Professional Documents
Culture Documents
By
ENGG01201601037
MASTER OF TECHNOLOGY
of
HOMI BHABHA NATIONAL INSTITUTE
July, 2019
i AA
Homi Bhaba National Institute
As members of the Thesis Examining Committee, we recommend that the dissertation submitted
Light Water Reactor Spent Fuel Pool” be accepted as fulfilling the dissertation requirement
Final approval and acceptance of this dissertation is contingent upon the candidate’s
submission of the final copies of the dissertation to the HBNI.
I hereby certify that I have read this thesis prepared under my direction and recommend that
it may be accepted as fulfilling the thesis requirement.
Date:
Place: Mumbai
i
DECLARATION
I, hereby declare that the investigation presented in the thesis has been carried out by me. The work
is original and has not been submitted earlier as a whole or in part for a degree / diploma at this or
ii
DEDICATIONS
To my beloved family
iii
ACKNOWLEDGEMENTS
Dr. Onkar Gokhale, SO/E Core Safety Studies Section, Reactor Safety Division, BARC,
Mumbai for providing the incentive to pursue the M. Tech project with their invaluable
Srilatha G, SO/E NPCIL for fostering me towards the pertaining literature and developing
I express deep sense of gratitude to Dr. Deb Mukhopadhyay, SO/H, Group Head, Core
Safety Studies Section, Reactor Safety Division, BARC, Mumbai. Throughout the whole project,
his patient guidance, constant encouragement and meticulous attention to detail provide me with
tremendous motivation.
APPLICATIONS-RP, NPCIL, Mumbai for providing his expert comments on my results and
work.
iv
CONTENTS
SYNOPSYS ………………………………………………………………………...... xx
NOMENCLATURE…………………………………………………………………......... xvii
v
2.3.2 THE BOELTER ET AL. CORRELATION [16] ......................................................... 31
2.3.3 THE SHAH CORRELATION [16] ............................................................................. 31
2.3.4 HUGO CORRELATION [16] ..................................................................................... 32
2.4 SUMMARY OF EVAPORATION CORRELATIONS ..................................................... 35
2.5 GAP AREAS ...................................................................................................................... 38
2.6 APPROACH ....................................................................................................................... 40
2.7 SUMMARY ........................................................................................................................ 42
CHAPTER-3. DESCRIPTION OF SFP THERMAL HYDRUILIC MODEL AND
SIMULATION TOOL ............................................................................................................ 43
3.1 SIMUTATION TOOL-RELAP5 ........................................................................................ 43
3.2 AT-REACTOR SFP DESIGN UNDER CONSIDERATION ............................................ 43
3.3 COMPUTATIONAL DOMAIN:........................................................................................ 45
3.4 SFP RACK DESIGN .......................................................................................................... 45
3.5 NODALISATION OF PLANT SCALE SFP ..................................................................... 47
3.6.1 INITIAL CONDITION:................................................................................................... 48
3.6.2 POWER REQUIREMENT .............................................................................................. 50
3.7.2 BOUNDARY CONDITIONS FOR PRE-TEST SIMULATIONS ................................. 50
3.8 PRE-TEST RELAP5 RESULTS AND DISCUSSION ...................................................... 52
3.9 SUMMARY ........................................................................................................................ 55
CHAPTER-4. DESIGN OF EXPERIMENTAL SETUP ................................................ 57
SFP ....................................................................................................................................... 88
vi
5.1 DESCRIPTION OF EXPERIMENT PROCEDURE ......................................................... 88
5.2 1ST SET OF EXPERIMENT ............................................................................................... 89
5.3 2ND SET OF EXPERIMENT ............................................................................................ 104
5.4 3RDSET OF EXPERIMENT ............................................................................................. 118
5.5 4TH SET OF EXPERIMENT ............................................................................................ 139
5.6 5TH SET OF EXPERIMENT ............................................................................................ 150
5.7 6TH SET OF EXPERIMENT ............................................................................................ 153
5.8 SUMMARY ...................................................................................................................... 167
CHAPTER-6. ASSESSMENT OF VARIOUS EVAPORATION CORRELATIONS
vii
REFERENCES .................................................................................................................... 214
viii
LIST OF FIGURES
Figure 3 Natural circulation patterns (a) for loss of cooling (b) for loss of coolant accident. ........ 9
Figure 4 Partly (a) and completely (b) uncovered fuel assemblies in undamaged state. .............. 10
Figure 5 Schematic illustration of partly (a) and completely (b) uncovered fuel assemblies in
Figure 6 Phenomenology of SFP accidents and region of investigation for present work ........... 18
Figure 7 Phenomenology of SFP accidents and region of investigation for present work.. ......... 18
Figure 10 (a) Representation of one rack enclosing one FA inside SFPRELAP5 Nodalisation of
Figure 14 Fluid velocity versus time variation in junctions, 823 discharge of FA rack, 824
discharge of upper plenum, 825 discharge of downcomer, 826 discharge of lower plenum ....... 55
Figure 16 Experimental setup (a) Outer tank section (b) Inner Tank section (c) Immersive U-tube
heater ............................................................................................................................................. 75
Figure 17 Thermocouple Location in the setup (a) Location Planes (b) Sensor ID/location ....... 77
ix
Figure 19 Temperature versus time variation of fluid in scaled overboard water volume ........... 82
Figure 20 Temperature versus time variation of fluid in scaled downcomer water volume ........ 83
Figure 21 Temperature versus time variation of fluid in scaled lower plenum water volume ..... 84
Figure 22 Experimental setup for flow and heat transfer assessment of SFP (a) Electrical and
Figure 23 Free board liquid level temperature distributions Lt=2794mm Pt=10KW ................... 90
Figure 24 Heater shroud inlet and outlet temperatures Lt=2794mm Pt=10KW ........................... 91
Figure 27 Evaporation flux versus average top plane temperature, Lt=2794mm Pt=10KW ........ 92
Figure 28 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 29 (a) to (e) Represents local temperature field within the experimental setup operating
Figure 30 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 31 Free board liquid level temperature distributions Lt=2763mm Pt=6.66KW .............. 105
Figure 34 Evaporation flux versus average top plane temperature, Lt=2763mm Pt=6.66KW ... 106
Figure 35 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 36 (a) to (g) Represents local temperature field within the experimental setup operating
x
Figure 37 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 38 Free board liquid level temperature distributions Lt=2767mm Pt=3.33KW .............. 119
Figure 41 Evaporation flux versus average top plane temperature, Lt=2767mm Pt=3.33KW ... 120
Figure 43 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 44 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 45 (a) to (g) Represents local temperature field within the experimental setup operating
Figure 47 Free board liquid level temperature distribution Lt=2046mm Pt=10KW ................... 140
Figure 50 Evaporation flux versus average top plane temperature Lt=2046mm Pt=10KW ....... 141
Figure 51 (a) to (e) Represents local temperature field within the experimental setup operating
Figure 52 (a) to (e) Represents local temperature field within the experimental setup operating
Figure 53 (a) and (e) Represents local temperature field within the experimental setup operating
Figure 54 Free board liquid level temperature distribution Lt=2027mm Pt=6.66KW ................ 152
xi
Figure 57 Liquid level versus time Lt=2022mm Pt=3.33KW ..................................................... 154
Figure 58 Evaporative mass flux versus time Lt=2022mm Pt=3.33KW .................................... 155
Figure 59 Evaporation flux versus average top plane temperature Lt=2022mm Pt=3.33KW .... 155
Figure 60 (a) to (e) Represents local temperature field within the experimental setup operating
Figure 61 (a) to (e) Represents local temperature field within the experimental setup operating
Figure 62 (a) to (d) Represents local temperature field within the experimental setup operating
Figure 63 SF model, HNC model comparison with Boelter experiment data ............................ 169
Figure 65 Comparison of Hugo’s model against present experimental data for 10KW heater
Figure 66 Comparison of Hugo’s model against present experimental data (evaporation flux
versus average top plane temperature)for 10KW heater power and 2794mm initial liquid level
..................................................................................................................................................... 171
Figure 67 Comparison of Hugo’s model against experimental data for 10KW heater power and
Figure 68 Comparison of Hugo’s model against present experimental data (Evaporation flux
versus average top plane temperature) for 10KW heater power and 2046mm initial liquid level
..................................................................................................................................................... 172
Figure 69 Comparison of developed correlation against Hugo’s model and experimental data
xii
Figure 70 (a) Comparison of developed correlation (Evaporation flux versus average top plane
temperature) against Hugo’s model and experimental data (P=10KW, initial liquid level
2046mm) (b) Parity plot between experimental data versus correlation predictions ................. 178
Figure 71 Comparing developed correlation against Hugo’s model and experimental data
Figure 72 (a) Comparison of developed correlation (Evaporation flux versus average top plane
temperature) against Hugo’s model and experimental data (P=10KW, initial liquid level
2794mm), (b) Parity plot between experimental data versus correlation predictions ................ 180
Figure 73 Comparing developed correlation (Evaporation flux versus average top plane
temperature) against experimental data (P=6.66KW, initial liquid level 2763 mm) .................. 185
Figure 74 (a) Comparing developed correlation (Evaporation flux versus average top plane
temperature) against experimental data (P=3.33KW, initial liquid level 2767mm, (b) Parity plot
Figure 75 Comparing developed correlations with experimental data (P=10KW, initial liquid
Figure 76 Comparing developed correlations (Evaporation flux versus average top plane
temperature)with experimental data (P=10KW, initial liquid level 2794mm) ........................... 191
Figure 77 Comparing developed correlations with experimental data (P=6.66KW, initial liquid
Figure 78 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=6.66KW, initial liquid level 2763mm) ....................... 192
Figure 79 Comparing developed correlations with experimental data (P=3.33KW, initial liquid
xiii
Figure 80 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=3.33KW, initial liquid level 2767mm) ....................... 193
Figure 81 Comparing developed correlations with experimental data (P=10KW, initial liquid
Figure 82 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=10KW, initial liquid level 2046mm) .......................... 194
Figure 83 Comparing developed correlations with experimental data (P=3.33KW, initial liquid
Figure 84 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=3.33KW, initial liquid level 2022mm) ....................... 195
Figure 86 Computational Domain (a) Overall 2D-StructuredMesh, (b) Magnified view of heater
Figure 88 Location of measurement planes within the computational domain .......................... 205
xiv
LIST OF TABLES
Table 4 Comparison of Non-dimensional numbers for plant scale and experimental scale ------- 69
---------------------------------------------------------------------------------------------------------------- 134
Table 17 Comparison of power based evaporation correlation with the Hugo’s correlation. --- 175
Table 18 Comparison of diffusion based evaporation correlation with Hugo’s correlation ----- 184
Table 19 Comparison of Power based and Diffusion based correlations -------------------------- 188
xv
Table 22 Boundary conditions---------------------------------------------------------------------------- 202
xvi
NOMENCLATURE
LIST OF ABBREVIATIONS
CV Control Volume
DM Demineralised Water
FA Fuel Assembly
HT Heat Transfer
MT Mass Transfer
xvii
NC Natural Circulation
Development
Programme
RH Relative Humidity
Association
xviii
LIST OF SYMBOLS
Ts Surface temperature
Pt Total pressure
Y’ Absolute humidity
Liquid density
Le Lewis number
xix
SYNOPSYS
The Spent Fuel Pools (SFPs) are designed to provide safe, stable and secure storage of irradiated
spent fuel assemblies discharge from the nuclear reactor before it is reprocessed. Generally SFPs
are designed as a large accident hardens structure with thick walls. The interest to the safety of
onsite storage of spent fuel increases after occurrence of Fukushima accident on March 11 2011.
The loss in cooling capability in the pool is leading by the water level drop, eventually causing
uncovery of Spent Fuel Assemblies (SFSs) and ultimately failure of SFAs. Simulating the
inherent nuclear SFP response to a loss of pool cooling accident requires precise estimation of
heat source (Decay heat generated by SFAs stored in the pool) and heat sink (Pool water
evaporation from air-water interface and via concrete structure) terms. Order of heat removal due
to surface evaporation is significantly higher than other possible modes of heat transfer such as
wall conduction, air convection, and radiation. This enumerates significant influence of precise
prediction of surface evaporation flux on SFP heat up rate. Existing correlations for predicting
evaporation flux from water surfaces are only calibrated for conditions specific to swimming
pools. Literature survey highlights the limited availability of evaporation flux correlations
specific to SFP applications. A new model of evaporation from warm SFP is presented in this
work. Presented model is applicable from low operating temperature (T = 50oC) (Low mass
transfer regime) up to high operating temperature (T > 90oC) (High mass transfer regime). Two
different empirical correlations are developed that is nearly explicit in solving for pool
temperatures. Limited availability of high temperature evaporation flux data and non-availability
of information on SFP response during the prolonged loss of cooling accident in SFP restricts
our understanding of involved thermal hydraulic phenomena during the course of accident. Only
available data for loss of cooling accident is from Fukushima accident. A timely external
xx
intervention restricts the Fukushima NPP SFP from entering into low liquid level condition.
Several constrains like high radiological risk, availability of very limited measurement sensors,
operational and regulatory limitations causes difficulty in assessment of flow and heat transfer in
NPP SFP. Design of experimental facility for flow and heat transfer assessment has been
convection. Scaling parameters such as Richardson number, Stanton number and geometrical
dimensional scaling ratios are conserved. Results from pre-test simulation studies with RELAP-5
system code are used to finalise the design of experimental setup. Experimental studies have
been performed to develop high temperature evaporation mass flux data for various operating
Empirical correlations are developed for the estimation of surface evaporation flux with two
different approaches. They are diffusion based approach (similar to Hugo’s correlation) and
decay power based approach, each specific to SFP conditions during the course of accident.
