Sie sind auf Seite 1von 7

Progress in Nuclear Energy 85 (2015) 511e517

Contents lists available at ScienceDirect

Progress in Nuclear Energy


journal homepage: www.elsevier.com/locate/pnucene

Effect of high density dispersion fuels on transient behavior of MTR


type research reactor under multiple reactivity transients
Rubina Nasir a, Muhammad Kamran Butt b, Sikander M. Mirza c, Nasir M. Mirza c, *
a

Department of Physics, Air University, PAF Complex, E-9, Islamabad 44000, Pakistan
Department of Nuclear Engineering, Pakistan Institute of Engineering & Applied Sciences (PIEAS), Post Ofce Nilore, Islamabad 45650, Pakistan
c
Department of Physics & Applied Mathematics, Pakistan Institute of Engineering & Applied Sciences (PIEAS), Post Ofce Nilore, Islamabad 45650, Pakistan
b

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 13 January 2015
Received in revised form
25 July 2015
Accepted 27 July 2015
Available online xxx

The effect of fuel on transient behavior under simultaneous multiple reactivity insertion in a typical MTR
type research reactor have been studied. The PARET code was used for three different high density fueled
cores. The fuels included oxide (U3O8-Al), silicide (U3Si2-Al) and aluminide (UAlx-Al) respectively. In rst
part single ramp reactivity insertions were done and results were found in good agreement with already
published data. Then simultaneous multiple reactivity insertion transients were investigated with totally
disabled control systems. Multiple reactivities were inserted in the IAEA benchmarked MTR type
research reactor and effects on reactor power and temperatures (fuel, clad and coolant) were studied.
The simultaneous accidents included ramp reactivity (large and small) plus beam tube ooding, a ramp
reactivity plus a core movement and a ramp reactivity plus a sample ejection accident respectively. In all
cases, the reactivity rises to peak values and then starts decreasing slowly; however when the ramp is
large the peak values also become high and power also shows multiple peaks. The core having oxide fuel
has slightly higher power peak values while core with silicide and aluminide fuels show similar behavior.
The fuel temperatures in all cases were much higher for the core having oxide fuel as compared to other
two cores.
2015 Elsevier Ltd. All rights reserved.

Keywords:
Research reactors
U3O8eAl
U3Si2eAl and UAlx-Al fuels
Multiple ramp reactivity accidents

1. Introduction
Research reactors are prone to severe accidents with possible
release of mega-Curies of hazardous ssion product activity to the
environment. Uncontrolled large reactivity insertion in a reactor
can produce such an accident. Efforts have been made in past to
determine the reactivity insertion limits for various research reactors both experimentally and theoretically. The PARET code was
initially developed as a part of the SPERT program to study reactor
transients theoretically (Obenchain, 1969). The code validation was
done by comparing its predictions with the actual experimentally
measured transient response in various SPERT cores. The computed
values of reactor power at the peak burst using the PARET code
showed a good agreement with the corresponding experimental
data (Cullen et al., 1986).
In the RA-2 critical facility in Argentina a super-prompt critical
excursion occurred in September 1983. This accident took place

* Corresponding author.
E-mail addresses: nasirmm@yahoo.com, nmm@pieas.edu.pk (N.M. Mirza).
http://dx.doi.org/10.1016/j.pnucene.2015.07.018
0149-1970/ 2015 Elsevier Ltd. All rights reserved.

while the reactor core was being recongured through a series of


fuel shufing all assumed to be in the sub-critical state (Waldman
and Vertullo, 1987). This accident demonstrated the need for
knowledge of the safe limit for the reactivity insertion.
The Reduced Enrichment for Research and Test Reactor (RERTR)
program was started to convert Highly Enriched Uranium (HEU)
cores to new Low Enriched Uranium (LEU) cores. Several studies
were carried out to nd effects of reactivity, power and ow rate
induced reactor accidents (Hainoun et al., 2010; Iqbal et al., 1997;
Metos et al., 1992; Mirza et al., 1998; Mirza, 1997; Nasir et al.,
1999). Metos et a1., studied the transient response of the High
Enriched Uranium (HEU) fueled core, Low Enriched Uranium (LEU)
fueled core and mixed cores for ramp insertions of reactivity. In
their studies, they used generic 10 MW MTR cores along with
theoretically calculated estimates of the reactivity feedback coefcients. Similarly, a series of benchmark transients for IAEA
10 MW research reactor have been analyzed using a simulation
model. They studied ramp reactivity insertion transients with
initial power set at 1 W and scram level at 12 MW (Gaheen et al.,
2007). Their computed values for peak reactor power, and clad
peak temperatures by using RELAP5 and RETRAC-PC showed good

