Post on 17-Jul-2020
transcript
1
SIMULATION OF CANDU FUEL
THERMAL-HYDRAULIC BEHAVIOR
DURING SPENT FUEL BAY LOSS OF COOLING EVENTS
Cătălin Zălog
Reactor Physics and Safety Analyses Group
Cernavoda NPP
Romania
IAEA TM on the Phenomenology, Simulation and Modelling of Accidents in Spent Fuel Pools (EVT 1701772)
Vienna, Austria
2 – 5 September 2019
ABSTRACT
After the Fukushima accident, one of the highest priorities for CERNAVODA NPP was to investigate events that
can lead to Spent Fuel Bay (SFB) loss of cooling and loss of coolant inventory.
In CANDU plants, fueling is performed on-power. Daily, fresh fuel bundles are loaded in core and spent fuel
bundles are discharged from the core, transferred and stored in SFB. Due to the SFB limited storage capacity,
bundles having 6 years or more of cooling time are transferred to the Dry Storage Facility. Thus, as per design,
a maximum number of around 38,000 fuel bundles can be stored, at any time, in SFB.
Following a loss of class III and class IV power sources (e. g. Station BLACKOUT):
▪ The cooling and purification systems for SFB water become unavailable;
▪ Consequently, the pool water temperature increases up to boiling;
▪ Due to boiling and vaporization, the pool water inventory and level will decrease in time;
▪ The decrease of coolant level can leave uncovered a number of fuel bundles, degrading their cooling.
The present paper reviews the analysis methodology and results for a typical event of SFB loss of cooling.
Methodologies used in the analysis and results presented are focused:
➢ upon the CANDU fuel thermal-hydraulic behavior during the event and
➢ upon its potential radiological hazard.
2
INTRODUCTION
No matter which is the initiation condition, a certain sequence of events develops for a Spent Fuel Bay (SFB) loss
of cooling accident:
▪ The cooling system of the SFB water becomes unavailable;
▪ Because fuel decay power cannot be removed, the water temperature increases up to boiling;
▪ If SFB cooling remains unavailable, the bay water vaporizes and its level decreases in time;
▪ If FSB cooling still remains unavailable, water must be added to maintain the level above fuel, keeping its
normal cooling conditions;
▪ If FSB cooling and water addition remain unavailable, the water level decreases continuously until top fuel
bundles remain uncovered, experiencing degraded cooling conditions (natural convection in air);
At any time, actions must be taken to restore SFB coolant inventory and cooling capacity.
Two distinct domains are considered for the analysis of fuel thermal-hydraulic behavior:
➢ normal cooling conditions (fuel bundle is covered by the bay cooling water, in normal operating conditions)
and
➢ degraded cooling conditions (fuel bundle is covered by boiling water or remains uncovered)
To evaluate the potential radiological hazard of the fuel stored in the bay, an estimate of the radioactive
inventory and a decay power calculation were performed.
3
HEAT LOAD AND RADIOACTIVE INVENTORY OF THE SPENT FUEL BAY
METHODOLOGY
Calculation Assumptions
Burnup and in-core irradiation power are the key parameters in obtaining fission products inventory and decay
power for a fuel bundle. Because spent fuel bundles discharged in the bay have different exit burnups and were
irradiated at different powers, it is unreasonable to perform inventory and decay power calculation for each
bundle. Consequently, prior calculations, the typical spent fuel bundle (exit burnup and irradiation power) is
defined, assuming that all fuel bundles stored in the Spent Fuel Bay are identical with it.
➢ Typical bundle exit burnup (170 MWh/kgU) is obtained from a statistical analysis on the spent fuel
bundles discharged during the first five years of operation (Fig. 1).
➢ In-core typical irradiation power is, conservatively, assumed to be the nominal design power peak of 800
kW/bundle.
