts the levels of H20

in the individual compartments according to the magnitude of the signal
coming frcm the interstitially placed incore platinum self-powered
detectors of the Hilborn type14]’ [51. There is one detector associated
with each of the 14 compartments (a passive spare detector is also
installed to provide a backup). To ensure that the signal from each of
these individual controlling detectors is representative of the power in
the zone being controlled (see Figure 2.4-2 and Figure 2.4-3) the
detector signal is periodicallir renormalized to agree with the measured
integrated power in the zone as,obtained from the flux mapping system.
The latter is a system of 102 self-powered vanadium flux detectors that
are located in 26 vertical interstitial assemblies. These detectors are
‘
typically about 30 cm long and hence provide essentially a point measure-
ment of the therxal neutron flux. The software in the reactor

.
 - 0 - TDAI-244

regulating system is designed to convert these 102 point measurements to

a local flux distribution throughout the reactor. A typical arrangement
of these flux detectors are shown in Figure 2.4-4 and 2.4-5. The latter
view is just one of the radial planes and is representative of a plane
containing the largest number: of detectors. Other radial planes have
fewer detectors.

2.4.2 Hechanical Control Absorbers

The zone control system is normally designed to provide a

reactivity control capability of about + 3 mk since this is sufficient
to compensate for routine reactivi:y perturbations that occur on a semi-.
continuous basis. For certain less frequent events the reactor fegu-
lacing system requires nnre reactivity range than the zone control
system can provide. Therefore, two additional control absorber systems
are provided which are also operated by the.reactor regulating systen.

t, The system used to extend the range of control in the

negative direction is the mechanical cohtrol absorber rod system. This
system consists of four control absorber devices which physically are
the same as the shutoff rod devices, but they do not form part of the
shutdown system. These control absorbers are normally fully withdrawn
from the reactor while the reactor is operating under normal steady
state full power conditions. They are activated only when circumstances
demand rapid reduction of the reactor power at a rate or over a range
that cannot be accomplished by filling the liquid zone control system at
the maximum possible rate. Modes of insertion range from driving the
rods in pairs to all four being dropped in by gravity following release
of an electromagnetic clutch in a manner similar to the operation of the
shutoff rod,s. The mode of insertion depends on the nature of the event
demanding a’ rapid power reduction.

Since the power coefficient of reactivity is negative in

the CANDU reactor a power reduction tends to increase reactivity and the
reactivity worth of the mechanical control absorber system is chosen So
that the combined effect of this system and the zone control system
 -9- TOAl-

aciing together will reduce power to a very low value without requiring
activation of either of the shutdown systems.

The positions .of these absorbers are shown in

Figures 2.3-l and 2.4-4.

2.4.3 Adjuster Rod Absorber System

To extend the range of the reactor regulating system in

the positive direction beyond tha i available from the tone control
system, the reactor is designed to operate with a group of absorber rods
fully inserted in the reactor during normal full power operation. This
system of rods is called the adjuster rod system and in the 600 HWe
reactor consists of 21 stainless steei rods. If more positive reactivity
is required than the zone control system can provide, these rods are
withdrawn in groups as necessary.

There are two circumstances where the reactivity decreases

relative to the normal steady state power condition to a degree that
demands withdrawal of some or all of the adjuiter rods to permit continu-
ing operation of the reactor.

(a) Fuelling machines being unavailable for a period of

more than about one week after which the reactivity
decrease due to burnout of the fuel will typically
exceed the range available in the zone control
system.

(b) Transient increases in the concentration of’

xenon 135 follcwing a reduction of reactor power.

The adjuster system is nominally designed to have sufficient

reactivity to compensate for the increase in xenon 135 concentration

that occurs within 30 minutes following a reactor shutdown. Such a

system provides capability to operate with fuelling machines unavailable
for about a month.
 - 10 - TDAI-244

Since the adjuster rods are normally fully in the core

their positions in the reactor and the distribution of absorbing
materials amongst the rods are chosen to flatten the power distribution
in an optimum manner so as to minimize the variation in the discharge
burnup of the fuel that is necessary to achieve the design power shape.
The average to maximum channel power ratio in the reactor is a parameter
which is chosen during the conceptual design stage and it determines the
number of channels chat ara provided in the reacior to achieve a given
total output, wi:hout overrating any one channel.

The 21 adjusters are grouped into seven banks, not all

composed of an equal number of adjusters. The banks are chosen such
that the reactivity worth of any one bank does not exceed the range of
the zone control sys:em. The reactivity worth of the complete system is
about I5 mk. The maximum reactivity change rate associated with moving
one bank of adjusters is < 0.1 mk per second.

The positons of the adjuster rods are shown in Figures

2.3-1 and 2.4-4.

2.4.4 Hoderator Poison

Moderator poison is used to hold down excess reactivity

during the initi~al fresh‘ fuel condi:ions or during and following shut-
down to compensate for lower than normal 135 Xe levels due to decay.
Boron is used in the former and gadolinium is used in the latter situ-
ation. The burnout rate of gadolinium during operation at full power
following an extended shutdown period is comparable to the xenon growth
rate in terms of reactivity, hence the need to rexwe poisor, by ion
exchange at a fairly rapid and controlled rate is much less hemanding.
Poison can be added to the moderator for these purposes either
automatically or manually.