Post-test CFD simulation incorporating newly developed evaporation models as heat sink
boundary condition is performed to simulate the pool water heat-up transient during the course of
accident. Heat up rate as calculated from CFD techniques is found to be consistent with the
xxi
CHAPTER-1. INTRODUCTION
This thesis covers the study of thermal hydraulic phenomena of typical light water reactor spent
fuel pool under extended station blackout conditions. Spent fuel storage pool is one of the
important auxiliary facilities of a nuclear power plant and serves as short-term fuel storage
before reprocessing. SFP assumed to be large accident harden structures, severe accidents
involving SFPs are generally regarded as highly improbable events still there are two principal
categories of accidents that may lead to loss of adequate cooling of the spent fuel in a spent fuel
pool, (a) Loss of Cooling accident (b) Loss of coolant accident. This work covers loss of cooling
Spent fuel storage pool is one of the important auxiliary facilities of a nuclear power plant and
serves as short-term fuel storage before reprocessing. The storage is provided at two locations,
firstly in At-reactor Spent Fuel Pool (SFP) and then in away from reactor SFP. About one-
fourth to one-third of the total fuel from the reactor core becomes spent after every refueling
outage. Post outage fuel is discharged from the reactor every 12 to 18 months depending on the
reactor fuel cycle. Spent fuel is stored in SFP until decay heat and radioactive field associated
with it decreases. The water in the pool shields the radiation emitted from the spent fuel. The
heat generated is rejected through a close loop cooling water circulation and purification system
which comprises of heat exchanger and ion exchange equipment’s. Spent fuel pool cooling and
purification system is a part of facility auxiliaries and is responsible to maintain desired normal
operating pool water temperature and water purity. The At-reactor spent fuel pool is present
1
either adjacent to the reactor dry well within the main containment for VVER 1000 design or in
SFP assumed to be large accident harden robust monolithic structures, severe accidents
involving SFPs are generally regarded as highly improbable events. The safety and security of
spent fuel pools are continuously re-assessed as new information becomes available or the
operating conditions of the plants or pools change. For example, the terrorist attacks in the USA
on September 11, 2001, prompted studies on the vulnerability of spent fuel storage facilities to
potential terrorist attacks in many countries. More recently, the Fukushima Daiichi nuclear
accident that followed after the Tohoku earthquake on March 11, 2011, has renewed
international interest in the safety of spent nuclear fuel stored in SFPs under prolonged loss-of-
cooling conditions, although the SFPs and the fuel stored in the pools remained safe during the
accident.
There are two principal categories of accidents that may lead to loss of adequate cooling of
b) Sudden loss of the pool water inventory by leaking (loss-of-coolant accident) due to
In case of loss of spent fuel pool cooling system, heat exchange between pool water and air in
the fuel pool building by natural convention of air above the pool and enthalpy loss due to
surface evaporation of water is dominant mode of heat transfer. If the decay power of stored fuel
assemblies in SFP is high, under Extended Station Blackout Period (ESBO) conditions this mode
of heat loss is insufficient to remove the residual power of the assemblies. The pool water is
2
heated until it reaches the saturation temperature which causes boiling under the free surface.
The steam produced is mixed with air in the fuel pool building hall which can pressurize the
containment or fuel pool building as per the design. Due to deteriorated heat removal from the
spent fuel pool water, fuel assembly’s starts overheating and this may leads to loss of fuel pin
integrity if proper cooling is not restored. Without water injection, the accidents may become a
severe accident.
Safe storage of spent fuel assemblies in the facilities for intermediate storage is very
important. Some of these facilities are not covered with a leak-tight confinement for example
away from reactor spent fuel storage facility. Hence the consequences of overheating and
melting of fuel in the spent fuel pools can be very severe. On the other hand, due to low decay
heat of fuel assemblies, the processes in pools are very slow. Therefore, the accident
management measures play a very important role in case of some accidents in spent fuel pools.
Almost all power reactors have some type of at-reactor pool with active cooling that allows
storage of recently discharged spent fuel until its decay power is low enough for transportation of
the fuel to intermediate off-site storage. The fuel residence time in the at-reactor pool depends on
the applied strategy for spent fuel management, availability of away from reactor SFP and dry
storage. It varies from a few years up to several decades. All SFPs are made of large robust
monolithic structures. The pool walls are generally made of more than 1m (up to 1.5m) thick
concrete with stainless steel liner. The SS liner thickness varies from 3mm in the periphery to 6
mm in the bottom). Fuel assemblies are stored vertically in the racks (PWR, VVER, and BWR
FAs) that provide spacing for coolant flow and in some cases also for criticality control. The
pools are filled with several additional meters of water above the spent fuel to provide biological
3
shielding. An active cooling and purification system maintains optimal conditions for the stored
fuel. Pool water is drawn from the overboard water volume present above the racks by the
suction header to the intake strainers of pumps present in the upper part of the pool and discharge
of cooling and purification system is introduce near the pool bottom via diffuser header.
Differences in pool design exist not only between reactor technologies, but also between
generations for the same plant type and often vary from site to site. Important design differences
are the elevation of the pool with respect to grade and whether it is located inside or outside the
primary containment of the reactor, the design of fuel transfer paths to the reactor, and the use of
borated or demineralized water as coolant. In spite of these differences, the fundamental design
parameters and safety provisions are consistent. Figure 1 schematically illustrates the aerial view
of BWR -SFP.
(a)
4
(b)
Figure 1 (a) BWR-SFP Aerial, (b) Schematic of SFP cross-section. [Source: www.nrc.gov]
In principal there are two types of accidents that may lead to loss of adequate cooling of the
spent fuel in an SFP malfunction or unavailability of the pool cooling system (loss-of-cooling
accident) and sudden loss of the pool water holdup inventory by leaking (loss-of-coolant
accident) due to structural damage in pool walls. The two types of accidents are similar with
regard to inherent phenomena, but the rate of accident progression is faster for the loss-of-
coolant accident due to sudden decrease in pool water inventory. Un-rectified accidents are
expected to evolve from a single dominant phenomenon in the early stages to a progressively
illustrates the phenomenology of SFP accidents. The accidents can be segregated into three
phases, in which different phenomena dominate the course of events. During the first, pre-
uncovery, phase, the spent fuel assemblies (FAs) are submerged under water and the
phenomenology is dominated by the thermal-hydraulics of the SFP. The second phase involves
5
uncovery of the fuel storage racks, which leads to significant heat-up of the FAs and the storage
racks, There is a possibility of criticality issues in the SFP due to change of the coolant density
and levels. The third phase is dominated by damage and disintegration of the spent fuel, storage
racks and possibly also other structures in the pool. The duration of each phase depends strongly
on the type of accident and on the decay power of the spent fuel stored in the rack.
6
1.2.1 PRE UNCOVERY PHASE (PHASE-1)
In the first phase, for both types of accidents involve loss of water from the spent fuel
pool until the racks which hold the spent fuel assemblies start to get uncovered. Since the
fuel remains submerged under water and is effectively cooled, it does not experience any
damage or degradation during this phase which may cause loss of fuel pin integrity. In
PWR and VVER1000 plants, boric acid used for reactivity control in both the reactor and
the SFP, provides sub criticality margin in the pool. In case of loss of coolant accident,
coolant loss will dilute the boric acid concentration in SFP, low boric acid concentration
may leads to criticality in the SFP also which would provide an additional source of heat
and radiation, and also generate an inventory of short-lived fission products in the fuel
that could add extra inventory to the radioactivity release later in the accident if fuel pin
integrity gets compromised. In addition to the risk of criticality, safety issues for the pre-
contaminants from the pool water due to decrease in the solubility as it heats up. If the
pool water level drops to less than about half a meter above the spent fuel assemblies,
biological shield gets compromised which lead to strong radiation field near the SFP.
High radioactive field could prevent access to the SFP building and restricts corrective
measures, surveillance and control. Furthermore, decrease in water level in the SFP could
make it difficult to recover cooling of the SFP by restarting the normal cooling systems,
because of pump cavitation or loss of suction to the intake strainers present in the upper
7
1.2.1.1 THERMAL-HYDRAULICS
rack designs for spent LWR fuel have a closed cell design, in which each fuel assembly is
borated steel. Since this closed cell design does not allow lateral or cross-flow across the
FA stored in the racks, lateral or cross flow can only occur in water volume present above
and below the racks. The overall shape of the natural convection flow pattern in the pool
depends largely on the location of free paths for water to flow downwards and on the
As the pool water heats up, the evaporation rate at the pool surface will increase.
Evaporation of water from the pool surface is considered the dominating mechanism for
heat removal from an SFP with unavailable cooling system. Several models and
correlations are presently available for the estimation of evaporative mass flux from pool
water air interface. Evaporation correlations developed for modeling high evaporative
mass flux in turbulent natural convection regime is discussed in later sections. Measured
data on pool temperature and water loss for the SFPs at the Fukushima Daiichi nuclear
power station after the 2011 accident have been used for validating the models.
It is important to note that the evaporation rate depends on the natural convection
heat transfer from the lower part of the pool to the pool surface. Natural circulation
currents which are responsible for natural convective heat transfer, leads to energy
deposition in overboard water volume present above the racks. Thermal hydraulic
analysis predicts flow instabilities and thermal stratification of water volume above the
pool. Figure 3 schematically illustrates the natural circulation patterns in the pool.
8
Thermal hydraulic analysis results suggest that bulk boiling does not necessarily occur in
an SFP with inoperable cooling system if the heat load is low and the pool building or
pool temperatures well below the boiling point. But for high thermal loads such as for At-
reactor SFP, boiling may occur in the free board water volume above racks. Evaporative
heat loss depends upon evaporative mass flux from pool air-water interface; this mass
flux is proportional to the difference in saturated partial pressure of water in the air-water
film and partial pressure of water in the air. If the building is not properly ventilated it
will get saturated with steam, which deteriorates the evaporative mass flux therefore
Figure 3 Natural circulation patterns (a) for loss of cooling (b) for loss of coolant
accident. [Source: Organisation for Economic Co-operation and Development. (August
25, 2017). PIRT on Spent Fuel Pools under loss of cooling and loss of coolant accident
conditions. (NEA/ CSNI/ R (2017) x)]
9
1.2.2 UNCOVERY PHASE (PHASE-2)
1.2.2.1 THERMAL-HYDRAULICS
In the second phase of accident, the spent fuel assemblies start to get uncovered. It may
not lead to immediate heat-up of the uncovered part of the fuel. Analysis suggest that
uncovered part of FA can be cooled by steam flow and water level swell from boiling in
the lower immersed part of a fuel assembly with low or moderate decay power. However,
for a fuel assembly with high decay power, or for situations with low water levels, the
steam production will not be sufficient to cool the upper part of the FA, and much higher
Figure 4 Partly (a) and completely (b) uncovered fuel assemblies in undamaged state.
[Source: Organisation for Economic Co-operation and Development. (August 25, 2017).
PIRT on Spent Fuel Pools under loss of cooling and loss of coolant accident conditions.
(NEA/ CSNI/ R (2017) x)]
10
Under prolonged un-mitigated conditions the spent fuel assemblies ultimately get
uncovered and the water level drops below the bottom plate of the storage racks, under
this scenario natural convection by air is only left cooling mechanism. By analyzing the
natural circulation airflow in completely drained SFPs and the surrounding building,
highlights that a large scale flow pattern develops inside the pool building: hot air-water
vapor mixture exiting the top of the fuel assemblies forms a plume that rises to the
ceiling. It then spreads laterally within a hot layer. If the layer of hot air-water vapor
mixture beneath the ceiling is evacuated by the ventilation system or by opening roof
hatches, the air in the building may remain thermally stratified as cool air enters at lower
elevation to replace the hot air that exits through the ceiling. The cool air is then drawn
into the SFP, where it spreads laterally beneath the racks and enters the FAs from below.
However, if the building ventilation is inadequate, the analyses suggest that the room will
gradually heat up and the hot gas layer will ultimately drop into the SFP, ultimately
1.2.2.2 THERMAL-MECHANICS
The high temperature experienced by the fuel during the uncovery phase will expedite
cladding creep and oxidation. The creep deformation, which is governed by the internal
gas overpressure in the fuel rod, will cause the cladding tube to expand in its radial
direction. This deformation may ultimately become unstable, if the diameter of the tube
increases at any axial position, the local stress is increased due to the larger diameter and
the reduced wall thickness, provided that the rod internal pressure does not reduce
significantly. This observed positive feedback phenomenon enhances the creep rate,
which may lead to a local runaway deformation (“ballooning”) which results in cladding
11
creep rupture. However, also in cases with stable and limited creep deformation, this
thermal expansion reduces the cross-sectional area (“Blockage”) for coolant flow through
the fuel assembly, which will increases the available cladding surface area exposed to
oxidants, and leads to cracking and spallation of the protective oxide layer at the cladding
outer surface.
The prominent oxidizing agents in the SFP environment are steam, oxygen and nitrogen.
These species may oxidize structural materials of FAs and storage racks, among which
zirconium alloys are the most abundant. The following exothermic reactions between the
As indicated by above equations, all the oxidation reactions are exothermic. The reaction
rates, and thus the rate of heat released in the reactions, increase exponentially with
12
temperature. When the cladding temperature reaches 1100–1200 K, the reaction heat
total heat load. Temperature feedback effects on the oxidation processes may then initiate
fire.
The transition point from the uncovery to the fuel damage phase of the accident is not
very sharp or somewhat indeterminate. The rod integrity may be lost by cladding creep
rupture or excessive oxidation. If the accident remains unattended, the damage may
progress, and ultimately leads to severe consequences. The damage phenomena are
be noted that the conditions are significantly different in SFP accidents. For example, the
decay heat is much lower and neighboring fuel assemblies may have very different heat
13
Figure 5 Schematic illustration of partly (a) and completely (b) uncovered fuel
assemblies in damaged state. [Source: Organisation for Economic Co-operation and
Development. (August 25, 2017). PIRT on Spent Fuel Pools under loss of cooling and
loss of coolant accident conditions. (NEA/ CSNI/ R (20z17) x)]
1.2.3.1 THERMAL-HYDRAULIC
The thermal-hydraulic conditions in the SFP may change considerably during the fuel
damage phase. As damage progresses in the upper part of the FA, debris may relocate
downward and may restrict the axial flow through the fuel assembly. Melting and
candling of the rack material in the damaged region may open pathways for cross-flow
between adjacent rack cells. The flow paths and thermal-hydraulic conditions in the
damaged fuel rack thus become complex and difficult to model via computer simulations.
In addition, the exothermic oxidation reactions will significantly add to the local heat
14
1.2.3.2 FISSION PRODUCT RELEASE
When UO2 fuel is discharged from the reactor during the refueling shutdown, depending
upon the burn-up major of the spent fuel mass still consists of UO2. The rest includes
fission products (FPs) and Trans-uranium elements, many of them being radioactive.