Computer Codes
➢ Calculations, both for radioactive inventory and for decay power, were performed with the ORIGEN-S
computer code (Ref. /2.1/), using a proprieties library specific for CANDU fuel (Ref. 2.2).
➢ For the typical spent fuel bundle, evolution of the decay power, up to 6 years cooling time, is given in Fig. 2
and, as an example, time evolution of I-131 inventory is given in Fig. 3.
4
HEAT LOAD AND RADIOACTIVE INVENTORY OF THE SPENT FUEL BAY
RESULTS AND CONCLUSIONS
Spent Fuel Bay Decay Power
➢ Decay power evolution in SFB is obtained by summing the contribution of all discharged spent fuelbundles, up to the total filling of the bay.
➢ All bundles in the bay are considered to be identical with the typical spent fuel bundle.
➢ It was assumed the refueling program from operation, close to a rate of 2 channels/FPD (16 bundles/FPD).
At any time, while new bundles, with high decay powers, enter the bay, decay power from bundles alreadystored decreases. Fig. 4 shows that SFB maximum heat load is, with a large margin, below the designheat exchangers capacity (2 MW). Also, it is noticeable that the heat load has a consistent decrease during shut-down periods.
Spent Fuel Bay Fission Products Inventory
➢ Fission products inventory in the SFB is obtained by summing the contribution of all discharged spent fuelbundles, up to the total filling of the bay.
➢ All bundles in the bay are considered to be identical with the typical spent fuel bundle.
➢ It was assumed a theoretical, continuous refueling rate of 2 channels/FPD (16 bundles/FPD), close to thevalue obtained in operation.
Fig. 5 shows that short-lived isotopes (like I-131) levels out in the early stage of bay filling, while totalinventory (with prevalent contribution from long-lived isotopes) has a continuous increase, yet with a decreasingslope.
5
SPENT FUEL BAY COOLANT BEHAVIOR
INTRODUCTION
No mater which is the initiation event for a SFB loss of cooling accident, a certain sequence of events develops:
➢ The cooling system of the SFB water becomes unavailable;
➢ Because the stored fuel decay power cannot be removed, the water temperature increases up to the boiling point;
➢ If SFB cooling remains unavailable, in the boiling conditions, the bay water vaporizes and its level decreases in time.
METHODOLOGY
Calculation Assumptions
➢ When SFB cooling is lost, the reactor is shutdown and SFB overall power source decreases due to decay.
➢ In SFB, bundles are stored on steel trays, in stacks of 19 trays each. Total fuel and trays volume is of 110 m3;
➢ Total bay water volume is of 1620 m3;
➢ Above bay water, it is the normal atmospheric pressure;
➢ Initial temperature of the air above water is 300C, initial water temperature is 380C and earth temperature is 120C;
➢ The decay power generated by the fuel stored in SFB is assumed to be transferred to the metallic structure of the fuel bundles,storage trays, bay water, air above water and bay walls.
Computer Codes
No computer codes were used. Assuming fuel in the bay as the heat source, heat transport equations were used, considering:
➢ homogeneous calorimetric heat transfer to the bay water and fuel bundles metal structure,
➢ homogeneous, bulk convection and conduction with the bay surrounding environment.
Sequence of Events
The events simulated are:
➢ heat transfer from fuel to the surrounding elements (bay water, metal, air, walls) up to water boiling point and
➢ heat transfer from fuel to the surrounding elements (water, metal, air, walls) during bay water evaporation (conservatively, it wasassumed that all heat from fuel is used for water evaporation).
6
SPENT FUEL BAY COOLANT BEHAVIOR
RESULTS AND CONCLUSIONS
Bay Water Temperature Increase
Using the methodology and assumptions described above, it was obtained a rate of water average temperature
increase of 0.83 deg./h (Fig. 6). Thus, SFB water average temperature will increase to the boiling point in about
75 hours.