512

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

agreements. Also, analysis of the reactivity insertion accident (RIA)


for the IAEA 10 MW reactor and the Egypt's second Research
Reactor (ETRR-2) have been carried out for super prompt-critical
transients (Khater et al., 2007).
Recently, single reactivity insertion accident at startup for the
IAEA 10 MW MTR system was studied using the MERSAT computer
code (Hainoun et al., 2010). For the range of inserted ramp reactivity values, for only $1.5/0.5 s value the peak power reached the
highest value of 133.66 MW in 0.625 s and subsequently sub-cooled
boiling and void formation occurred. Comparison of peak values
showed good agreement with results from RELAP5/Mod 3.2 and
PARET code.
Recently many researchers have reported transient analysis of
MTRs for single transients (Mirza et al., 1998; Muhammad and
Majid, 2010; Nasir et al., 1999). In their work, many scenarios
were simulated for the transient behavior of an MTR under
controlled and uncontrolled conditions. Controlled scenarios are
the one in which safety system trip is enabled where uncontrolled
are the one in which safety system trip is disabled.
The Pakistan Research Reactor-I (PARR-I) was also upgraded
from 5 MW HEU to 10 MW LEU core under the international RERTR
program. Extensive experimental measurements of various parameters of the upgraded Pakistan Research Reactor-1 (PARR-I)
core were carried out including the measurement of various coefcients of reactivity (Iqbal et al., 1997). These measurements
show that for the PARR-I upgraded core, the fuel temperature coefcient of reactivity is about 21% lower as compared to HEU type.
Similarly, the void coefcient of reactivity is about 31% higher and
the moderator temperature coefcient of reactivity is about 24%
lower as compared with the corresponding previous estimates for
HEU core. Also, the rod drop times and both the differential and the
integral worth of the PARR-I control rods have also been measured
experimentally (Iqbal et al., 1997; Mirza et al., 1998). Simulations
showed that the LEU core is more sensitive to perturbations at low
power as compared to transients at full power (Mirza et al., 1998).
This work aims at analysis of simultaneous multiple reactivity
insertion transients and for a typical MTR type reactor using oxide
(U3O8eAl), silicide (U3Si2eAl) and aluminide (UAlx-Al) fuels
respectively. As per standard practice, both Start-up as well as
Reactor at Full Power reactivity insertion accidents have been
considered. Using parameters of IAEA benchmark 10 MW MTR type
system having LEU core, the PARET code was employed to simulate
the transients. In rst part single ramp reactivity insertions were
studied and results were compared with published data. It was
assumed that the control rod derive mechanism malfunctioned and
control rods ejected from the core (partially or fully out position)
and got stuck there which posed a ramp type positive reactivity
insertion into the core. Then at the same time another component
failed and it led to another reactivity insertion simultaneously.
These credible accidents due to simultaneous & multiple component failures were simulated for three different cores and their
effect on power, net reactivity of the system, fuel & clad temperatures were investigated. The effects of fuel type were carried out for
multiple failures in the same system for three different accident
scenarios and these included: ramp reactivity (large and small) plus
beam tube ooding, a ramp reactivity plus a core movement, and a
ramp reactivity plus a sample ejection accident respectively.

Reactor has nominal power of 10 MW. It has 5  6 matrix (as


shown in Fig. 1) in which 21 fuel element and 4 control fuel element
can t in. Reactor core is reected by graphite on two opposite sides
with a thickness of one fuel element. Reactor pool is lled with
reactor grade puried water. A summary of the design and thermal
hydraulic parameters is given as Table 1. Core has ve ux traps.
One ux trap is at the center of the core where the other four ux
traps are at each corner of the core. Central ux trap has aluminum
box of dimensions of 7.7 cm  8.1 cm. This box has a square water
hole of dimensions 5.0 cm  5.0 cm. Standard Fuel Element (SFE) of
IAEA Benchmark reactor uses aluminide (UAlx-Al) fuel with uranium density of 4.40 g/cm3.
3. Theoretical model
The PARET code has been used for the simulation of reactivity
transients, which employs the coupled point kinetic model along
with thermal and hydrodynamic equations (Obenchain, 1969).
One-dimensional conductive heat transfer from the fuel to clad and
the convective heat transfer from the clad surface to the coolant are
solved by this code for all of the specied number of node. The code
automatically switches from the normal convective heat transfer
model to the nucleate, transition or stable lm boiling models
depending on the local values of the coolant temperature.
In the case of low enriched uranium core, the two-channel
model was used to analyze the core; one channel representing
the hottest plate and ow channel while the other average
channel representing the remaining fuel plates in volume weighted
sense. Each ow channel with the entrance effects was selected to
contain 21 axial and 7 radial nodes for all transients. The axial
source distribution was represented by 21 axial regions and a
chopped cosine shape which has an axial power peaking factor of
1.5 for both the average channel and the hot channel. For the hot
channel, this axial distribution was multiplied by the other specied hot channel factors i.e. 1.4 nuclear and 1.2 engineering. The
moderator heat source was taken as 4.5% of the total energy being
deposited directly in the moderator. A linear approximation was
used for all of the reactivity feedback coefcients. The reactor was