With regard to their release behavior in reactor accidents, the fission products are usually
1. Volatiles: release of volatiles (Xe, Kr, Cs, I) from the fuel is usually complete
before the fuel starts to melt. The release is not significantly influenced by
2. Semi-volatile and low-volatile FPs: the release rate of semi-volatile FPs (Mo,
Rh, Ba, Pa, Tc) and low-volatile FPs (Ru, Ni, Sr, Y, La, Ce, Eu) is very sensitive
3. Non-volatile FPs: negligible release before the fuel melts (Zr, Nd, Pr).
Release of volatiles from the SFP during the accident is therefore proportional to the
extent of fuel heat-up and degradation, i.e. the fraction of fuel exposed to high
Due to the presence of relatively low heat loads and heating rates involved in SFP loss-
of-cooling and loss-of coolant accidents, it is expected that metal in the pool inventory
(structural material) will be oxidized before reaching melting temperatures. However, this
15
will depend on the storage rack design and the structural material of construction. For
example, aluminum, with a melting point as low as 930 K, is used in some rack designs
The SFP may also contain spent B4C-bearing control rods (BARs for VVER1000), which
makes eutectic reactions between the B4C absorber material, stainless steel and zirconium
alloys possible
If melting occurs, the molten material will flow downwards (“candle”) and
solidify in cooler regions of the FA. The accompanying loss of support may also lead to
relocation of partially degraded cladding and fuel material, and it will open pathways for
1.3 MOTIVATION
spent nuclear fuel storage pool design, operation, maintenance, and accident response.
1. During light water reactor nuclear power plant design, sizing of the demineralized water
source must consider for the makeup requirements for the SFP based on the expected
evaporation rate. In the absence of significant pipe leaks, SFP evaporation will be the
largest source of water loss from the NPP. This is a particularly important consideration
2. Once the NPP has been operated, radioactive tritium builds up in all of the NPP coolant
systems. Coolant is transferred between the reactor coolant system and other NPP
systems including the SFP by a number of processes such as during refueling outages the
Reactor pressure vessel and SFP are directly connected to execute refueling. Evaporation
16
will carry a portion of this tritium inventory to the SFP building’s ventilation exhaust
contributing to the site boundary radiation dose. Predicting the impact of this contributor
to offsite dose requires an accurate estimate of the evaporation rate from the SFP.
3. Maintenance on the SFP cooling system may require that the system be shut down to
allow work on pumps, valves, Ion exchange columns and filters. The SFP water heats up
as decay heat produced in the stored irradiated fuel exceeds the ambient losses. Usually
available for the maintenance is needed to ensure the cooling equipment can be returned
to service before this temperature limit is reached. Although some heat is lost from the
pool due to conduction, convection, and radiation heat transfer, the dominant heat transfer
unnecessarily.
4. Following a severe accident such as the prolonged loss of electrical power that occurred
during the 2011 Fukushima accident, neither cooling nor normal makeup to the SFP will
competing needs for reactor pressure vessel (RPV) injection for core cooling, containment
injection for containment cooling, and SFP makeup to maintain the spent fuel covered with
water. An accurate method of estimating cumulative evaporation losses from the SFP is
essential for determining when to shift emergency makeup from the RPV or containment
to the SFP. Premature diversion of water to the SFP could result in unnecessary additional
17
damage to the core or containment; late makeup could result in damage to spent fuel and
thermal hydraulics of spent reactor pool during first phase of the accident under ESBO
condition and to achieve realistic thermal hydraulic model of the spent fuel pool in pre-
uncovery phase.
The study in this report is generic with regard to reactor fuel and fuel storage rack design.
Present work considers loss of cooling accident in an at-reactor spent fuel pool of typical
design, VVER 1000. Understanding developed by the present work can also be implied to
The investigation into the divergence between the observed spent fuel pool evaporation
rates and those predicted by existing correlations makes it imperative to study the
thermal-hydraulics behaviour of LWR SFP during initial uncovery phase. The objective
pool levels.
from the SFP tends to over predict the evaporation. The over prediction is mainly
because of lack of understanding of evolution of flow and heat transfer in the pool
experimental setup details [pool size, numbers of FAs, power distributions among
FAs and rack dimensions] will be worked out. RELAP5 simulation tool will be
2. CFD model development for the scaled down experimental setup to be done
3. Design, fabrication and experimentation for flow and heat transfer assessment of
SFP.
experimental observations.
19
5. Assessment of existing models of RELAP5 code and CFD tool and evaporation
1.5 DELIVERABLES
1. Experimental data base generation for early stage of loss of cooling accident in
SFP.
parameter code for pool natural convection phenomena and SFP evaporation.
The flow of the thesis starts with Chapter-2 which includes literature review on the
evaporation and its related applications. This chapter compares the applicability of
applications and highlights the possible gap areas. This is followed by Chapter 3 which
comprises of VVER1000 SFP Thermal Hydraulic Simulation model and Simulation tool
used for the analysis. Chapter 4 comprises of the design of the experimental setup for
heat and flow assessment of VVER1000 SFP. This chapter includes the non-dimensional
scaling laws and similarity criteria followed during the design of the experimental setup.
This chapter also covers the electrical equipment’s and measuring instruments utilised in
the experimental setup. This is followed by Chapter 5, experiment for flow and heat
20
transfer assessment of SFP. This chapter comprises of description of operating
various experimental correlations against experimental data. This chapter also covers
simulation of the experimental setup. Finally the thesis ends with Chapter 8 stating the
1.7 SUMMARY
Spent fuel storage pool is one of the important auxiliary facilities of a nuclear power plant
and serves as short-term fuel storage before reprocessing. Spent fuel pool cooling and
normal operating pool water temperature and water purity. SFP assumed to be large
accident harden robust monolithic structures, severe accidents involving SFPs are
generally regarded as highly improbable events. There are two principal categories of
accidents that may lead to loss of adequate cooling of the spent fuel in a spent fuel pool,
(1) Loss of cooling (2) Loss of coolant accident. The accidents can be segregated into
three phases, in which different phenomena dominate the course of events. During the
first, pre-uncovery, phase, the spent fuel assemblies (FAs) are submerged under water
and the phenomenology is dominated by the thermal-hydraulics of the SFP. The second
phase involves uncovery of the fuel storage racks, which leads to significant heat-up of
the FAs and the storage racks, There is a possibility of criticality issues in the SFP due to
change of the coolant density and levels. The third phase is dominated by damage and
disintegration of the spent fuel, storage racks and possibly also other structures in the
21
pool. The duration of each phase depends strongly on the type of accident and on the
decay power of the spent fuel stored in the rack. For simulating NPP-SFP response during
loss of cooling accident, precise estimation of heat source and heat sink term is necessary.
As evaporation from air water interface is dominating mode of heat removal, precise
estimation of evaporative mass flux is desirable. The investigation into the divergence
between the observed spent fuel pool evaporation rates and those predicted by existing
22
CHAPTER-2. LITERATURE SURVEY
Spent Fuel Pool (SFP) loss-of-cooling accidents feature very long time horizon–ranging
in the order of days if not weeks–before reaching degraded conditions. This is the main
reason why many nuclear power plant operators have not taken any further action to
improve safety related to the SFP accident response. In addition, conducted probabilistic
risk analysis show very low core damage frequencies based on the available time for
recovery actions, but also on its high degree of versatility (as there are multiple available
means to provide alternative cooling in the case of an accident) and relatively low degree
of complexity (as the safety of the SFP relies on covering the FAs (fuel assemblies) with
water). By not considering the above, both IAEA and WENRA have issued
which nuclear plants shall prevent conditions that result in large and/or early radioactive
releases.
These organizations have warned nuclear power plant operators not to solely base their
arguments in low-frequency numbers but to support them with highly reliable safety
counter measures against severe accidents has increased after the events that unfolded in
Fukushima-Daiichi.
A widen approach to risk is now emphasized, where events falling under risks assumed in
the operation of the plant is now being taken into consideration. A broader approach to
risk is now emphasized, where events falling under residual risks (i.e. risks assumed in
the operation of the plant). In a similar accident event, severe damage to fuel assemblies
took place during an incident at Unit 2 of PAKS nuclear power plant in Hungary [3]. The
23
assemblies were being cleaned in a special tank below the water level of the spent fuel
storage pool in order to remove crud build-up. The first evaluation of the event showed
that the severe fuel damage happened due to inadequate coolant circulation within the
cleaning tank. The OECD-IAEA PAKS Fuel Project [3] aimed to support the
PAKS-2 event. Nuclear Energy Agency (NEA) committee on the safety of nuclear
installations ‘Phenomena Identification and Ranking Table (PIRT) on Spent Fuel Pools
coolant accidents in spent fuel pools. This is done by applying a PIRT process
importance and of high uncertainty, and therefore pose sufficient risk to merit new
Kaliatka A. [4] modelled IGNALINA NPP Unit 2 spent fuel pool, using system thermal
hydraulic code RELAP5/MOD3 and severe accident analysis code ASTEC under
possible consequences of fuel overheating due to the leakage of water from a spent fuel
pool. In his capacity Kaliatka A. evaluate the inherent response of spent fuel pool with
the accident progression. Important parameters such as temperature variation within the
pool, reflooding characteristics and possible hydrogen release (if accident enters into
Phase-3) during the course of accident are also estimated. In the computational model of
SFP, heat loss through concrete walls is modelled by using suitable boundary conditions
24
but no effect of heat and mass loss due to surface evaporation is taken into account. On a
fuel pool and demonstrated the applicability of GOTHIC code in low pressure systems.
ChihiroYanagi [6], and Hung, T-C, et al[7] used of Computational Fluid Dynamics
(CFD) tool to demonstrate possible natural convection loops develop in the pool,
ChihiroYanagi used ANSYS FLUENT CFD package to evaluate the air velocity profile
over air-water interface by using free slip boundary conditions over the pool air water
modelled which in later be utilised for defining evaporative heat loss from the air-water
multiple experiments on a small test section made of transparent Plexiglas. The height
and the width of its cross section were 20 mm and 100 mm, respectively, and the test
section length was 1,000 mm. Data acquired by the experiments was used to validate
‘General Ocean Circulation’ model and found to be in good agreement with model
predictions. The natural convection (NC) heat transfer correlation proposed by Kataoka et
al. [8] is used to evaluate NC heat losses from air-water interface by the circulating air in
the containment. Important observations includes, uniform water temperatures within the
pool except inside the regions near the water surface and the bottom.
Tzu-Chen Hung et al. [7] develop three dimensional transient CFD model for predicting
cooling ability of spent fuel pools. This study demonstrates the applicability of porous
model approach for approximating complex fuel assembly geometry. To present the
pressure drop of the flow passing through the fuel assembly, the effective permeability is
calculated from darcy’s law for porous medium, which simplified the fuel region. The
25
effective convection coefficient from air water interface (evaporation + convective loss)
is also modelled based on empirical correlations (Fuji and Imura). This study predicts the
occurrence of local boiling near the exit of high decay power fuel assemblies stored in the
pool, however bulk boiling is not observed. It is found that the location of high decay
power fuel assembly’s significant influences the NC patterns and air water interface
DENOPI was performed by HungvMutelle, with the aim to study the behaviour of spent
fuel pools under loss of cooling and loss of coolant conditions.In order to study boiling
and natural convection flows in pool conditions, scaling analysis has been performed to
determine the dominant terms of the dimensionless equations that describe the whole
physical system. Use of multiphase CFD modeling to have a better understanding of the
different convection loops and boiling conditions and to help in specifying the
modelling works, validation and benchmarking of computer codes. This work mainly
addresses the two phase conditions which may exist in high power fuel assemblies stored
in SFP.
ChihiroYanagi [9], modeled Fukushima Daiici nuclear plants unit 2 and 4 SFP.
Evaporative heat flux correlation derived by ChihiroYanagiin his previous studies [6] was
employed as a boundary condition at the air water interface. Heat and mass transfer
analogy was used to model heat loss due to NC flow of air over SFP air-water interface
(Based on the natural convection heat transfer correlation proposed by Kataoka et al. [8]).
26
Measured data on pool temperature and water loss for the SFPs at the Fukushima Daiichi
nuclear power station after the 2011 accident had been used for validating the models.
The correlation for evaporation heat fluxes, which was derived from the analogy between
turbulent natural convection heat transfer and mass transfer of vapor, overestimates water
evaporation heat fluxes, which were calculated to give good agreement between
calculated and measured values, agreed well with the correlation proposed by Fujii et al
[10] for a limited conditions. Estimated decreasing rate of the water level calculated with
80% decay heat agreed well with measured values. This signifies the overestimation of
decay heat by the ORIGEN 2.2. In view of the above observations, it is imperative to
conclude that evaporative heat flux model develop by ChihiroYanagi could not simulate
Tadashi Fuji, Yoshiyuki Kataoka and MichioMurase [10] evaluate the system pressure
response of a water wall type containment cooling system, which is one of the passive
safety systems (Concept for passive safety systems with no active components have been
investigated for next generation light water reactors). In this study the evaporation and
Two different experimental setup were used to obtain the temperature profiles neat the
heat transfer surfaces and measure the evaporation and condensation heat transfer
coefficients. An essential result of this study includes the comparison of results of two
apparatuses to confirm that the size of heat transfer surface is not affecting the heat
transfer characteristics within these tests. This study proposed experimentally driven heat
transfer coefficient (HTC) correlations for evaporation and condensation. The HTC’s
27
were expressed by the ratio of the steam to non-condensable gas logarithmic mean
concentration, which considered the steam and gas concentration gradient from heat
Naoureddine Boukadida and Sassi Ben Nasrallah [11] explore the validity of heat and
mass transfer (HT-MT) analogy applicable to evaporative heat and mass loss scenarios.