Bay Water Level Decrease
At boiling temperature, if cooling is not restored, due to vaporization, the bay water level decreases with a rate
of about 24.2 cm/day. After about 16 days from the loss of cooling, the bay water is with 1 meter above fuel
and, at about 20 days, the water level reaches the top surface of the fuel bundles (Fig. 7). If water level is
not maintained by addition, fuel bundles will begin to loose their coverage, entering in a degraded cooling
conditions (natural convection in air). Fuel bundles from the top trays will be fully uncovered in about 2 days.
Water Addition Flow to Maintain SFB Water Level Above Fuel
After reaching boiling, due to vaporization, bay water inventory decreases with 0.84 kg/s. Thus, at any time
before the water level reaches fuel top surface, water must be added at a rate, at least, equal with vaporization.
Conclusions
The analysis shows that, if SFB cooling is lost, water temperature increases with about 1 degree/h, reaching
boiling in about 2.5 days. If cooling is not restored, the pool water evaporates and, in about two weeks, its level
reaches one meter above fuel stack. As, with one meter of water above fuel stack, staff access is not restricted by
radiological constraints, it was concluded that, in case of Spent Fuel Bay Loss of Cooling, enough time (more
than two weeks) remains for compensatory measures, i.e. to supply alternative cooling sources.7
FUEL THERMAL-HYDRAULIC BEHAVIOR IN THE SPENT FUEL BAY
METHODOLOGY
CANDU Fuel Description
The CANDU fuel is designed as a bundle of 37 fuel elements distributed in a cylindrical geometry. The main characteristics of theCANDU fuel are given in Table 4.1, while its geometry is presented in Fig. 8.
Computer Codes
For CANDU fuel safety analyses, a set of two computer codes, developed in CANADA, are mainly used.
➢ ELESTRES (Ref. /4.1/) is a fuel-performance computer code. It is used to predict the on-power behavior of a CANDU fuelelement, for a given geometry and power-burnup history, under in-core Normal Operating Conditions (NOC). The code performsthe following major calculations:
▪ pellet and sheath temperatures,
▪ fission-gas release,
▪ internal pressure and
▪ sheath strain.
➢ The ELOCA (Ref. /4.2/) code assesses the thermal-mechanical response of a CANDU fuel element under Transient Conditions(TC). As part of its input, ELOCA requires:
▪ via a transfer file, the initial physical conditions of the fuel, as calculated by ELESTRES,
▪ the power generation history and
▪ the boundary conditions history (i.e. coolant temperature, coolant pressure and sheath-to-coolant heat transfer coefficient).
The code performs its major calculations for:
▪ expansion, contraction, cracking and melting of the fuel,
▪ variations in the fuel element internal pressure,
▪ changes in the fuel-to-sheath heat transfer,
▪ deformation of the sheath,
▪ chemical reaction of Zr with H2O and UO2 and
▪ Beryllium-assisted cracking of the sheath.
8
FUEL THERMAL-HYDRAULIC BEHAVIOR IN THE SPENT FUEL BAY
METHODOLOGY
Fuel Failure Mechanisms
Seven fuel failure mechanisms have been identified to be of relevance to the Design Basis Accidents (Ref. /4.3/)
and to be, also, used in analyses of Spent Fuel Bay loss of cooling events:
1. Sheath failure by overstrain,
2. Low ductility sheath failure,
3. Beryllium-assisted crack penetration,
4. Oxygen embrittlement,
5. Fuel melting,
6. Sheath melting,
7. Failure due to oxidation and overstrain under oxide cracks, including oxidation in a mixed air/steam
environment following bundle ejection.
9
FUEL THERMAL-HYDRAULIC BEHAVIOR IN THE SPENT FUEL BAY
NORMAL COOLING CONDITIONS
It was simulated the behavior of a fuel element from the outer ring of typical fuel bundle discharged from core and
transferred to the Spent fuel Bay operating in normal conditions.