2. Reactor description
Under RERTR program IAEA devised a benchmark MTR type and
its major parameters have been reported in its core conversion guide
books (IAEA, 1990). This system is a pool type reactor with high
density low enriched silicide (U3Si2eAl) fuel. All the other design
parameters are same as that of HEU fuelled research reactors.

Fig. 1. The core conguration of a typical MTR system.

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

513

Table 1
Design & thermal hydraulic parameters of IAEA benchmark MTR type reactor.
Active core height
Space at the grid plate per fuel element
Fuel element cross-section
Meat dimensions
Thickness of support plate
Number of fuel plates per fuel element
Number of fuel plates per control element
Identication of the remaining plate positions of the control element

60.0 cm
7.7 cm  8.1 cm
7.6 cm  8.05 cm including support plate
6.3 cm  0.051 cm  60.0 cm
0.475 cm
23 identical plates, each 0.127 cm thick
17 identical plates, each 0.127 cm thick
4 plates of aluminum, each 0.127 cm thick in the position of the rst, the third, the
twenty-rst, and the twenty-third standard plate position; water gaps between
the two sets of aluminum plates.
Enrichment 20 w/o U-235; 390 g U-235 (ru 4.40 g/cm3) per fuel element (23
plates)
46.7372
10 MWth

Specications of the LEU fuel


Mass of U in core (kg)
Total power
Thermal hydraulic data:
Volumetric Flow Rate (m3/h)
Mass ow rate (kg/s)
Core ow area (m2)
Core inlet temperature ( C)
Core outlet temperature ( C)
Pressure at core mid-plane (bar)
Saturation temperature ( C)
Heat transfer area (m2)
Average heat ux (W/cm2)

1000
275.97
0.0788
38
47.6
1.7
115.1
39.91
25.06

assumed initially at full power of 10 MW and at a constant coolant


ow rate respectively, in all transients.
In the current study, the correlations used in the PARET code
were SeidereTate correlation for single phase ow, JenseLottes
correlation for two phase ow, original PARET model for transient
two phase ow, and original PARET model for DNB calculations. The
original Departure from Nucleate Boiling (DNB) correlation available in PARET code has been employed. In the case of single phase
ow, the revised model for the laminar ow has been selected
(Obenchain, 1969; Woodruff et al., 1996). The neutronic & design
parameters were taken from the IAEA TECHDOC-643 (IAEA, 1990).
3.1. Thermal and hydrodynamic model
The core was discretized into 21 axial and 7 radial nodes per
channel and the calculations for the heat conduction through the
fuel meat and clad were performed using the modied PARET code.
In the case of heat transfer across the clad-coolant interface, the
local ow and temperature dependent correlation were employed.
The value of the thermal conductivity and heat capacities calculations for three different core types are shown in Table 2 respectively
(IAEA, 1990; Muhammad and Majid, 2010). The coolant ow and
the convective heat transfer problem was solved by employing the
modied momentum integral method (MIM) in the PARET code
which solves the governing mass, momentum and energy conservation equations along with the pressure balance equation for the
entire domain of interest at each time step. The MIM model uses
channel averaged mass ow rate in the momentum equation
(Woodruff, 1982). Also, the coolant properties once evaluated at
some reference pressure are assumed temperature dependent only
for the rst step and the same values are used throughout the rest

of calculations. Time step control was varied to achieve the optimum CPU time and small error.
The PARET code is generally not applicable to situations where
there is strong spaceetime coupling of neutron ux. The code also
uses steady-state heat transfer correlation and it is limited in its
prediction of thermal crisis by the fact that it employs the steadystate DNB correlation (Woodruff, 1982, 1984). Also it uses an
incompressible hydrodynamic model and a simplied void volume
generation equation. Therefore, hydrodynamic outputs of PARET
code should be interpreted as qualitative indicators. However,
comparison of PARET/ANL and RELAP/Mod3.3 codes for the several
reactivity transients have shown that the predictions of PARET code
are fairly close to the RELAP/Mod3.3 (Woodruff et al., 1996).