This study is focused to quantify the effect of radiation heat loss from the air-water
surface and its effects on the application of heat and mass transfer analogy. Essential
outcome from this study highlights the conditional applicability of HT-MT analogy only
at low free stream temperatures and vapor concentrations. Study demonstrates the effect
In the wake of Fukushima a number of simple models were proposed to simulate SFP
evaporation and warm-up, particularly Daiichi Unit 4 (SFP4) because it was so heavily
loaded with spent fuel. All of these models were based on some elements of heat and/or
mass transfer theory, some without consideration of high rate mass transfer and some
with .Some of them relied on calibration constants from outside heat- and mass-transfer
theory and some did not. In spite of being based on heat and mass-transfer theory, these
Bruce Robert Hugo [12] develops new evaporation model based on diffusion mass
compared with other correlations for evaporation under conditions of high water
temperature and forced air flow. Analysis of the Fukushima Daiichi Unit 4 SFP response
to a prolonged loss of cooling had provided the only high temperature evaporation data
28
with forced air flow conditions present. The diffusion based evaporation model is
[13]. M.M Shah [13] with aim to develop and validate robust evaporation correlation
from many types of water pools and vessels, discussed the basis of the development of
mass/heat flux from indoor and outdoor swimming pools (occupied and unoccupied),
spent nuclear fuel pools, decorative pools, water tanks, and spills. Evaporation model
developed by M.M Shah [13] holds good for un-occupied swimming pools but shows
deviation from high temperature evaporation data with forced air flow measured at
M. Quinn Brewster [14] review previous literature available evaporation models and
develop a new, robust heat and mass-transfer model for warm water pools in air under
turbulent natural convection flow applicable to the high mass-transfer rate regime. M.
Quinn Brewster [14] reviewed many models like Mc-Adams, Shahs, Horizontal natural
convection model (HNC), Stagnant film model, Hugos model etc.Based on this
comparison, their assessment is that the HNC model and Hugo’s model come closest to
modeling the actual flow. Bruce R. Hugo and William C. Kinsel [12] developed diffusion
evaporative mass flux data generated by the Boelter et al [15] is fitted in this model for
the estimation of calibration constants. Boelter et al [15] generated evaporative mass flux
versus temperature data from a small vertical electrically heated cylinder filled with water
in the stagnant atmosphere. Hugos [12] refined his model by using evaporative mass flux
29
data measurements from the SFP at the Columbia Generating Station nuclear power plant
(CGS).
evaporation rate depends on pool air water interface temperature. Thus thermal
hydraulics of SFP plays important role in the evolution of the air-water surface
temperature which directly influence the mass and heat transfer through concentration
Other available evaporative mass flux correlations [16] and their limitations reported in
2011 for use in designing heating, ventilation, and air conditioning equipment HVAC for
Hugo [12] used this equation at Columbia generation station to estimate evaporation
losses from the SFP but results appeared to under predict the experienced evaporation
rate. Additionally, the air velocity profile over the SFP is not usually known and is non-
evaporation from pre occupied swimming pools through the use of activity factors; newer
studies such as by Shah 2002 have shown that for unoccupied pools carrier correlation
over-predicts the evaporation rate at low air velocities such as those typical of SFPs.
Since the Carrier correlation was developed for swimming pools for which a water
30
40°C to 50°C and the even higher SFP temperatures that could be reached during a
The Boelter correlation (Boelter, Hordon, & Griffin, 1946) for evaporation from a calm
water surface in still air is considered here because it was published only with the high
This equation is applicable for water temperatures above 27°C; the complete correlation
includes two other relations for water temperature ranges well below those typical of
reactor SFPs. The evaporation data published by Boelter et al diverge substantially from
above a water temperature of 88°C; this deviation was attributed due to the phenomena
associated with boiling. But the temperature range of 88°C to 100°C is applicable when
evaluating SFP evaporation under severe accident conditions such as occurred at the
rate. Eqn. 9 was developed using an analogy between heat and mass transfer; it assumes
that natural convection is the dominant mechanism for water vapor transport.
31
Shah had made an extensive comparison of evaporation correlations with published
evaporation data (Shah 2002). Eleven empirical correlations were evaluated including the
Carrier and Boelter correlations discussed early, and an earlier version of Shah’s
correlation. Shah compared these correlations with 11 sets of published evaporation data.
With the exception of the Boelter data (Boelter, Hordon, & Griffin, 1946), none of these
data sets included evaporation rates for high water temperatures above 33.4°C. The Shah
correlation was shown to provide the best overall fit to this body of published data, with
the Boelter correlation providing the second best fit. Accordingly, the Shah correlation is
used as the “benchmark” for evaluating the proposed diffusion based evaporation
correlation.
The analogy between heat and mass transfer limits down the application of Shah
Correlation to SFP applications. Shah does recommend a correlation that includes the
effect of air velocities above 0.15 m/s to allow for calculation of evaporation rates for
outdoor swimming pools. Shah also provides additional corrections for occupied
Hugo’s developed its evaporation correlation based on diffusion mass transfer theory.
his model by using evaporative mass flux data bases on the makeup water requirements
from the SFP at the Columbia Generating Station nuclear power plant (CGS).
32
qr,s qc,s m’’ s
qc,u
u
qc,u = m"hfg+qr,s+qc,s
Where, Decay heat power = qc,uW.D. W.D = Cross section area of the interface, qc,u is
decay heat under quasi steady state, qr,u is radiative heat loss, qc,s convective heat loss, m"
pressure of water in air-water film at interface, P1,eis partial pressure of water in air. It is
essential to note that Hugo’s refined his model by using evaporative mass flux data bases
on the makeup water requirements from the SFP at the Columbia Generating Station
nuclear power plant (CGS). Hugo’s back estimated the evaporation rate from the make-
up water requirement at CGS-NPP and also measured the tritium activity in the SFP
building over the course of few months. Hugo’s also quantifies the amount of heat loss by
conduction and convention. But in all of his estimations the SFP is in normal operating
condition with operational cooling system. It is also imperative to note that the decay heat
33
generated by the discharged fuel assemblies stored within the pool is not very high (as
compared to full core dump scenarios) moreover pool water cooling system maintains
constant level of overboard water above the fuel storage racks. Thus Hugo’s studies are
not able to capture the effect of falling liquid level in the pool and high decay heat on the
surface evaporation. In this context it is essential to conclude that the Hugo’s model is
only applicable when overboard water level above fuel assemblies is significant.
34
2.4 SUMMARY OF EVAPORATION CORRELATIONS
It is a recommended equation
considered in the
35
developed correlation
benchmarked with
E = 0.00005(Pw− Pa)
only low
Not applicable to SFP
temperature (33.4oC)
applications as
evaporation data.
Developed using an analogy
Shah correlation is
Shah correlation
between heat and mass transfer
Correlation not applicable under
may under-predict
High water temperature
the evaporation rate
scenarios
at high water
temperature.
36
flux data based on prolonged surface
Columbia
Generating Station
37
2.5 GAP AREAS
In the pursuance of literature survey, few areas in the pre-uncovery phase, for which further
proposed to simulate SFP evaporation and heat up, particularly Daiichi Unit 4 (SFP4).
All of these evaporation models were based on some form of heat and/or mass transfer
theory and further some with and without consideration of high rate mass transfer regime.
In spite of being based on common heat and mass-transfer theory, these various
evaporation models predict a wide variety of results for SFP4. Therefore in order to
perform realistic thermal hydraulic analysis of SFP under ESBO condition, precise
2. Most of the evaporation models are calibrated and validated either by using high
generated by Boelter et al. For best estimation purpose, benchmarking and validation of
measured and in house available high temperature evaporative heat flux data.
3. Literature available evaporation data relevant to SFP application is only available for
scenarios where overboard water level above racks is significant. Effect of falling liquid
level and presence of high decay heat source on the surface evaporation is not explored
extensively.
38
2.5.2 MULTI-DIMENSIONAL INTERACTION OF DIFFERENT
1. This phenomenon includes transfer of heat, mass and momentum between buoyant
plumes and stagnant fluid zones within the SFP, local erosion of thermally stratified
zones, and interaction of adjacent thermal plumes, whose interests stand, most of all, in
the prediction of liquid local boundary conditions near the air water interface for an
2. The natural circulation loops that develop in the SFP will depend, in large, on the non-
uniform water heating caused by the distribution of fuel assemblies in the pool based on
their decay power and fluid dynamic interaction between zones with different water
temperature. The interaction is complex and properly scaled experiments are needed to
codes.
1. In case of BWR/PWR/VVER fuel assemblies with closed cell storage rack designs,
where the cells can be viewed as parallel and independently heated vertical shrouds that
2. Natural circulation flow may become unstable in this configuration, and flow reversal
may possible in rack cells stored with low-power fuel assemblies. The main safety
concern is that these flow instabilities may leads to local boiling in rack cells with
39
perturbed flow, which may potentially lead to loss of sub criticality by local void
generation.
3. This phenomenon is qualitatively well known from heat exchangers and other nuclear
reactor applications but the knowledge level for it is very low for flow geometries
specific to spent fuel storage racks and conditions expected for SFP accidents.
2.6 APPROACH
40
Tier 2 Scaling analysis of natural convection loops
41
2.7 SUMMARY
This chapter covers the literature survey on pertaining topic in a condensed form. Both
experimental and analytical work specific to surface evaporation is compiled It was observed
that in the wake of Fukushima accident a number of simple evaporation models were proposed to
simulate SFP evaporation and heat up, particularly Daiichi Unit 4 (SFP4). All of these
evaporation models were based on some form of heat and/or mass transfer theory and further
some with and without consideration of high rate mass transfer regime. In spite of being based
on common heat and mass-transfer theory, these various evaporation models predict a wide
variety of results for SFP4. Therefore in order to perform realistic thermal hydraulic analysis of
SFP under ESBO condition, precise estimation of surface evaporative heat/mass flux is required.
Most of the evaporation models are calibrated and validated either by using high temperature
evaporative mass flux data measured during Fukushima accident or generated by Boelter et al.
For best estimation purpose, benchmarking and validation of existing literature available
evaporative correlations is required by using precisely measured and in house available high
temperature evaporative heat flux data. Literature available evaporation data relevant to SFP
application is only available for scenarios where overboard water level above racks is significant.
Effect of falling liquid level and presence of high decay heat source on the surface evaporation is
42
CHAPTER-3. DESCRIPTION OF SFP THERMAL HYDRUILIC MODEL
AND SIMULATION TOOL
The RELAP5 (Reactivity Excursion and Leak Analysis Program) is a simulation tool that allows
users to model the coupled behaviour of the reactor coolant system and the core for various
operational transients such as anticipated transient without scram, loss of offsite power, loss of
feed water, and loss of flow and postulated accidents that might occur in a nuclear reactor . A
generic modelling approach is used that permits simulating a variety of thermal hydraulic
systems. The code includes many generic component models from which general systems can be
simulated. The component models include pumps, valves, pipe, heat releasing or absorbing
structures, reactor point kinetics, electric heaters, jet pumps, turbines, separators, accumulators,
and control system components. In addition, special process models are included for effects such
as form loss, flow at an abrupt area change, branching, choked flow, boron tracking, and non-
This section covers thermal hydraulic analysis specific for pre-uncovery phase during the loss of
dimensional thermal-hydraulic code RELAP5. It is now known that the pre-uncovery phase is
dominated by thermal-hydraulic phenomena, FA stored vertically in the racks and lateral or cross
flow is only permitted in water volume present above and below the racks. Geometric details and
normal operating conditions of VVER 1000 At-reactor SFP are tabulated in the Table-2.
43
Table 2 SFP Dimensions and design parameters (Plant scale)
no
NOC, m
5 Types of racks 3
6 Capacity of Compartment -1, number of FAs 214 (77 FA capacity rack ×2+60
FA capacity ×1)
7 Capacity of Compartment -2, number of FAs 368 (77 FA capacity rack ×4+60
FA capacity ×1)
10 Total thermal load consider for present analysis (at 17.14 MW [ANSI 5.1 ]
44
3.3 COMPUTATIONAL DOMAIN
In order to avoid common cause failure or Postulated Initiating Event (PIE) failure, SFP of
VVER-1000 LWR NPP is divided into two compartments with separate cooling and water
purification system. The fuel assemblies (FAs) are stored in vertical orientation in borated
stainless steel racks, submerged in water that provides cooling to FAs as well as biological
protection. The total height of water column is 19m above fuel racks FA racks are located 400
mm above SFP bottom floor level. Cooling water enters the bottom of the rack, rises through
vertical FAs and gets mixed in the upper water column available over the FA racks. Closed loop
cooling water circulation maintains the pool water temperature at 50oC during normal operating
conditions and 70oC during refuelling outage. In order to understand the thermal hydraulics of
pool under ESBO conditions, one such rack completely filled with FAs is modelled by RELAP5
system code.
Different type’s storage rack designs with different storage capacity are used for the storage of
spent LWR fuel inside the SFP. Installed racks are have a closed cell design, in which each fuel
assembly is enclosed in a separate cells with walls made of, borated steel. Since this closed cell
design does not allows lateral or cross-flow across the FA stored in the racks. Lateral or cross
flow can only occurs in water volume present above and below the racks. There is a provision in
the pool to store leaky fuel assemblies in the sealed canisters, vertically oriented in the rack.
45
Table 3 Fuel Assemblies storage rack specifications
2 Length, m 2.73
3 Width, m 1.92
4 Height, m 4.2
5 Pitch, mm 300
46
3.5 NODALISATION OF PLANT SCALE SFP
The structured approach to the SFP elements in RELAP5 is based on a threefold, bottom-up
arrangement. Active and non-active rods are located within the FA which is in turn stored in
channels or racks, and racks are placed in the SFP. Each rack is homogenized. The RELAP5
code approximates the different fuel elements belonging to one fuel rack into one single fuel
element (Riser). Shortcoming in this approach relates to the lack of natural convection between
fuel assemblies belonging to the same rack. Lateral movement of the water is only allowed in the
overboard and lower water volumes. RELAP5 specific model develop for the computational
domain is shown in Figure 10. The Spent fuel pool with internals is modelled using multiple
1. Volume 904: Simulates overboard water volume present above the rack (Upper
plenum).
4. Volume 805: Simulates free spaces around the fuel rack through which water
5. Volume 806: Simulate water volume present below the racks (lower plenum)
47
All volumes are interconnected to the overboard volume of the pipe component by time
dependent junctions. The decay heat generation by the fuel assemblies is modelled as a heat
structure, 801 which is subjected to a constant power. Surface evaporation of pool water from air
water interface is modelled by using time dependent junction. Enthalpy loss due to evaporation
from the overboard volume is modelled by using heat structure, 802. Other time dependent
(a) (b)
Figure 10 (a) Representation of one rack enclosing one FA inside SFP RELAP5 Nodalisation of
48
a. Flow Area = 5.24 m2
49
4. Lower plenum–Volume 806
1000 light water reactor spent fuel pool which receives fuel assemblies discharged from the
reactor during each refuelling outage and has a capacity to accommodate discharged fuel
is also a provision to accommodate entire reactor core during planed reactor pressure vessel
inspection or during emergency conditions. Total thermal load estimated by ANSI 5.1 for
entirely exhausted capacity of typical LWR-SFP (Fuel assemblies from 8 RSDs plus entire
reactor core) considered in the present analysis is 17.14 MW for 582 fuel assemblies.