METHODOLOGY
➢ For NOC simulation with ELESTRES code, it was selected the typical spent fuel bundle:
➢ For TC simulation with ELOCA code:
• the power transient was assumed to be the fuel decay curve after discharging from core,
▪ the coolant properties were assumed to be those of the bay water, in normal operating conditions:
▪ pressure = 1 atm (equals the atmospheric pressure above the water surface),
▪ temperature of 40 oC (conservatively, higher then the normal operating setpoint of 38 oC) and
▪ heat transfer coefficient sheath-to-coolant corresponding to free convection in liquids
(conservatively, was selected the lowest value of the domain: 50 W/m2K).
RESULTS AND CONCLUSIONS
The results of the simulation with ELOCA code show that no failure occurs and fuel element integrity is preserved.
For example, see Fig. 9 that checks the overstrain mechanism (criteria 1 and 7).
10
FUEL THERMAL-HYDRAULIC BEHAVIOR IN THE SPENT FUEL BAY
DEGRADED COOLING CONDITIONS
It was simulated the behavior of a typical fuel bundle discharged from core and transferred to the Spent Fuel
Bay which operates following a loss of cooling event. Two types of degraded cooling were considered:
a) cooling in boiling water (if the bundle is still covered by the water) and
b) cooling in air (if the bundle becomes totally uncovered by the water).
METHODOLOGY
➢ For the NOC simulation with the ELESTRES code, it was selected the typical spent fuel bundle:
➢ For the TC simulation with the ELOCA code:
• the power transient was assumed to be the fuel decay curve after discharging from core ,
▪ the coolant properties were assumed to be those of the water from the bay, after a loss of cooling
event:
a) cooling in boiling water
▪ pressure = 1 atm (equals the atmosferic pressure above the water surface),
▪ temperature of 100 oC (boiling conditions) and
▪ heat transfer coefficient sheath-to-coolant corresponding to free convection in liquids
(conservatively, was selected the lowest value of the domain: 50 W/m2K).
b) cooling in hot air
▪ pressure = 1 atm (equals the atmosferic pressure above the water surface),
▪ temperature of 100 oC (air above boiling water) and
▪ heat transfer coefficient sheath-to-coolant corresponding to free convection in gases
(conservatively, was selected the lowest value of the domain: 2 W/m2K).11
FUEL THERMAL-HYDRAULIC BEHAVIOR IN THE SPENT FUEL BAY
DEGRADED COOLING CONDITIONS
RESULTS AND CONCLUSIONS
a) cooling in boiling water
The results of the simulation with ELOCA code show that, in conditions of cooling in boiling water, no failure
occurs and the fuel element integrity is preserved. For example, see Fig. 10 that checks the overstrain
mechanism (criteria 1 and 7).
b) cooling in hot air
The results of the simulation with ELOCA code show that (Fig. 11), in conditions of cooling in hot air, failure
occurs due to the overstrain mechanism (criterion 1).
12
CONCLUSIONS
The analyses showed that if, at all times, spent fuel is kept entirely covered by the pool water,
even at boiling, NO FAILURE occurs.
Consequently, in case of a Spent Fuel Bay Loss of Cooling accident, all efforts must be made to
keep constant the pool water level (i. e. inventory) above fuel.
At Cernavoda NPP, enough time (about two weeks) is available for adequate compensatory
action to be taken.
13
Thank you !
14
REFERENCES
/2.1/ Oak Ridge National Laboratory - ORIGEN-S User’s Manual, NUREG/CR-0200, Vol 3.