3.2. Reactor physics model


The reactor dynamics is based on point kinetic model equations
with continuous reactivity feedback from the thermal and hydrodynamic model in the PARET code. The point reactivity assumptions
have been found to be true for small cores. The governing equations
for the reactor power, P, and number of precursors in the reactor, C,
are following:
6
X
dPt rt  b

Pt
li Ci t
dt
L
i1

(1)

dCi t bi
Pt li Ci t;
dt
L

(2)

i 1; 2; 3; /; 6:

where, b and bi are the total and i'th group delayed neutron

Table 2
Thermal conductivity and heat capacities for the fuel and clad materials used in various systems.
Fuel meat/clad material

Thermal conductivity (W/m-K)

Heat capacitya (J/cm3-K)

UAlx-Al
U3Si2-Al
U3O8-Al
Clad (Al-6061)

50.0
91.1
8.0
180

2.072
1.929
2.072
2.069

Cp a bT where, T is in K degree.

0.0011
0.0007
0.0011
0.0012

T
T
T
T

514

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

fractions respectively. The neutron generation time is L and the li is


the decay constant for i'th group precursors. In the case of the
super-prompt critical transients, the delayed neutrons play only a
minor role and under the prompt critical approximation, the above
equations reduce to (Ott and Neuhold, 1985):

dPt rt  b

Pt
dt
L

(3)

The time dependent reactivity r(t) appearing in both Eqs. (1)


and (3) is composed of two parts: the externally controlled reactivity and the reactivity feedbacks coming from the Doppler
broadening of resonances in the fuel, effects of moderator temperature and the void temperature effects. For various scenarios
studied in this work, the feedback reactivity values are calculated
by the PARET code dynamically using the instantaneous state
values of various reactor parameters. In the neutronic model, six
groups of delayed neutrons have been used in this study. The
respective values of the reactivity coefcients, b, and L for various
cores are shown in Table 3 (IAEA, 1990; Muhammad and Majid,
2010). In these calculations a single ramp reactivity insertions
were modeled by the following function:

8
<0
rt gt
:
rm

for t  0
for 0  t  Dt
for t  Dt

(4)

Where, g is the ramp rate, Dt is the time duration for which the
ramp is applied and rm is the nal inserted reactivity. Similarly,
multiple simultaneous ramp reactivity insertions were modeled by
the following function:

8
<0
rt gt bt
:
rm

for t  0
for 0  t  t1
for t  t1

(5)

Where, g and b are two different the ramp rates between time t 0
and t1. The second ramp is applied without any delay time and rm is
the nal sum of inserted reactivity.
4. Fuel properties
Thermal conductivity and heat capacities for various fuels and
clads are shown as Table 2. Thermal conductivity for the oxide
(U3O8-Al) fuel is lowest among all and its heat capacity is close to
the value for aluminide (UAlx-Al) fuel. The silicide (U3Si2-Al) fuel
has the highest thermal conductivity and its heat capacity is comparable with other fuel types. Also aluminum clad shows high
thermal conductivity and good heat capacity.
The reactivity coefcients and other parameters for three
different fuels are given as Table 3 (IAEA, 1990; Muhammad and
Majid, 2010). There is a small difference between the L values of
different fuels and it is highest for silicide fuels, having the value of
44.0 ms followed by the aluminide fuel at 44.03 ms, and the oxide
fuel has the least L value of 44.01 ms. Similarly, The beff is maximum
at 0.007185 for UAlx-Al, followed by 0.007184 of U3O8-Al, and