1. DECAY POWER
To simulate decay heat generated by FAs stored in the rack, present analysis considered
constant power source as the boundary condition. This boundary condition is enabled
by using power versus time card (reactor kinetics equation is not enabled) of RELAP5
50
for 801 heat structure. The net power supplied to the heat structure 801 is 1.86 MW
2. EVAPORATION MODEL
HEAT LOSS
To simulate evaporative heat loss from the overboard water volume, present analysis is
considered variable heat transfer coefficient (HTC) as the left boundary condition (BC)
and very high heat transfer coefficient as a right boundary condition for heat
structure802. Due to high HTC on the right boundary, left boundary of heat structure
802will map similar temperature as of fluid volume to which it is attached (Pipe volume
90405). Hugo’s model is used for the estimation of evaporative mass and heat flux from
the air water interface. Estimated evaporative heat loss as a function of surface
m[kg/sec] = m"Hugo*AMT/3600
qevaporation[J/sec] = m*hfg
qevaporation= haverage*AHT*(Tsurface-Tambient)
MASS LOSS
To simulate mass of water loss due to surface evaporation, time dependent junctions are
connected with every water filled Control Volume (CV) of pipe component 904 (CV
51
volume above racks, mass flow rate through these junction as a function of fluid volume
temperature is used as a boundary condition. Based on the liquid level in each control
volume (measured by liquid void fraction of CV) only one out of five time dependent
junction will actuate at a time to remove mass of water from un-voided control volume.
Due to evaporative mass loss, liquid level in control volume will fall, based on the
liquid void fraction of CV, subsequent junction will actuate and previous junction will
be latch close. In this manner continuous mass loss due to evaporation and resultant fall
This section incorporates the RELAP5 thermal hydraulic simulation results for actual plant scale
spent fuel pool geometry. Variation in the temperature of fluids within the control volumes and
fluid velocities at the junctions with respect to time are presented in the following Figures.
450.00
400.00
350.00
803010000
300.00
803020000
Temperature, K
250.00 803030000
803040000
200.00
803050000
150.00
0.00
0.00 2,000.00 4,000.00 6,000.00 8,000.00 10,000.00 12,000.00
Time, sec
52
Discussion on Figure 11
From the above graph it can be infer that nearly no thermal stratification is observed in the 5
control volumes representing fuel assembly storage rack (803010000 to 803050000).Pool heat-
450.00
400.00
350.00
300.00
Temperature, K
904010000
250.00 904020000
904030000
200.00 904040000
904050000
150.00
0.00
0.00 2,000.00 4,000.00 6,000.00 8,000.00 10,000.00 12,000.00
Time, sec
Discussion on Figure 12
From the above graph it can be infer that the thermal stratification is observed in all five control
volumes representing overboard water volume above rack (904010000 to 904050000). Thermal
stratification in the 904 pipe volume is may be due to very low fluid circulation velocity. From
Figure 14 it can be observed that the fluid circulation velocity at the inlets and discharges of all
pipe volumes (Junctions) is approximately 0.08 m/sec before the inception of boiling (nearly at
5100 sec). This low circulation velocities is accountable for weak natural circulation currents and
53
poor mixing in the overboard volume. It is important to note that RELAP5 is one dimensional
system code therefore cross flow within the control volume cannot be modelled. Exact
circulation pattern can only be observed by employing CFD techniques. Due to the convection
boundary condition representing the heat loss due to surface evaporation, temperature of top
most water filled control volume (904050000) is lowest among all. It is principal to notice that
temperature of the topmost water layer differs from bulk water temperature hence thermal
hydraulics of pool plays important role in energy deposition and the development of surface
water temperature.
450.00
400.00
350.00
805010000
300.00
Temperature, K
805020000
250.00 805030000
200.00 805040000
805050000
150.00
100.00
50.00
0.00
0.00 2,000.00 4,000.00 6,000.00 8,000.00 10,000.00 12,000.00
Time, min
Discussion on Figure 13
54
Single Phase Boiling region
Heat-up region
Velocity,
m/sec
Time, sec
Figure 14 Fluid velocity versus time variation in junctions, 823 discharge of FA rack, 824
discharge of upper plenum, 825 discharge of downcomer, 826 discharge of lower plenum
Discussion on Figure 14
Fluid circulation velocities in all four junctions (823-826) are nearly identical. Prior to the
inception of boiling (up-to 5100 sec), nearly steady fluid velocity of 0.08 m/sec is observed in
all the junctions and highly fluctuating fluid velocities are observed in two phase region. Low
fluid circulation velocities will establish weak natural circulation currents in the SFP. This weak
natural circulation flow within the SFP may become unstable at some regions, such as flow
reversal may be possible in few rack cells with stored with low-power fuel assemblies
3.9 SUMMARY
This chapter covers the pre-test thermal hydraulic analysis specific for pre-uncovery phase
during the loss of cooling accident scenario in VVER 1000 NPP-SFP. Thermal hydraulic
55
specific model is developed for the computational domain. In the computational domain, FA
stored vertically in the racks and lateral or cross-flow is only permitted in water volume present
above and below the racks. To simulate decay heat generated by FAs stored in the rack, constant
power source as the boundary condition is used. Decay heat is estimated by using ANSI 5.1. This
boundary condition is enabled by using power versus time card. To simulate evaporative heat
loss from the overboard water volume, present analysis is considered variable heat transfer
coefficient (HTC) as the left boundary condition (BC) and very high heat transfer coefficient as a
boundary condition for heat structure. To simulate mass of water loss due to surface evaporation,
time dependent junctions are connected with every water filled Control Volume (CV) of pipe
component representing overboard water volume above racks, mass flow rate through these
junction as a function of fluid volume temperature is used as a boundary condition. Based on the
liquid level in each control volume (measured by liquid void fraction of CV) only one out of five
time dependent junction will actuate at a time to remove mass of water from un-voided control
volume. Due to evaporative mass loss, liquid level in control volume will fall, based on the
liquid void fraction of CV, subsequent junction will actuate and previous junction will be latch
close. In this manner continuous mass loss due to evaporation and resultant fall in the water level
in the overboard volume is modelled. Thermal stratification is only observed in the overboard
water volume. Nearly negligible thermal stratification is observed in other volumes.). Steady
fluid velocity of 0.08 m/sec is observed in all the junctions and highly fluctuating fluid velocities
are observed in two phase region. Low fluid circulation velocities will establish weak natural
56
CHAPTER-4. DESIGN OF EXPERIMENTAL SETUP
Similarity laws and scaling criteria is an important technique for the design and development of
scaled test facilities in order to simulate the behaviour observed in prototype plant. Several
constrains like high radiological risk, availability of very limited measurement sensors,
operational and regulatory limitations causes difficulty in assessment of flow and heat transfer in
NPP SFP. In the view of the above mentioned restrictions and for the assessment of flow and
heat transfer phenomena within the NPP SFP, design of one scaled test facility is conceived.
In this section scaling criteria for single phase flow loops have been developed with special
emphasis on natural circulation mode. General scaling laws for modelling NPP systems had been
proposed by Nahavandist al. (1979). In general most of the experimental facilities are design in
such a way that the power to volume ratio remains conserved. The essential requirements of this
scaling philosophy are described by Zuber (1980) and Karwat (1985). The main limitation of this
approach is the conservation of geodetic elevation of the prototype and the test facility, based on
the available power supply to derive the test facility, volume scaling ratio is obtained. As a result
of this test facility has a flow cross-section area scaled by volume ratio. It is recognised that the
reduced loop diameters can result in the distortion of certain physical phenomena observed in the
prototype plant.
It is essential to note that one of the basic problems in understanding the behaviour of
natural convection system is the lack of generally acceptable scaling laws which are non-loop
specific (Zirvin, 1981a). Initial attempt to develop scaling laws for natural convection loops was
made by Heisler (1982) and later on by Ishii and Kataoka (1984). As mentioned earlier,
57
significant limitations of power to volume scaling approach such as, required conservation of
actual natural convection circulation height and the effect of loop diameter on scaling distortion
limits its applicability in the scenarios where circulation height is substantial. P. K. Vijayan, H.
Austregesilo (1994) [17] investigates the effect of loop diameter on scaling distortion and
demonstrate through experimental studies is that P/V scaling principal adequately describe only
the steady state behaviour of NC system. P. K. Vijayan (1992) carried out experiments in three
rectangular, natural circulation loops, based on these experiment results and using simple one
dimensional theory as used by Zvirin (1981) [18], it is shown that the scaling laws for single
phase NC loops can be developed. Theoretical investigation of the single phase natural
circulation phenomena showed that the transient and stability behaviour can be simulated only if
the diameter ratio Dp/Dm is also conserved. Developed scaling laws are then tested against
experimental data from various loops and results from such studies upheld the validity of
developed scaling approach Brief description on the derivation of scaling laws and associated
The similarity criteria have been obtained from conservative equations of mass momentum and
energy, boundary conditions, and geometry of a system. For whole natural circulation loop
simulation it is very important to consider not only the local condition but also integral effects
over an entire loop and components therefore it is necessary to include integral balances in the
similarity analysis. From these, the geometrical similarity groups are derived. One dimensional
approach is used in the flowing derivation, where only coordinate runs around the loop. For
58
horizontal heater Horizontal condenser loop configuration, origin is at the beginning of the
(3) Boussinesq approximation is valid, i.e. fluid properties can be treated as a constant in the
transient and convective term but as a variable in the body force term.
(12)
i.e. mass flow rate in the loop is independent of space coordinate and is only a function of time.
Using this and Boussinesq approximation integral momentum equation can be written as
∮ (13)
In writing the above equation, it is assumed that the form losses are negligible as compared to the
wall shear. The effect of form losses (if any) can be incorporated in the above equation by using
(14)
Where the constants p and b taken different values for laminar and turbulent flow. For example,
p = 64 and b=1 for laminar flow, while p and b are respectively 0.316 and 0.25 for turbulent flow
(Blasius equation)
59
(Heater Section) (15)
The momentum and energy equations are then non-dimensionalized using the flowing
substitutions:
, , ,τ ,
With these substitutions, non-dimensional momentum and energy equation can be obtained as
∮ (18)
= (19)
(20)
(21)
It is imperative to note that above mentioned non dimensional groups are modified and differ
somewhat from their classical definitions. The term QH/(ACpμ) has a dimensions of temperature
For the steady state conditions, , therefore ωss=1, the steady state temperature
60
θss (S) = θcl+ (Heater) (22)
Using these equations, the integral in equation 18 can be evaluated and it can be shown that
∮ for the natural circulation loop considered in the present derivation. Using this
result in equation 18, the following equation for the flow rate under steady state conditions can
be obtained.
For the above development it is imperative to note that steady state Reynolds number data
plotted against the product of Grm(D/L) on a log-log graph falls on a straight line with slope
equals to r and intercept equals to C. Knowing C and r, the constants p and b can be evaluated as
b = 3- , p = . These equations can be used to obtain an empirical correlation for the friction
factor of natural circulation flows. For simulating the transient and stability behaviour of natural
circulation flows, the coefficients of the equations 18-21 must be same. Thus for simulating
(24)
(25)
[∮ ] (26)
( )
61
In the view of above developments, for simulating the transient behaviour of the momentum
steady state behaviour needs the simulation of the product Grm(D/L).Therefore in order to
simulate the steady state and transient behaviour stability behaviour of natural circulation
( ) ( )
( ) ( )
( ) ( )
( ) ( )
In compiled formation
(27)
For constant P/V scaling approach, equation 27 states that, if Hp = Hm, then the transient and
stability behaviour can be simulated only if Lp = Lm, (Lh)p=(Lh)m, and Dp= Dm. As mentioned
earlier, in case of P/V scaling approach (Dp/Dm) ≠ 1. Hence scaled test loops based on the P/V
62
4.3 SINGLE PHASE SIMILARITY LAWS FOR NATURAL CONVECTION
BY M. ISHII AND KATAOKA [Source: Ref [19]].
The similarity criteria for a natural circulation system can be obtained from the integral effects of
the local balance equations along the entire loop. For a single phase flow case, a method similar
to that is developed by Heisler&Singer [20] and Heisler [21] for a liquid metal system is applied
to develop similarity criteria. There are three equations which are governs for the natural
convection loop: the fluid energy equation; the solid energy (heater rod or fuel rod) equation;
and, the fluid momentum equation. Each conservation equation is derived in the fashion as
recommended by M. Ishii and Kataoka [19]. The equations are normalized and the terms are
divided by the “driver term” such that the resulting groups are dimensionless.
In the following derivation, subscripts o and r denote the reference constant value and
representative variable of a system. The component and solid are denoted by subscripts i and
s. Using the Boussinesq assumption for a single phase natural convection system, the fluid is
considered incompressible, except in the gravitational term in the momentum equation. Then the
conservation laws can be expressed by the following set of simplified balance equations based on
a one-dimensional formulation.
Continuity equation
(28)
(29)
∑ ∑
63
Fluid energy equation for section
(30)
( )
(31)
̇
=h( (32)
In the above equations, ur is the representative velocity of the system corresponding to the
velocity of the section having cross sectional area ao, .lh is the equivalent total length of the hot
fluid section. The above set of equations can be non-dimensionalized by introducing the
, , , , , , , , ,
Continuity equation
(33)
64
Momentum balance equation
(34)
(∑ ) ∑( )
(35)
(36)
Fluid-solid boundary condition (
(37)
The similarity groups appeared in the above equations and their significance are defined below.