/2.2/ Oak Ridge National Laboratory - CANDULIB-AECL: Burnup-Dependent ORIGEN-S Cross-Section Libraries for CANDUReactor Fuel Characterization, RSICC Data Package DLC-210
/4.1/ G. G. Chassie - ELESTRES-IST 1.0 User’s Manual, TTR-733, rev. 1, 2002
/4.2/ D. J. Caswell, A. F. Williams and W. R. Richmond - ELOCA 2.2 User’s Manual, AECL Report 153-113400-UM-001, Rev. 0,2005
/4.3/ A. F. Williams – State of the Art Report on Fuel Sheath Failure Mechanisms for Design Basis Accidents, COG-09-2068,October 2011
/4.4/ G. Sachs and J. D. Lubahn, - Failure of Ductile Metals in Tension, Transactions of the American Society of MechanicalEngineers, vol. 68, p. 271, 1946
/4.5/ D. G. Hardy – High Temperature Expansion and Rupture Behavior of Zircaloy Tubing, National Topical Meeting on WaterReactor Safety, Salt Lake City, Utah, CONF-730304, p. 254, 1973
/4.6/ C. E. L. Hunt – The Limit of Uniform Strain or Onset of Ballooning in Fuel Sheath Ballooning Tests, AECL Report CRNL-1187, 1974
/4.7/ D. G. Hardy – The Effect of Neutron Irradiation on the Mechanical Properties of Zirconium Fuel Cladding Alloys in Uniaxialand Biaxial Tests, AECL Report CRNL-537, 1970
/4.8/ M. F. Ashby – A First Report on Deformation-Mechanisms Map, Acta Mettalurgica, vol. 20, pp. 887-897, 1972
/4.9/ W. R. Richmond et al. – ELOCA 2.2: Theory Manual, COG IST Report SQAD-07-5033, AECL Report 153-113400-COG-012,rev. 0, 2007
/4.10/ E. Kohn and W. R. Clendening – A Model for Predicting the Behavior of Zircaloy-4 Fuel Sheath Ruptures at BrazedAppendages, Westinghouse Canada Limited Report CWAPD-313, CANDEV 78-06, 1978
/4.11/ A. Sawatzky – A proposed Criterion for the Oxygen Embritlement of Zircaloy-4 Fuel Cladding, Zirconium in the NuclearIndustry (4th Conference), ASTM STP 681, p. 497, 1979
/4.12/ MATPRO – Library of Material Properties for Light-Water-Reactor Accident Analysis, Idaho National Engineering andEnvironment Laboratory Report INEEL/EXT-02-00589, Vol. 4, Rev. 2.2, 2003
/4.13/ P. J. Hayward and I. M. George – Determination of Melting Points of Zircaloy-4/oxygen alloys, COG Report COG-98-159,1998
/4.14/ IAEA – Thermo physical properties database of materials for light water reactors and heavy water reactors , IAEA-TECDOC-1496, 2006
15
16
Table 1. – Summary of Fresh Fuel Bundle Characteristics
Parameter Best-Estimate Standard ErrorStatistical
Distribution
PELLET
Number of pellets in a fuel element 30 + 1 N/A N/A
Pellet diameter [mm] 12.222 0.0262 Normal
Pellet density [g/cm3] 10.59 0.0024 Normal
235U enrichment [wt%] / [atom %] 0.711 / 0.72 N/A N/A
234U [atom %] 0.0055 N/A N/A
238U [atom %] 99.2745 N/A N/A
Pellets stack length in fuel element [mm] 481.71 0.433 Uniform
FUEL ELEMENT and BUNDLE
Number of fuel elements in a bundle 37 N/A N/A
Sheath outside diameter[mm] 13.095 0.014 Uniform
Sheath inside diameter [mm] 12.31 0.014 Uniform
Filling gas pressure [atm] 1.0 N/A N/A
He content in the filling gas [volumetric fraction] 0.9001 - -
Bundle U mass [kg] 19.3286 0.0005 Normal