minimum at 0.007181 for U3Si2-Al fuel. The amount of U and Pu


isotopes in burned fuel vary for different fuels and remain
responsible for the variation in the beff values. Also, the shape of
neutron uxes of all the groups is same with small change in
magnitude (Muhammad and Majid, 2010). The thermal neutron
ux at the mid of the central ux trap is almost same for different
fuels and alloying elements have little effect on neutron ux
spectrum.
The reactivity feedbacks due to change in fuel temperature
(Table 3) of silicide fuels are nearly same as that of the aluminide
fuel whereas the oxide fuel has reactivity feedback about 2% more
than that of the aluminide fuel. The Doppler broadening causes
higher epithermal neutron absorption in fuel and the epithermal
ux in oxide fuel is high therefore this fuel has high Doppler
feedback coefcient. As seen from the values in Table 3, the
moderator temperature feedback coefcients of silicide and oxide
fuels show little change over the aluminide fuel. The moderator
density feedback coefcients of all the fuels are also nearly same.
5. Results and discussion
Under steady-state conditions, the clad temperature, coolant
outlet temperature and the Maximum DNBR are 73.8  C, 53.2  C
and 10.6 respectively, for all fuel types. The difference in the fuel
maximum temperatures of various fuels is due to the difference in
the fuel thermal conductivity. Effect of lower thermal conductivity
of oxide fuel is visible from the maximum fuel temperature which
is 7.7 K higher than the value for aluminide fuel. The maximum fuel
temperature of the silicide fuels is 0.5 K lesser than the aluminide
fuel.
In the case of multiple reactivity insertions, safety analysis was
done for oxide (U3O8-Al), silicide (U3Si2-Al), and aluminide (UAlxAl) fuels and effects on power and temperatures were studied. It
was shown in several previous works (Woodruff et al., 1996; Mirza
et al., 1998; Muhammad and Majid, 2010; Nasir et al., 1999) that
maximum clad temperatures become more than 250  C and can go
further to melting conditions for reactivity insertions rates of $1.0/
0.5 s or more. Therefore, these ramp reactivity insertions were
termed as large ramp rates and all reactivity insertion rates less
than $1.0/0.5 swere called as small ramp rates. In this work, we
have used $0.4/0.5 s as small ramp and $1.0/0.5 s or more as large
ramps. In this work, the proposed accidental scenarios which could
occur as simultaneous accidents are following:
1) Large and Small Ramp Reactivity Insertion plus Beam Tube
Flooding;
2) Ramp Reactivity Inertion plus a Core Movement Startup
Accident;
3) Ramp Reactivity Insertion plus a Sample Ejaction Accident.

5.1. Ramp reactivity insertion & beam tube ooding


These systems can have ve to six radial beam tubes. The beam
tubes are generally plugged with shielding blocks and lled with

Table 3
Reactivity coefcients & other parameters for the IAEA benchmark cores with different fuel types.
Parameter

UAlx-Al

U3Si2-Al

U3O8-Al

Density (g/cm3)
Coolant temperature (  106  Dk/k  K)
Coolant Density (  106  Dk/k  K)
Fuel temperature (  106  Dk/k  K)
beff (x 1000)
L (msecond)

4.40
95.96
283.87
26.72
7.185
42.03

4.40
96.435
283.449
26.774
7.181
44.04

4.40
96.40
283.716
27.386
7.184
44.01

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

515

demineralized water. However, when an experiment is to be setup,


water is drained and plugs are removed. The transition from the air
lled to water lled state adds a positive reactivity into the core. In
general, it is impossible to visualize an event in which more than
one beam tubes is ooded simultaneously within a very short
duration. However, we have assumed that a crack develops in one
of the air lled beam tube having maximum reactivity effects. Then
it's ooding and transition from air lled to water lled state adds a
reactivity of 0.363% Dk/k. It has been assumed that this reactivity is
added within 0.25 s (IAEA, 1990). These data values are assumed to
be valid for the IAEA 10 MW Benchmark core since PARR-I has
somewhat similar core conguration with central ux trap and
operates at 9 MW power level. In this scenario both single ramp
reactivity plus beam tube ooding occur together and the coupled
effects are analyzed. We have assumed that the reactor was operating at steady state power of 10 MW and the reactor trip system
was disabled.
In rst part we have compared effects of a single ramp reactivity
insertion with a ramp reactivity insertion along with the beam
tube ooding. In next step, the effect of a small ramp reactivity
insertion ($0.4/0.5 s) along with the reactivity insertion due to
beam tube ooding transient on power and temperatures were
studied for different fuel types. Then nally a large ramp reactivity
insertion ($1.3/0.5 s) plus the beam tube ooding transient was
simulated for three cores and effects of fuel type were studied.
5.1.1. Small ramp along with beam tube ooding
A small ramp reactivity ($0.4/0.5 s) was inserted with a simultaneous $0.5/0.25 s due to the beam tube ooding. In the tube
ooding the water starts owing into the beam tube due to a hypothetical crack or an accidental switching of motorized operated
valve that can ood the beam tube. This ooding can cause a positive reactivity insertion in the core. Several experiments have
shown that the beam tube can insert a positive reactivity of about
$0.5 within 0.25 s. Therefore we have assumed ooding with the
insertion rate equal to $0.5/0.25 s (IAEA, 1990).
The transient response of the core to such a coupled accident
under uncontrolled conditions, have been investigated for reactor
at full power. The cumulative effect of both the small ramp ($0.4/
0.5 s) and beam tube ooding ($0.5/0.25 s) on power and temperatures as a function of time are shown as Fig. 2a and b respectively.
The response has been compared with single reactivity of $0.4/0.5 s
for oxide fueled core. Here the rise in power translates into an
accumulation of energy and it starts changing the temperatures.
Then fuel temperature (Fig. 2b) rst rises to a peak value and then
shows a secondary peak as well. The oscillations after rst peak
subside after about 0.8 s and power rises monotonically in the
absence of any scram. The reactivity rises to a peak value and then
starts decreasing slowly. The power rises to a peak value and then
keeps on rising monotonically in the absence of reactor scram
distinct power peaks are observed in multiple reactivity transient
cases. The rst power peak is about 3 times higher than the peak
seen in single ramp reactivity transient. In the case of this uncontrolled system, during next 100 s the reactivity keeps on decreasing
and power keeps on rising slowly. The fuel and clad temperatures
remain close to each other and follow nearly same pattern in these
transients. The time taken to reach the rst peak remains less than
0.5 s for the ramp reactivity insertion of $1.4/0.5 s. Maximum clad
temperatures at rst and second peak are 148.8  C & 147.4  C
respectively. The results for single reactivity insertion are in good
agreement with already published data (Woodruff et al., 1996).
5.1.2. Large ramp with simultaneous beam tube ooding
Then we initiated simultaneously a large ramp reactivity of $1.3/
0.5 s along with beam tube ooding (equal to a ramp of $0.5/0.25 s).