It should be noted that this system is completely

independent of the very high speed liquid poison injection system which
 - 11 - TDAI-244

is used as a shutdown system. lo the regulating system func:ion the

poison is inserted into the piping used to circulate the moderator
whereas in the poison injection system, the poison is injected through
nozzles that are installed horizontally across the core, and a compl’etely
independent source of poison is used.

2.4.5 Shutdown Systems

The 600 f?We reactor is equipped with two physically inde-

pendent shutdown systems. These systems are designed to be both func-
tionally different and geometrically separate. These, differences are
achieved by using vertically oriented mechanical shutoff rods in one
system and horizontally oriented liquid poison injection nozzles in the
second system.

2.4.6 Shutoff Rods

The shutoff rods are tubes made up of a cadmium sheet

sandwiched betwean two concentric steel cylinders. The rods are
inserted into perforated circular guide tubes which are permanently
fixed in the core. The location of’the rods are shown in Figu~re 2.3-l.
The diameter of the rods is the maximum that can be physically accom-
modated in the space between the caiandria tubes (about 113 mm), when
space for the guide tubes and appropriate clearances are allowed for.
The outermost four rods are about 4.4 m long while the rest are about
5.4 m long. The rods are normally fully withdrawn from the core and are
held in position by an electromagnetic clutch. When a signal for
shutdown is received the clutch releases and the rods are ini:ially
accelerated by a spring and then fall by gravity into the core.

2.4.7 Liquid Poison Injection System

The alternative way of shutting down the reactor is

through high speed injection of a solution of gadolinium in heavy water
into the calandria. This is accomplished by opening high speed valves
which normally are closed and retain the solution at high pressure in a
 - I2 - TDAI-244

vessel outside of the calandria. When the valves are open the liquid
poison is injected into the reactor moderator through six horizontally
oriented nozzles that span the core and are located in positions as shown
in Figures 2.3-2 and 2.3-3. The nozzles are designed to inject the
poison in four different directions in the form of a large number of
individual jets. This disperses the poison rapidly throughout a large
fraction of the core. The gadolinium solution is typically held in the
pressure vessel at a concentration of about 8000 g of gadolinium per fig
of heavy water.

2.4.0 Regional Overpower Protective System

Another important consideration in the design of the

reactor core is the requirement to provide an array of seif-powered flux
detectors for application in the regional overpower protective system.
This is done to insure that localized overrating of the fuel does not
occur due to abnormal operation of the reactor, or as the result of
.
malfunctions in the regulating system causing an uncontrolled power
increase to occur.

A separate array of detectors is provided for each of the

two shutdown systems. Those associated with the shutoff rod system are
on carrier tubes that are~verticaliy oriented while those which activate
the poison injection system are on horizontally oriented carrier tubes.
This complies with the philosophy of maximum independence of the two
shutdown systems. The detectors for the regional overpower proteciive
system are “prompt” platinum detectors like those used for the spatial
control system in the regulating system function. Typical positions of
these flux detectors are shown in Figures 2.4-6 and 2.4-T. The latter
shows just one of the diametral “planes” of detectors.
 - 13 - TDAI-244

FIGURE 21.1 REACTOR ASSEMBLY

 FIGURE 2.1.2 SCHEMATIC OF CANDU.PHW LATTICE

,“., i., ,,, “,, ,,, .

 -

- 15 - TDAI -244

FIGURE 22.1 CONCRETE CALANDRIA VAULT

 - 16 -

FIGURE 23-l REACTOR GA PCAN

 TDAI-244

-.
-_
-. .
--
-.
--
-_
to*cm.“ll”
.
noyII**
-.Jr-

3
FIGURE 23-2 REACTOR GENERAL ASSEMBLY (SECTION)
 - 18 - TDAI-244

FIGURE 2.3.3 REACTOR LAYOUT - ELEVATION

 - Calandria size unchanged
- Uniform displacemenl of all fuel channels
- Moderator density and temperature

26 27 20 29 30 91
Lauice pilch (cm)

FlGUnE 2.4.1 VARIATION OF REACTIVITY WITH LATTICE PITCH FOR

CANDU.PHW LATTICE

_, Ii ,,,. . ,.
 - 20 - TDAI -244

VFD 2

VFD 9
n /

NokTH /ll’I\

FIXED END OF
FUEL CHANNELS

SDSZ=N
THIS SIDE

FIGURE 2.4.2 RELATION OF ZONE CONTROL UNITS (ZCU) TO THE FOURTEEN

ZONES AND THE REACTOR ZONE CONTROL DETECTOR
ASSEMBLIES VFD 2,3,9,18.23,25
 - 21 - TDAI-244

___-_ --
---m-y F
_----
 - 22 - TDAI-244

VFD- VERTICALFLUXOnEciOR
AA - AOJ”S7ER100
c* - COHTROLABsoR8Eil no0
TClJ- ZONECoNfnoL “Nri

FIGURE 24-4 VEilTICAL FLUX DETECTOR ASSEMBLY LOCATIONS

 - 23 - TDAI-244

FIGURE 24.5 FLUX MAPPING DETECTOR LOCATIONS - VIEW 1

 -24- TDAI-244

r-ii4

FIGURE 24-6 CALANDRIA PLAN SHOWING SDS1 AND SDS2 Dc7ECTORS

(TOP VIEW)
 -25- TDAI -244

FIGURE 24-1 (continued) - VIEW 3