Bundle UO2 mass [kg] 21.927 0.0005 Normal
Bundle Zy mass [kg] 2.1524 0.00005 Normal
Bundle length [mm] 495.896 0.001 Normal
Bundle outer diameter [mm] < 102.77 0.003 Uniform
Volume displaced by the fuel bundle [cm3] 2459.83 5.50 Normal
Figure 1 – Statistics on discharged spent fuel bundles
17
0.01% 0.18%0.74% 0.60%
1.19%
2.69%
3.91%
7.10%
20.59%
29.52%
19.42%
7.77%
4.91%
0.94%0.22% 0.15% 0.03% 0.01%
0%
5%
10%
15%
20%
25%
30%
35%3 23 43 63 83 103
123
143
163
183
203
223
243
263
283
303
323
343
Discharge burnup [MWh/kgU]
Num
ber
of b
undl
es
Average Burnup = 169.48 MWh/kgU
Standard Error = 0.28 MWh/kgU
Figure 2 – Decay heat for the typical spent fuel bundle
18
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Cooling time (days)
kW
atts/b
un
dle
(C
oo
ling
tim
e: 1
h -
30
da
ys)
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0 500 1000 1500 2000 2500
Cooling time (days)
kW
atts/b
un
dle
(C
oo
ling
tim
e: 3
0 d
ays -
6 y
ea
rs)
Cooling time: 1h - 30 days
Cooling time: 30 days - 6 years
It is assumed a delay time of 1 hour (~ 0.04 days) between
the moment when the bundle is discharged from the reactor
core and the moment when the bundle enters the Spent Fuel
Bay. (6.24 kW/bundle)
Note that, at the moment when the fuel bundle is discharged
from the reactor core, its decay power is 28.55 kW.
Figure 3 – I-131 inventory for the typical spent fuel bundle
19
0
100
200
300
400
500
600
700
800
900
1000
1100
1200
0 5 10 15 20 25 30
days
I-1
31
in
ven
tory
(T
Bq
/bu
nd
le)
0-3
0 d
ay
s
1.E-20
1.E-18
1.E-16
1.E-14
1.E-12
1.E-10
1.E-08
1.E-06
1.E-04
1.E-02
1.E+00
1.E+02
1.E+04
0 100 200 300 400 500 600
days
I-1
31
in
ven
tory
(T
Bq
/bu
nd
le)
0-
600
da
ys
0 - 30 days
0 - 600 days
Figure 4 – Spent Fuel Bay Decay Power
20
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
0 365 730 1095 1460 1825 2190 2555 2920
Time from first refueling (days)
Sp
en
t F
uel B
ay D
ecay P
ow
er
(MW
)
Figure 5 – Spent Fuel Bay Activity
21
0.0E+00
2.0E+06
4.0E+06
6.0E+06
8.0E+06
1.0E+07
1.2E+07
1.4E+07
0 500 1000 1500 2000 2500 3000
Duration from first refueling (days)
Sp
en
t F
uel B
ay T
ota
l A
cti
vit
y (
TB
q)
0.0E+00
2.0E+04
4.0E+04
6.0E+04
8.0E+04
1.0E+05
1.2E+05
1.4E+05
Sp
en
t F
uel B
ay I-1
31 A
cti
vit
y (
TB
q)
Total activity (TBq)
I-131 activity (TBq)
Figure 6 – Pool water average temperature following a SFB loss of cooling event
22
0
10
20
30
40
50
60
70
80
90
100
110
0 6 12 18 24 30 36 42 48 54 60 66 72 78 84
Timp [h]
Te
mp
era
tura
[°C
]
Figure 7 – Pool water level following a SFB loss of cooling event
23
95.000
95.500
96.000
96.500
97.000
97.500
98.000
98.500
99.000
99.500
100.000
0 2 4 6 8 10 12 14 16 18 20 22
Timp [zile]
Niv
el [m
]
Figure 8 – CANDU Fuel Bundle
24
Figure 9 – CANDU Fuel Behavior during normal cooling conditions (in cold water)
25
Figure 10 – CANDU Fuel Behavior during degraded cooling conditions (boiling water)
26
Figure 11 – CANDU Fuel Behavior during degraded cooling conditions (in air)
27