Fig. 2. Comparison of (a) power (MW) and (b) fuel temperature (oC) as a function of
time for single and multiple (small ramp ($0.4/0.5 s) along with beam tube ooding)
transients in U3O8eAl fueled LEU core (initial power 10 MW; scram was disabled).

Again the transient response of the core to such a situation, was


investigated at power level of 10 MW without reactor scram. The
effect of single and multiple reactivity transients on power and
temperatures as a function of transient time are shown as Fig. 2a
and b respectively. During rst 0.3 s, the net reactivity rst rises to
the value of $0.8 and then another peak ($0.7) is observed before it
starts decreasing. In the case of multiple reactivity transients the
power rises from 10 MW to 65 MW in rst 0.3 s and then another
peak (of 63 MW) is formed near 0.52 s; however, for the single
ramp, the rst power peak is low and occurs near 0.56 s. The fuel
temperatures as a function of time are shown as Fig. 2b respectively. In the case of multiple reactivity transient case, the rst and
second peak fuel temperatures (212  C and 225  C respectively)
remain well below the melting points. Only one prominent peak is
observed for the single ramp case. In this transient, the integrity of
the core remains intact as the clad surface temperature remains far
below the melting point. However the coolant started reaching at
saturation point within 0.5 s after the start of this accident and
causes nucleate boiling. A comparison of large ramp reactivity
insertion coupled with beam tube ooding scenario with the small
ramp reactivity insertion coupled with beam tube ooding is
shown in Fig. 2. The large ramp reactivity case shows appreciable
oscillations reaching about 2.3 times higher peak power as
compared with the small ramp reactivity insertion case.
The effects of different fuel types in the core were studied with
large ramp ($1.3/0.5 s) along with beam tube ooding ($0.5/0.25 s).
In the multiple reactivity transient case, the power and temperatures as a function of transient time are shown as Fig. 3 respectively.

516

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

Fig. 3. Comparison of power (solid lines) and fuel temperature (lines plus symbols) as
a function of time (during large ramp ($1.3/0.5 s) with simultaneous beam tube
ooding) for different fuel/clad cores at initial power of 10 MW (scram was not
enabled).

Within 0.5 s, two peaks are observed for both the power and fuel
temperature proles. The reactivity rises to peak values and then
starts decreasing slowly; however when the ramp is large the peak
values also become high. The power also shows two peaks where
the second peak remains higher as compared to the rst peak in all
cases. The power rises to a peak value in less than a second and then
keeps on rising monotonically in the absence of reactor scram for
all cases. The U3O8eAl fueled core shows slightly higher power
peaks as compared to other two cores The U3Si2eAl fueled core
remained lowest in power rise during the transient. The fuel temperatures (Fig. 3) remained far below the melting temperatures.
The fuel temperatures in the core with oxide (U3O8eAl) fuel
remained at higher power peaks as compared to other two cores
due to its low thermal conductivity values. The U3Si2eAl fueled
core is lowest in fuel temperature rise during the multiple reactivity
transient. Fuel temperatures remain below 165  C because of good
thermal conductivity of both silicide (U2Si3eAl) and aluminide
(UAlx-Al) fuels. In these multiple reactivity transients, the coolant
reached at onset of nucleate boiling and attained the saturation
temperature.