65
Biot number Wall convection / Conduction
In addition to the above defined physical similarity groups, several other geometrical similarity
Axial scale; ,
It is noted here that the hydraulic diameter di and the conduction depth are defined by
And
Where ai, asi and are the flow area, solid cross sectional area and wetted perimeter of ith section,
In the above development, the reference scales for the velocity and temperature change have
been used. The simplest way to obtain these scales is to use the steady state solutions. By taking
the heated section as a representative section, following solutions of temperature rise has been
obtained
66
Where subscript 0 denotes the heated section. Substituting the above expression into steady state
( ) ( )
(∑ )
fuel rack simulator section. Fuel rack simulator consists of 30vertical immersive U-tube heaters
surrounded by a cylindrical shroud to simulate decay power and flow area of 60 fuel assemblies.
Facility is equipped with an auto transformer and, input power metering device for heater power
control, multiple thermocouples for measuring fluid temperature profile, DP transmitter for level
measurement and emergency trip-protection systems. All sensors are hard wired to one Data
Acquisition System (DAS) with inbuilt data recorder and monitoring system. Detailed
67
Vertical Outer tank
SS tubes
simulating
FAs
Inner tank
Perspex
front
= = 3.50
In order to achieve desired hydraulic diameter for the experimental model, heater rod OD is
68
Where ΔTo is steady state temperature difference of the fluid in between inlet and outlet
reference section
Table 4Comparison of Non-dimensional numbers for plant scale and experimental scale
assembly, ao
69
8 Fluid velocity in the FA rack 7.22 x 10-2 m/sec 4 x 10-2m/sec
(RELAP5 Output)
(Tw+Tinfinity)/2
10 ρ, Cp, β at mean film 944 Kg/m3, 4.24 958 Kg/m3, 4.21 KJ/kgK,
temperature and P= 3 bar KJ/kgK, 0.00273 C-1 0.00075057 C-1 at P= 1.3 bar
As mentioned earlier in the section 4.2 and 4.3 conservation of non-dimensional numbers
specific to natural circulation loop scaling such as Richardson number, Stanton number is a
necessary requirement. It is imperative to note that for total input power of 3.3KW, Richardson
number, Stanton number remains conserved for the plant and experimental scale, i.e.
phenomena occurring in the actual At-reactor SFP with average decay power of 1.86MW per
rack.
70
4.6 DESIGN OF EXPERIMENTAL SETUP
The experimental setup consists of a flat bottom outer tank (SS 316) having dimensions of800
mm (L) X 510 mm (W) X 3000 mm (D) with wall thickness of 5mm. The outer tank encloses a
fuel assembly rack simulator made up of SS 316. Fuel assembly rack simulator section is having
dimensions of 660 mm (L) X 420 mm (W) X 1000 mm (D). Capacity of FA rack simulator is to
accommodate 60 fuel assemblies in vertical position. FAs are modelled using 60 vertical SS
base plate. Facility is equipped with 30 Immersive U-tube heaters enclosed within the cylindrical
shrouds. Each of 30 immersive U-tube heaters is mounted on the outer tank base plate via gland
sealed assembly. Heater mounting gland seal assembly is design in such a way to facilitate heater
removal and heater maintenance. One auto transformer is connected with the power distribution
bus of immersive heaters to regulate the heater power. Power variation is required simulate the
decay heat generated by each fuel assembly. Outer tank is filled with DM water and the water
temperature and level is continuously monitored using K type thermocouples and differential
pressure transmitter. In addition to tanks, piping and valves for filling and drainage are arranged
in required manner. Detailed placement locations of the thermocouples within the experimental
setup is shown in Figure 17 (b) and Table- 6. Entire set of measuring equipment’s are connected
A. Selection of Heater:
B. Number of Heaters:
71
The present analysis considers the capacity of 60 fuel assemblies per rack. One
Heater Specification:
enumerated below:
60 mm (Horizontal)
4 Bend radius 42 mm
5 Wattage 1000W
Amps
72
3000 mm
Inner Tank
800 mm
(a)
73
Heater Pin
Heater
Shroud
(b)
74
(c)
Figure 16 Experimental setup (a) Outer tank section (b) Inner Tank section (c) Immersive U-tube
heater
C. Instrumentation Specification-Thermocouple:
Requirement of Thermocouple:
fluid volume. Quantification of heat flux is done by first allowing the fluid
volume to attain pseudo steady state and then applying the 1-D conduction
(DAS)
75
Specification of Thermocouple:
The Specification of thermocouple used in experiment are K-type thermocouple with SS316
sheathed, 0.5mm tip diameter, 2m long. Simplex thermocouple with mineral insulation
Outer
Tank Plane-1 Thermocouple
Planes
400 mm
Plane-2
400 mm
Plane-3
400 mm
Plane-4
3000 mm
400 mm
Fuel Rack
Simulator 1000 mm
200 mm
(a)
76
(b)
Figure 17 Thermocouple Location in the setup (a) Location Planes (b) Sensor ID/location
77
3 Temperature Plane-3 (Diagonally located) 5 10 to 15
measurement meter
Level Measurement:
Measurement of water level in the tank is required to keep a track on the rate of level fall
due to surface evaporation of water from air water interface. Inventory tracking method is
used to estimate evaporative mass flux from the air water interface. To furnish the
78
requirement, a Differential Pressure Transmitter is used which measures the gauge
1. HART Compatible
This section covers the RELAP5 thermal hydraulic analysis of the experimental setup operating
under conditions carefully chosen on the basis of non-dimensionalized scaling performed before.
Thermal Hydraulic analysis is performed in a similar manner as done in Chapter-5 but with
different flow geometry and boundary conditions pertinent to experimental setup and non-
79
Pin length 3.5 m 1m
This section covers the thermal hydraulic simulation results of scaled experimental setup
operated at 3.2 KW. Variation in the temperature of fluids within the control volumes and fluid
velocities at the junctions with respect to time is illustrated in the following figures:
80
8030000
390.00
803010000 803020000 803030000
380.00
803040000 803050000
370.00
Temperature, K
360.00
350.00
320.00
0 5000 10000 15000 20000 25000
Time, Sec
Low thermal stratification is observed in the 5 control volumes representing scaled fuel assembly
storage rack in the experimental setup. (803010000 to 803050000). It can be inferred that the
heat-up time of water to attain saturation in scaled experimental setup is nearly 16000 sec.
81
9040000
390.00
904010000 904020000 904030000 904040000 905050000
380.00 Single phase Two phase
370.00 heat-up region region
Temperature, K
360.00
350.00
340.00
330.00
320.00
310.00
300.00
0 5000 10000 15000 20000 25000
Time, sec
Figure 19 Temperature versus time variation of fluid in scaled overboard water volume
Discussion on Figure 19
Thermal stratification is observed in all five control volumes representing overboard water
volume above rack (904010000 to 904050000). Thermal stratification in the 904 pipe volume is
may be due to very low fluid circulation velocity. Low circulation velocities is accountable for
weak natural circulation currents and poor mixing in the overboard volume. It is important to
note that RELAP5 is one dimensional system code therefore cross flow within the control
volume cannot be modelled. . Exact circulation pattern can only be observed by employing CFD
techniques. Due to the convection boundary condition representing the heat loss due to surface
evaporation, temperature of top most water filled control volume (904050000) is lowest among
all.
82
8050000
380.00
370.00
Single phase Two phase
heat-up region region
Temperature,K
360.00
350.00
330.00
805040000 805050000
320.00
0 5000 10000 15000 20000 25000
Time,Sec
Figure 20 Temperature versus time variation of fluid in scaled downcomer water volume
Discussion on Figure 20
83
8060000
380
370
Single phase Two phase
heat-up region region
360
Temperature, K
350
320
0 5000 10000 15000 20000 25000
Time,sec
Figure 21 Temperature versus time variation of fluid in scaled lower plenum water volume
Discussion on Figure 21
Nearly, no thermal stratification is observed in the 5 control volumes representing scaled lower
84
ELECTRICAL
PROTECTION
SWITICHGEAR
AND POWER
METERING
PANNEL
IMMERSIVE
U-TUBE
AUTO
HEATER
TRANSFORMER
POWER
DISTRIBUTION
PANEL
(COPPER BUS
BARS)
DRAIN
LINES
(a)
85
(b)
Figure 22 Experimental setup for flow and heat transfer assessment of SFP (a) Electrical and
Piping schematic (b) Final Photographs of experimental setup
4.9 SUMMARY
This chapter covers the scaling criteria for single phase flow loops with special emphasis on
natural circulation mode. Similarity laws and scaling criteria is an important technique for the
design and development of scaled test facilities in order to simulate the behaviour observed in
86
prototype plant. Several constrains like high radiological risk, availability of very limited
flow and heat transfer in NPP SFP. In the view of the above mentioned restrictions and for the
assessment of flow and heat transfer phenomena within the NPP SFP, design of one scaled test
facility is conceived. This chapter covers the limitations of power to volume scaling approach
and also includes the derivation of scaling laws and associated mathematical procedure published
by two different authors P. K. Vijayan, H. Austregesilo (1994) and M. Ishii and I. Kataoka in a
(1994),it is observed that the steady state Reynolds number data when plotted against the product
of Grm(D/L) on a log-log graph falls on a straight line and importantly, for simulating the
simulated, whereas simulation of the steady state behaviour requires only the simulation of the
product Grm(D/L). Based on the mathematical derivation scaling ratios specific to NC system is
derived and later validated experimentally. M. Ishii and I. Kataoka proposed the conservation of
Richardson number and Stanton number specific to NC system. Design of the test facility is
mentioned earlier. This chapter also covers the design aspects of the experimental setup such as,
experimental facility is performed for the verification of scaling laws. From the simulation
observed.
87
CHAPTER-5. EXPERIMENT FOR FLOW AND HEAT TRANSFER
ASSESSMENT OF SFP
In total six set of experiments with two different liquid levels and three different heater powers
are conducted. Procedure for conducting experiment is similar for all set of experiments. Initially
outer tank is filled up to the desired liquid level and desired heater power level is set on the auto-
transformer. Heater power is reconfirmed with installed power meter in the electrical distribution
panel. After initiating the experiment at the desired operating parameters, system is set to run
continuously for days until boiling temperature is achieved. Measurements from all the sensors
are logged automatically by the Data logger at the frequency of 1min. As the experimentation
facility is installed in a well-ventilated shed, Relative humidity data from the Indian Metrological
department Thane Maharashtra measuring centre is also logged at regular intervals during the
& 9.
88
6 2022 3.33 Temperature and level
1 2794 0.59
2 2763 0.27
3 2767 0.46
4 2046 0.23
5 2000 0.28
6 2022 0.18
Operating parameters of experimental setup during 1st run are tabulated in Table-10. Sensors
Identification tag number and their mounting locations within the experimental setup are
2 Power level 10 KW
89
3 Initial Relative Humidity 0.59
TE01
100
TE02
TE03
TE04
90
TE05
TE06
80 TE07
TE08
70 TE09
Temperature C
o
TE10
60 TE11
TE12
50 TE13
TE14
40 TE15
TE16
TE17
30
TE18
TE19
20
TE20
0 100 200 300 400 500 600 700
Time, min
90
100
TE22
TE23
TE24
80 TE26
TE27
Temperature, C
TE33
TE34
60 TE35
TE36
TE37
TE38
40
20
0 100 200 300 400 500 600 700
Time, min
Level
2800
2780
2760
2740
2720
Level,mm
2700
2680
2660
2640
2620
2600
100 200 300 400 500 600 700
Time, min
91
25
Experimental
kg/m2 Hr
20
Evaporatuion rate, kg/m2hr
15
10
0
15 115 215 315 415 515 615 715
Time, min
Experimental 10KW
20
Evaporation Flux, kg/m2hr
15
10
0
0 10 20 30 40 50 60 70 80 90 100
Temperature, C
Figure 27 Evaporation flux versus average top plane temperature, Lt=2794mm Pt=10KW
92
Plane Temperature Profile Plane location
number
1
40
39
38
37
36
35
Temperatures, C
34
33
32
31
30
29
28
27
26
25 5
1 4
2 n
3 tio
Se
ns o 3 o ca
r Lo 2 rL
cat 4
n so
ion
5 1 Se
(a)
2
40
39
38
37
36
35
34
Temperature, C
33
32
31
30
29
28
27
26
25 5
1 4
2
3 n
Se 3 tio
ns o 2 o ca
r Lo 4 rL
c at 1 n so
ion 5 Se
(b)
93
3
40
39
38
37
36
35
Temperature, C 34
33
32
31
30
29
28
27
26
25 5
1 4
2 on
3
3
c ati
Se Lo
ns o 2 r
r Lo 4
n so
cat 1 Se
ion 5
(c)
40
39
38
37
36
35
34
Temperature, C
33
32
31
30
29
28
27
26
25 5
1 4
2 on
3
3
c ati
Se Lo
ns o 2 r
r Loc 4 so
n
atio
n 5 1 Se
(d)
94
35.2
12345
35.0 5
4
3
34.8 2
Temperature, C
1
Plane1
34.6 Plane2
Plane3
Plane4
34.4
Heaters
34.2
34.0
1 2 3 4 5
Thermocouple Location
(e)
24 3
33
32 N 4
31 A
30
29
26 5
28
27
26 27 6
25 6
1 5 1 2 3 4 5 6
2 4
3 on
Se 3 ati
ns 4 oc
or L 5 2 o rL
oc a ns
tion 6 1 Se
(f)
95
Heater shroud inlet
Thermocouples 33 to 38
40
39
33 1
38
37
36 34 2
35
34
Temperature, C
35 3
33
32 36 4
31
30
37 5
29
28
27 38 6
26
25 6 1 2 3 4 5 6
1 5
2 4 on
Se
3 3 c ati
ns or L 4 Lo
2 or
o cat 5 ns
ion 6 1 Se
(g)
Figure 28 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2794mm Pt=10KW, at t=100min
96
Plane Temperature Profile Plane location
number
1
62
61
60
59
58
57
56
Temperature, C
55
54
53
52
51
50
49
48
47
46
45 5
1 4
2 n
3 tio
Se 3 o ca
ns o 2 rL
r Lo 4 n so
cat
io 5 1 Se
n
(a)
2
62
61
60
59
58
57
56
Temperature, C
55
54
53
52
51
50
49
48
47
46
45 5
1 4
2 o n
3 ti
3 ca
Se
ns 2 r Lo
or L 4 n so
oc a Se
tion 5 1
(b)
97
3
62
61
60
59
58
57
Temperature, C
56
55
54
53
52
51
50
49
48
47
46
45 5
1 4
2 n
3 tio
Se 3 o ca
ns o 2 rL
r Lo 4 n so
cat
ion 5 1 Se
(c)
4
62
61
60
59
58
57
56
Temperature, C
55
54
53
52
51
50
49
48
47
46
45 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 rL
oc a
4
n so
tion 5 1 Se
(d)
98
59.0
58.8
58.6 12345
58.4
Temperature, C
58.2
58.0 Plane 1
57.8
Plane 2
Plane 3
57.6 Plane 4
57.4
Heaters
57.2
57.0
1 2 3 4 5
Thermocouple Location
(e)
56
55
54 24 3
53
52
51 N 4
50 A
49
48
47 2 5
46 6
45 6
1 5
2 4 on 27 6
3 ati
Se 4
3 Loc
ns or L 2 or
oc a 5 ns 1 2 3 4 5 6
tion 1 Se
6
(f)
99
Heater shroud inlet
62
Thermocouples 33 to 38
61
60 33 1
59
58
57 34 2
56
Temperature, C
55 35 3
54
53
52 36 4
51
50
49 37 5
48
47
46 38 6
45 6
1 5
2 1 2 3 4 5 6
4 on
3 ati
Se 4
3 Loc
ns or L 2 or
oc a 5 ns
tion 1 Se
6
(g)
Figure 29 (a) to (e) Represents local temperature field within the experimental setup operating
with Lt=2794mm Pt=10KW, at t=300min
100
Plane Temperature Profile Plane location
number
1
80
79
78
77
76
75
74
Temperature, C
73
72
71
70
69
68
67
66
65 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(a)
2
80
79
78
77
76
75
74
Temperature, C
73
72
71
70
69
68
67
66
65 5
1 4
2 on
3 ati
Se 3 Loc
ns or L 4
2
s or
oc a n
tion 5 1 Se
(b)
101
3
80
79
78
77
76
75
74
Temperature, C
73
72
71
70
69
68
67
66
65 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(c)
4
80
79
78
77
76
75
74
Temperature, C
73
72
71
70
69
68
67
66
65 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(d)
102
79.2
79.1
79.0
78.9
12345
78.8
Temperature, C
78.7
78.6
78.5 Plane1
Plane2
78.4 Plane3
Plane4
78.3
78.2 Heaters
78.1
1 2 3 4 5
Thermocouple Location
(e)
73 24 3
72
71 N 4
70
A
69
68
67 26 5
66
65 6
1 5 27 6
2 4 on
3 3 c ati
Se 4 Lo 1 2 3 4 5 6
ns or L 2 or
oc a 5 ns
tion 1 Se
6
(f)
103
Heater shroud inlet
Thermocouples 33 to 38
80
79
78 33 1
77
76 34 2
75
74
Temperature, C
35 3
73
72
36 4
71
70
69 37 5
68
67 38 6
66
65 6
1 5 1 2 3 4 5 6
2 4 n
tio
3 3 ca
Se 4 Lo
ns or L 2 or
oc a 5 ns
tion 1 Se
6
(g)
Figure 30 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2794mm Pt=10KW, at t=500min
number and their mounting locations within the experimental setup are enumerated in Figure 17
104
TE01
100 TE02
TE03
TE04
90 TE05
TE06
TE07
80 TE08
Temperature, C
TE09
TE10
70
TE11
TE12
TE13
60
TE14
TE15
50
TE16
TE17
TE18
40 TE19
0 200 400 600 800 1000 TE20
Time, min
DPT01
2780
2760
2740
2720
Level, mm
2700
2680
2660
2640
2620
2600
0 200 400 600 800 1000
Time, min
105
25
20
Evaporation rate, kg/m2hr
15
10
0
51 151 251 351 451 551 651 751 851 951 1051
Time, min
20
Evaporation flux, kg/m2hr
15
10
0
52.5 57.5 62.5 67.5 72.5 77.5 82.5 87.5 92.5 97.5
Temperature C
Figure 34 Evaporation flux versus average top plane temperature, Lt=2763mm Pt=6.66KW
106
Plane Temperature Profile Plane location
number
1
66
65
64
63
62
61
60
Temperature, C
59
58
57
56
55
54
53
52
51 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(a)
2
66
65
64
63
62
61
60
Temperature, C
59
58
57
56
55
54
53
52
51 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
oc a
4 ns
tion 5 1 Se
(b)
107
3
66
65
64
63
62
61
60
Temperature, C
59
58
57
56
55
54
53
52
51 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(c)
4
66
65
64
63
62
61
60
Temperature, C
59
58
57
56
55
54
53
52
51 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(d)
108
64.0
63.5
12345
Temperature, C 63.0
62.5 Plane 1
Plane 2
Plane 3
62.0
Plane 4
61.5 Heaters
61.0
1 2 3 4 5
Thermocouple Location
(e)
24 3
59
58
N 4
57
A
56
55
54 26 5
53
52
27 6
51 6
1 5
2 4 on 1 2 3 4 5 6
3 3 c ati
Se 4 Lo
ns or L 2 or
oc a 5 ns
tion 1 Se
6
(f)
109
Heater shroud inlet
66
Thermocouples 33 to 38
65
64 33 1
63
62 34 2
61
60
Temperature, C
35 3
59
58
57 36 4
56
55 37 5
54
53
52 38 6
51 6
1 5
2 1 2 3 4 5 6
4 on
3 ati
Se 4
3 Loc
ns or L 2 or
oc a 5 ns
tion 1 Se
6
(g)
Figure 35 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2763mm Pt=6.66KW, at t=200min
110
Plane Temperature Profile Plane location
number
1
80
79
78
77
76
75
74
Temperature, C
73
72
71
70
69
68
67
66
65 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(a)
2
80
79
78
77
76
75
Temperature, C
74
73
72
71
70
69
68
67
66
65 5
1 4
2 on
3 c ati
Se 3 Lo
ns 2 r
or L 4 n so
oc a Se
tion 5 1
(b)
111
3
80
79
78
77
76
Temperature, C 75
74
73
72
71
70
69
68
67
66
65 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(c)
4
80
79
78
77
76
75
Temperature, C
74
73
72
71
70
69
68
67
66
65 5
1 4
2 on
3 c ati
Se 3 Lo
ns 2 r
or L 4 n so
oc a Se
tion 5 1
(d)
112
77.4
77.2
77.0 12345
76.8
76.6
Temperature, C
76.4
76.2 Plane 1
Plane 2
76.0
Plane 3
75.8 Plane 4
75.6
75.4 Heaters
75.2
75.0
1 2 3 4 5
Time, min
(e)
74 24 3
73
72
25 4
71
70
69 N 5
68 A
67
66 27 6
65 6
1 5
2 4 n
tio
1 2 3 4 5 6
Se
3 3 o ca
ns or L
4
2 orL
oc a 5 ns
tion 1 Se
6
(f)
113
Heater shroud inlet
Thermocouples 33 to 38
80
79
78 33 1
77
76 34 2
75
74
Temperature, C
35 3
73
72
36 4
71
70
69 37 5
68
67
38 6
66
65 6
1 5 1 2 3 4 5 6
2 4 n
tio
3 3 ca
Se Lo
ns or L 4 2 or
oc a ns
tion 5 1 Se
6
(g)
Figure 36 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2763mm Pt=6.66KW, at t=500min
114
Plane Temperature Profile Plane location
number
1
90
89
88
87
86
Temperature, C
85
84
83
82
81
80
79 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(a)
2
90
89
88
87
86
Temperature, C
85
84
83
82
81
80
79 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(b)
115
3 90
89
88
87
Temperature, C 86
85
84
83
82
81
80
79 5
1 4
2 on
3 c ati
Se 3 Lo
ns or L 2 or
oc a
4 ns
tion 5 1 Se
(c)
4
90
89
88
87
86
Temperature, C
85
84
83
82
81
80
79 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(d)
116
88.6
88.5
88.4
12345
88.3
Temperature, C 88.2
88.1
88.0 Plane 1
87.9
Plane 2
Plane 3
87.8 Plane 4
87.7
87.6 Heaters
87.5
1 2 3 4 5
Thermocouple Location
(e)
85 24 3
84
N 4
83 A
82
81 26 5
80
79 6 27 6
1 5
2 4 n
tio
Se
3 3 o ca 1 2 3 4 5 6
ns or L
4
2 orL
oc a 5 ns
tion 1 Se
6
(f)
117
Heater shroud inlet
Thermocouples 33 to 38
90
89 33 1
88
34 2
87
86
Temperature, C
35 3
85
84 36 4
83
37 5
82
81
38 6
80
79 5 1 2 3 4 5 6
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
oc a Se
tion 5 1
(g)
Figure 37 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2763mm Pt=6.66KW, at t=800min
number and their mounting locations within the experimental setup are enumerated in Figure 17
118
4 Data Logging frequency 1 min
85 TE01
TE02
80 TE03
TE04
75 TE05
TE06
70 TE07
TE08
Temperature, C
TE09
65
TE10
TE11
60 TE12
TE13
55 TE14
TE15
50 TE16
TE17
TE18
45
TE19
TE20
40
500 1000 1500 2000 2500
35 Time, min
30
25
20
Figure 38 Free board liquid level temperature distributions Lt=2767mm Pt=3.33KW
15
DPT01
10
2800
5
2780
2760
2740
Level, mm
2720
2700
2680
2660
2640
2620
0 500 1000 1500 2000 2500
Time, min
119
5
4.5
4
Evaporative flux, kg/m2hr
3.5
2.5
1.5
0.5
0
493 993 1493 1993 2493 2993
Time, min
3.82
Experimental, 3.33KW
3.32
Evaporation flux, kg/m2hr
2.82
2.32
1.82
1.32
0.82
0.32
0 10 20 30 40 50 60 70 80 90
Temperature, C
Figure 41 Evaporation flux versus average top plane temperature, Lt=2767mm Pt=3.33KW
120
90
TE22
80 TE23
TE24
TE25
70
TE26
TE27
Temperature C
60 TE33
TE34
50
TE35
TE36
TE37
40 TE38
30
20
500 1000 1500 2000 2500
Time, min
121
Plane Temperature Profile Plane location
number
1
55
54
53
52
51
50
49
Temperature, C
48
47
46
45
44
43
42
41
40 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(a)
2
55
54
53
52
51
50
49
Temperature, C
48
47
46
45
44
43
42
41
40 5
1 4
2 on
3 c ati
Se 3 Lo
ns o 2 or
r Lo 4 ns
c at
ion 5 1 Se
(b)
122
3
55
54
53
52
51
50
Temperature, C 49
48
47
46
45
44
43
42
41
40 5
1 4
2 on
3 ati
Se 3 oc
ns or L 2 o rL
4 ns
oc a Se
tion 5 1
(c)
4
55
54
53
52
51
Temperature, C
50
49
48
47
46
45
44
43
42
41
40 5
1 4
2 n
3 tio
3 o ca
Se rL
ns 2
or L 4 n so
oc a Se
tion 5 1
(d)
123
52
51.9
51.8
51.7
51.6 c
Temperature, C
c
c
51.5 c
51.4 Plane 1
51.3 Plane 2
Plane 3
51.2
Plane 4
51.1
51 Heaters
50.9
0 1 2 3 4 5 6 7
Thermocouple Location
(e)
48 N 4
47 A
46
45 26 5
44
43
27 6
42
41
40 6 1 2 3 4 5 6
1 5
2 i on
4 at
3 c
Se 3 r Lo
ns o 4 s o
2 n
r Lo
cat 5 Se
ion 6 1
(f)
124
Heater shroud inlet
Thermocouples 33 to 38
55
54 33 1
53
52 34 2
51
50 35 3
49
Temperature, C
48
36 4
47
46
45 37 5
44
43 38 6
42
41 1 2 3 4 5 6
40 6
1 5
n
2 4 tio
3 o ca
Se 3 rL
ns or L
4
n so
2
o cat 5 Se
ion 6 1
(g)
Figure 43 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2767mm Pt=3.33KW, at t=750min
125
Plane Temperature Profile Plane location
number
1
65
64
63
62
61
Temperature, C
60
59
58
57
56
55 5
1 4
2 n
3 tio
Se 3 o ca
ns 2 rL
or L 4 n so
oc a Se
tion 5 1
(a)
2
65
64
63
62
61
Temperature, C
60
59
58
57
56
55 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 o rL
oc a
4 ns
tion 5 1 Se
(b)
126
3
65
64
63
62
61
Temperature, C
60
59
58
57
56
55 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
oc a
4 ns
tion 5 1 Se
(c)
4
65
64
63
62
61
Temperature, C
60
59
58
57
56
55 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
oc a
4 ns
tion 5 1 Se
(d)
127
62.6
62.5
1 2 3 4 5
62.4
Temperature, C
62.3
Plane 1
Plane 2
62.2
Plane 3
62.1 Plane 4
62
Heaters
61.9
0 1 2 3 4 5 6
Thermocouple Location
(e)
64 22 1
63
23 2
62
61 24 3
Temperature, C
60
N 4
59 A
58
26 5
57
56 27 6
55 6
1 5 1 2 3 4 5 6
2 4 tion
3 o ca
3 L
Se
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(f)
128
Heater shroud inlet
Thermocouples 33 to 38
65
64 33 1
63
34 2
62
35 3
Temperature, C
61
60
36 4
59
58 37 5
57
38 6
56
55 6 1 2 3 4 5 6
1 5
2 on
4 ati
3 3 L oc
Se
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(g)
Figure 44 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2767mm, Pt=3.33KW, at t=1250min
129
Plane Temperature Profile Plane location
number
1
79
78
77
76
75
Temperature, C
74
73
72
71
70
69 5
1 4
2 on
3 c ati
Se 3 Lo
ns or L 2 or
4 ns
o c at Se
ion 5 1
(a)
2
79
78
77
76
75
Temperature, C
74
73
72
71
70
69 5
1 4
2 on
3 c ati
Se 3 Lo
ns or L 2 or
4 ns
o c at Se
ion 5 1
(b)
130
3
79
78
77
Temperature, C
76
75
74
73
72
71
70
69 5
1 4
2 n
3 tio
Se 3 o ca
ns or L 2 orL
4 ns
o c at Se
ion 5 1
(c)
4
79
78
77
76
75
Temperature, C
74
73
72
71
70
69 5
1 4
2 on
3 c ati
Se 3 Lo
ns or L 2 or
4 ns
o c at Se
ion 5 1
(d)
131
77
76.9
76.8
Temperature, C
76.7
Plane 1
76.6
Plane 2
76.5 Plane 3
Plane 4
76.4
76.3 Heaters
76.2
0 1 2 3 4 5 6
Thermocouple Location
(e)
78 22 1
77
23 2
76
75 24 3
Temperature, C
74
N 4
73 A
72
26 5
71
70 27 6
69 6
1 5
n 1 2 3 4 5 6
2 4 atio
3 oc
Se 3 rL
ns o 4
2 n so
r Lo
cat 5 Se
ion 6 1
(f)
132
Heater shroud inlet
Thermocouples 33 to 38
79
78 33 1
77
34 2
76
75 35 3
Temperature, C
74
36 4
73
72 37 5
71
38 6
70
69 6 1 2 3 4 5 6
1 5
n
2 4 tio
3 ca
3 Lo
Se
ns o 4 or
2 ns
r Lo
cat 5 Se
ion 6 1
(g)
Figure 45 (a) to (g) Represents local temperature field within the experimental setup operating
with Lt=2767mm, Pt=3.33KW, at t=2250min
133
TE28
90 TE29
TE30
80
70
Temperature C
60
50
40
30
no
uaverage
134
From Figure 46, it is observed that heater 64.5oC
( )
W/m2K
7 Fluid properties at Tmean film = 67C, P= 1.3 β = 0.0005720 oC-1, ρ = 978.9 Kg/m3
bar
8 Richardson number 32
In the view of above results from Figure 21 to Figure 46 for three different input heater power
levels of 3.33, 6.66, 10 KW. From Figure 23, Figure 31 and Figure 38 it can be perceive that
very low thermal stratification is observed in the free board water volume. Local temperature
135
distribution within the pool at different elevations for various time instances are shown in Figure
28, Figure 29 and Figure 30 specific to heater power of 10KW highlights the evolution of the
flow field within the experimental setup with time. Change in the local temperature field at
various parallel locations with time confirms the existence of buoyant water plume circulation
within the setup. From the 3-axis plots (Thermocouple location vs. Temperature), it is imperative
to notice asymmetric (but not skewed) local temperature field during the initial heat-up phase
(t=100min), this observation highlights the probable existence of raising and expanding buoyant
plume in the central region of the experimental setup, out of phase local temperature profiles are
observed in the upper and lower region of the setup, this observation raises the probability of the
existence of two distinct natural convection regions incorporating multiple natural circulation
loops (possibly 2 in number per section). With time progression (t= 300min), observation of
skewed but in phase local temperature variation highlights the probable existence of raising and
expanding buoyant plume, symmetry of the temperature fields enhances the probability of the
existence of single natural circulation plume. At t=500min,out of phase variation and enhanced
skewness in local temperature variation observed among different planes raises the possibility of
existence of 2 natural circulation loops circulating in contrary direction to each other. Thus there
is a chances of the transition from 1 to 2 or multiple natural convection loops within the
experimental setup with time progression for heater power of 10KW (Simulating maximum
Local temperature distribution within the pool at different elevations for various time instances
are enumerated in Figure 35, Figure 36 and Figure 37 specific to heater power of 6.66KW
highlights the evolution of the flow field within the experimental setup with time. During initial
instance of heat up (t=200min),slightly skewed but in phase varying local temperature field in
136
the lower section of experimental setup highlights the probable existence of raising and
expanding buoyant plume. Observation of out of phase local temperature variation in the upper
planes increases the probability of the existence of 2 distinct natural circulating loops. With the
evolution of time (t=500min) in phase variation of the local temperature among different planes
is observed. This observation enhances the probability of the existence of single natural
convection loop within the system. Similar rising and expanding buoyant plume is observed in
In principal 3rd experimental run operating at heater power of 3.33KW simulates the
natural convection phenomena pertains to actual At- reactor SFP completely exhausted with
spent fuel (VVER1000, 582 FAs) generating average decay power of 17.14MW. From Figure
43, Figure 44 and Figure 45 it can be inferred that the local temperature field during the initial
that local temperature variation is in phase among different planes. This observation supports the
probable existence of single raising and expanding buoyant plume in the setup, further in phase
local temperature profile upheld the probable existence of single natural convection loop within
the setup. With the evolution of time (t=1250min) slight disturbance in the local temperature
field is observed but system regain similar behaviour as of initial instances in the latter stage of
experiment.