Then simultaneously a large ramp reactivity of $1.0/0.5 s is inserted


perhaps due to unintended removal of control rods and at the same
time the reactor core is moved towards the thermal column. The
neutrons that were leaking out from the core are now reecting
back from the thermal column and cause a positive reactivity
insertion. It is assumed that reactivity inserted because of movement of core towards thermal column is at a rate of 1$ per 0.615 s
(IAEA, 1990).
The coupled transient response of this scenario for a specic
ramp of $1.0/0.5 s plus the core movement towards thermal column
(equal to $1.0/0.615 s) is simulated for three different fueled cores.
Both transients occur simultaneously at low power of 1 W when
system scram is disabled. The results for power and fuel temperatures are shown in Fig. 4. The reactivity increases linearly within
rst 0.25 s and becomes constant there for a short duration. It then
decreases and forms a second peak ($0.8) near 1.2 s. In the case of
U3O8eAl fueled core, power rises from 1 W to a peak value within
0.9 s and the second power peak of occurs near 1.24 s. The fuel
temperatures show that the U3O8eAl fueled core remained at
higher peak values as compared to other two cores. The U3Si2eAl
fueled core is lowest in fuel temperature rise during the transient.
Fuel temperatures in U3O8eAl fueled core remain below 165  C
because of good thermal conductivity of fuel. First peak in fuel
temperatures is observed at 165  C. The clad temperatures also
show a prominent peak similar to the fuel temperatures and follow
the fuel temperatures. The clad temperatures remained below the
melting point during these coupled transients.
5.3. A ramp with simultaneous experimental sample ejection
The experiments inside or in the vicinity of the reactor core can
become a potential means of imparting a sudden increase in the
reactivity. It can be inserted by removal of an experiment while the
core is critical. In this work we consider experiment being done
inside the core and the reactivity is inserted from the removal of a
single in-pile experiment. The maximum sample worth is taken to
be 1$ and time taken for the sample ejection is 0.025 s but it mainly
depends upon the sample's weight. Therefore, the reactivity
insertion rate due to sample ejection has been considered to be 1.0$
per 0.025 s (IAEA, 1990).
In this scenario, the reactor is assumed to be operating at steady

5.2. Ramp with simultaneous core movement towards thermal


column
The reactor core can be moved from a position, in which it is
completely surrounded by water, to another position, in which a
portion of its water reector is replaced by graphite thermal column into the stall end operating position. This will add reactivity to
the core. In order to prevent the initiation of excursions through the
rapid movement of the core in the vicinity of the thermal column, a
micro-switch is installed in the bridge drive assembly, which scram
the reactor when the crank controlling bridge movement is
engaged. If the bridge scram interlock fails, the intensity of the
transient will depend upon the speed with which the reactivity is
added. The maximum speed at which the reactor bridge can be
moved by an average person is about 5e13 cm/s.The effect of
thermal column on the core reactivity can occurs only in the last
10 cm of the motion. Hence, we have assumed that such a movement will add a reactivity of about 0.728% Dk=k into the core in
about 0.615 s. In this transient it is assumed that reactor is initially
at very low power of 1 W and the reactor trip system is disabled.

Fig. 4. Comparison of power (solid lines) and fuel temperature (lines plus symbols) as
a function of time (large ramp ($1.0/0.5 s) with simultaneous core movement towards
thermal column transient) in different fuel/clad systems at initial power of 1 W (scram
was not enabled).

R. Nasir et al. / Progress in Nuclear Energy 85 (2015) 511e517

Fig. 5. Comparison of power (solid lines) and fuel temperature (lines plus symbols) as
a function of time (during a small ramp along with sample ejection accident) in
different fuel/clad systems at initial power of 10 MW (scram was not enabled).

state full power of 10 MW and reactor trip is disabled. A specic


ramp of $0.4/0.5 s is coupled with an accident in which an in-pile
experimental sample gets ejected and introduces a positive reactivity ($1.0/0.025 s) into the system. Both transients occur simultaneously and the system scram is disabled. The results for the
power and temperatures are illustrated in Fig. 5 for three different
cores. The core having oxide fuel shows that the reactivity increases
very sharply within rst 0.1 s and it then decreases monotonically
and forms a second peak. The power follows the net reactivity and
it rises to nearly 80 MW within 0.1 s; the second power peak of
45 MW occurs near 0.31 s and afterwards it remains less than
50 MW. The prominent peaks in fuel, clad and coolant temperatures are observed. The clad temperature remained far below the
melting temperatures. For the core having oxide fuel, power rises
from 1 W to 80 MW within 0.9 s and the second power peak occurs
near 0.35 s. The fuel temperatures show that the core with oxide
fuel achieved higher peak values as compared to other two cores.
All cores show similar transient behavior. The core having silicide
fuel remained lowest in fuel temperature rise during the multiple
reactivity transients.
6. Conclusions
When simultaneously two different reactivity transients are
coupled in the LEU fueled MTR type research reactors under
uncontrolled conditions, the self-control of the system decreases.
The power peaks and oscillates along with the net reactivity of
the reactor within rst second of the coupled accident. The
simultaneous accidents included ramp reactivity (large and
small) plus beam tube ooding, a ramp reactivity plus a core
movement and a ramp reactivity plus a sample ejection accident

517

respectively. In all cases, the reactivity rises to peak values and


then starts decreasing slowly; however when the ramp is large
the peak values also become high and power also showed multiple peaks.
When a ramp plus beam tube ooding occurred, the power
peaks remained less than 70 MW and fuel temperatures remained
less than 225  C during rst second for all fuel types. In the case of
ramp plus core movement transient under uncontrolled conditions,
the power peaks were found to be less than 75 MW and fuel
temperatures remained less than 190  C during rst second. Also, in
case of a ramp plus sample ejection accident under uncontrolled
conditions, the power peaks were less than 80 MW and fuel temperature remained below the level of 220  C in rst second. Hence
the uncontrolled MTR system under these coupled scenarios did
lead to nucleate boiling in some cases but did not show any clad
melting conditions. The core with oxide fuel showed slightly higher
power peak values while cores with silicide and aluminide fuels
were found similar in behavior. The fuel temperatures in all cases
showed much higher peak values for the core with oxide fuel as
compared to other two cores.
References
Cullen, D.E., Muranaka, R., Schmidt, J., 1986. Applications in Nuclear Data and
Reactor Physics, 756e891. World Scientic, Vienna, Austria.
Gaheen, M.A., Elaraby, S., Aly, M.N., Nagy, M.S., 2007. Simulation and analysis of
IAEA benchmark transients. Prog. Nucl. Energy 49, 217e229.
Hainoun, A., Ghazi, N., Abdul-Moaizb, B.M., 2010. Safety analysis of the IAEA
reference research reactor during loss of ow accident using the code MERSAT.
Nucl. Eng. Des. 240, 1132e1138.
IAEA, 1990. Research Reactor Core Conversion Guidebook, vol. 3. IAEA TECDOC-643,
Vienna.
Iqbal, M., Mirza, N.M., Mirza, S.M., Ayazuddin, S.K., 1997. Study of the void coefcient of reactivity in a typical pool-type research reactor. Ann. Nucl. Energy 24,
177e186.
Khater, H., Abu-EL-Maty, T., EL-Morshdy, S.E.-D., 2007. Thermal-hydraulic modeling
of reactivity accident in MTR reactors. Ann. Nucl. Energy 34, 732e742.
Metos, J.E., Pennington, E.M., Freese, K.E., Woodruff, W.L., 1992. Safety-related
Benchmark Calculations for MTR Type Reactors with HEU, MEU and LEU Fuels.
In: Research Reactor Core Conversion Guidebook, vol. 3. IAEA, Vienna.
Mirza, A.M., Khanam, S., Mirza, N.M., 1998. Simulation of reactivity transients in
current MTRs. Ann. Nucl. Energy 25, 1465e1484.
Mirza, S.M., 1997. Simulation of over-power transients in tank-in-pool type research
reactors. Ann. Nucl. Energy 24, 871e881.
Muhammad, F., Majid, A., 2010. Prospects of using different clad materials in a
material test research reactor e part 4 e the uncontrolled reactivity insertion
transients. Prog. Nucl. Energy 52, 332e338.
Nasir, R., Mirza, N.M., Mirza, S.M., 1999. Sensitivity of reactivity insertion limits with
respect to safety parameters in a typical MTR. Ann. Nucl. Energy 26, 1517e1535.
Obenchain, C.F., 1969. PARET e a Program for the Analysis of Reactor Transients. AEC
Research and Development Report, IDO-17282. USAEC, Idaho.
Ott, K.O., Neuhold, R.J., 1985. Introductory Nuclear Reactor Dynamics. American
Nuclear Society, Illinois.
Waldman, R., Vertullo, A., 1987. Reactivity accident analysis in MTR cores. In: Proc.
of the 10th International Meeting of Reduced Enrichment Research and Test
Reactors, Buenos Airs, Argentina. CNEA. September 28-October 1.
Woodruff, W.L., 1982. The PARET Code and the Analysis of the SPERT-i Transients.
ANL/RERTR/TM-4.
Woodruff, W.L., 1984. A kinetcs and thermal-hydraulics capability for the analysis of
research reactors. Nucl. Technol. 64, 196e206.
Woodruff, W.L., Hanan, N.A., Smith, R.S., Matos, J.E., 1996. A comparison of the
PARET/ANL and RELAP/MOD3 codes for the analysis of IAEA benchmark transients. In: Proc. Of the International Meeting on Reduced Enrichment for
Research and Test Reactors, Seoul, Korea, pp. 260e269. October 7-10.

Das könnte Ihnen auch gefallen