This presence of buoyant water plumes will enhance fluid mixing within the pool. This
phenomenon includes transfer of heat, mass and momentum between buoyant plumes and
stagnant fluid zones within the SFP, local erosion of thermally stratified zones, and interaction of
adjacent thermal plumes. Buoyant plume circulation will significantly affects the prediction of
liquid local boundary conditions such as surface temperature and surface replenishment rate near
137
the air water interface for an estimation of evaporative mass loss. From Figure 24, rise in local
temperature of water measured at the inlet and outlet of six shrouds shroud-heater assembly for
10KW input power experimental run, confirms the existence of natural circulation loops in the
lower portion of the setup which is responsible for the dissipates of heat generated by the electric
from 3rd experimental run data. It can be observed that experimental setup operating at 3.33KW
not observed. This deviation in the Stanton is due to the difference in between RELAP5
predicted heat transfer coefficient (HTC) and experimentally observed one. RELAP5 prediction
From Figure 26, Figure 27, Figure 33, Figure 34, Figure 40 and Figure 41, it is evident that with
the progression of time evaporative flux continues to increase. It is principal to note that as the
Pressure Transmitter. During the start-up phase initial fall in the evaporation flux is due to the
thermal expansion of the water is also observed but not shown in the plots. Due to the presence
of high liquid level (simulating inception of phase 1 accident in SFP), air water interface is
nearby to the upper end of the experimental setup. Vapour plume generated will continue get
evacuated from the experimental setup end therefore accumulation of the vapour column is not
observed. It is imperative to note that the evaporation flux varies non-linearly with fluid
temperature and sharp increases in the evaporation flux is observed beyond pool temperature of
65oC and 75oC for input heater power of 10KW and 6.66KW. From Figure 40, Figure 41, nearly
138
linear evaporation flux variation with time is observed only for input heater power of 3.33KW.
No high mass flux region is observed for input heater power of 3.33KW.
number and their mounting locations within the experimental setup are enumerated in Figure 17
2 Power level 10 KW
139
100
90
80
Temperature, C
70 TE11
TE12
60 TE13
TE14
50 TE15
TE16
40 TE17
TE18
30 TE19
TE20
20
0 100 200 300 400 500
Time, min
DPT01
2250
2200
2150
2100
Level, mm
2050
2000
1950
1900
1850
1800
100 200 300 400
Time, min
140
40
Experimental
35
kg/m2hr
30
Evaporation flux, kg/m2hr
25
20
15
10
0
76 126 176 226 276 326 376 426 476
Time, min
40
Experimental 10KW
35
30
Evaporation flux, kg/m2hr
25
20
15
10
0
40 50 60 70 80 90
Temperature, C
Figure 50 Evaporation flux versus average top plane temperature Lt=2046mm Pt=10KW
141
Plane Temperature Profile Plane location
number
3 z
56
55
54
53
52
51
50
Temperature, C
49
48
47
46
45
44
43
42
41
40 5
1 4
on
2 ati
3 oc
Se 3
o rL
ns or L 2 ns
oc a
4 Se
tion 5 1
(a)
4
56
55
54
53
52
51
50
Temperature, C
49
48
47
46
45
44
43
42
41
40 5
1 4 n
2 tio
3 o ca
Se 3 rL
ns or L 2 n so
oc a
4 Se
tion 5 1
(b)
142
53.5
53.4
53.3
53.2
Temperature, C
53.1
53
13
52.9 24
52.8
Heaters
52.7
52.6
52.5
0 1 2 3 4 5 6
Thermocouple location
(c)
24 3
49
48
N 4
47
A
46
45
44 26 5
43
42
41 27 6
40 6
1 5 1 2 3 4 5 6
n
2 4 a tio
3 oc
Se 3 rL
ns or L
4
2 n so
o cat 5 Se
ion 6 1
143
(d)
50 35 3
49
48
36 4
47
46
45 37 5
44
43
38 6
42
41
40 6 1 2 3 4 5 6
1 5
2 on
4 ati
3 3 L oc
Se
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(e)
Figure 51 (a) to (e) Represents local temperature field within the experimental setup operating
with Lt=2046mm, Pt=10KW, at t=150min
number
144
3
73
72
71
70
69
68
67
66
Temperature, C
65
64
63
62
61
60
59
58
57
56
55 5
1 4
on
2
3 c ati
3 Lo
Se
ns or
or L 2 ns
o cat
4 Se
ion 5 1
(a)
4
73
72
71
70
69
68
67
66
Temperature, C
65
64
63
62
61
60
59
58
57
56
55 5
1 4
on
2
3 c ati
3 Lo
Se
ns 2 s or
or L n
o cat
4 Se
ion 5 1
(b)
145
71.4
71.2
71
Temperature, C
70.8
31
70.6
42
70.4
Heaters
70.2
70
0 1 2 3 4 5 6
Thermocouple location
(c)
24 3
65
64
63 N 4
62 A
61
60
59 26 5
58
57
56 27 6
55 6
1 5
on
1 2 3 4 5 6
2 4
c ati
3 3 Lo
Se r
ns o 4
2 n so
r Lo
cat 5 Se
ion 6 1
(d)
146
Heater shroud inlet
Thermocouples 33 to 38
73
72
71 33 1
70
69 34 2
68
67
Temperature, C
66 35 3
65
64
63 36 4
62
61
60 37 5
59
58
38 6
57
56
55 6 1 2 3 4 5 6
1 5
n
2 4 tio
3 ca
3 Lo
Se
ns o 4 or
2 ns
r Lo
cat 5 Se
ion 6 1
(e)
Figure 52 (a) to (e) Represents local temperature field within the experimental setup operating
with Lt=2046mm, Pt=10KW, at t=250min
number
147
3
89
88
87
86
85
84
83
82
Temperature, C
81
80
79
78
77
76
75
74
73
72
71
70 5
1 4 n
2 tio
3 o ca
Se 3 rL
ns or L 2 n so
oc a
4 Se
tion 5 1
(a)
4
89
88
87
86
85
84
83
82
Temperature, C
81
80
79
78
77
76
75
74
73
72
71
70 5
1 4 n
2 tio
3 o ca
Se 3 rL
ns or L 2 n so
oc a
4 Se
tion 5 1
(b)
148
86.4
86.2
86
Temperaturwe, C
85.8
85.6
13
85.4
24
85.2
Heaters
85
84.8
0 1 2 3 4 5 6
Thermocouple location
(c)
24 3
81
80
79 N 4
78 A
77
76
75 26 5
74
73
72 27 6
71
70 6
1 5 1 2 3 4 5 6
2 n
4
a tio
3 3 oc
Se L
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(d)
149
Heater shroud inlet
Thermocouples 33 to 38
89
88
87 33 1
86
85 34 2
84
83
82 35 3
Temperature, C
81
80 36 4
79
78
77 37 5
76
75
74 38 6
73
72
71 1 2 3 4 5 6
70 6
1 5
n
2 4 tio
3 o ca
Se 3 rL
ns or L
4
n so
2
o cat 5 Se
ion 6 1
(e)
Figure 53 (a) and (e) Represents local temperature field within the experimental setup operating
with Lt=2046mm, Pt=10KW, at t=350min
number and their mounting locations within the experimental setup are enumerated in Figure 17
150
3 Initial Relative Humidity 0.28
151
100
90
TE11
80
TE12
TE13
70
TE14
Temperature, C
TE15
60
TE16
50
TE17
TE18
40 TE19
TE20
30
20
10
100 200 300 400 500 600 700
Time, min
DPT01
2200
2100
Level, mm
2000
1900
1800
100 200 300 400
Time, min
152
5.7 6TH SET OF EXPERIMENT
Operating parameters of experimental setup are tabulated in Table-16Sensors Identification tag
number and their mounting locations within the experimental setup are enumerated in Figure 17
153
100
95
90
85
80
75
TE11
Temperature, C
70
TE12
65 TE13
60 TE14
55 TE15
50 TE16
45
TE17
TE18
40
TE19
35 TE20
30
25
-200 0 200 400 600 800 1000 1200 1400 1600 1800
Time, min
DPT01
2250
2000
1750
Level, mm
1500
1250
1000
-200 0 200 400 600 800 1000 1200 1400 1600 1800
Time, min
154
25
15
10
0
0 200 400 600 800 1000 1200 1400 1600 1800
Time, min
25
Experimental
20
eEaporation flux, kg/m2hr
15
10
0
0 10 20 30 40 50 60 70 80 90 100
Temperature, C
Figure 59 Evaporation flux versus average top plane temperature Lt=2022mm Pt=3.33KW
155
Plane Temperature Profile Plane location
number
43
42
41
40
Temperature, C
39
38
37
36
35
34 5
1 4
2 ti on
3 o ca
3 L
Sen
2 s or
so n
r Lo
cat
4 Se
ion 5 1
(a)
4
43
42
41
40
Temperature, C
39
38
37
36
35
34 5
1 4
on
2
3 c ati
3 Lo
Se
ns o or
2 ns
r Lo
cat
4 Se
ion 5 1
(b)
156
41.8
41.6
41.4
41.2
Temperature, C
41
40.8
31
40.6
42
40.4
40.2 Heaters
40
39.8
0 1 2 3 4 5 6
Thermocouple location
(c)
42 22 1
41 23 2
40
Temperature, C
24 3
39
38 N 4
A
37
36 26 5
35 27 6
34 6
1 5 1 2 3 4 5 6
n
2 4 a tio
3 oc
Se 3 rL
ns or L
4
2 n so
o cat 5 Se
ion 6 1
(d)
157
Heater shroud inlet
Thermocouples 33 to 38
43
42 33 1
41 34 2
40
35 3
Temperature, C
39
38 36 4
37 37 5
36
38 6
35
34 6 1 2 3 4 5 6
1 5
2 on
4 ati
3 3 L oc
Se
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(e)
Figure 60 (a) to (e) Represents local temperature field within the experimental setup operating
with Lt=2022mm, Pt=3.33KW, at t=100min
number
158
3
56
55
54
53
Temperature, C 52
51
50
49
48
47
46
45
44 5
1 4 n
2 tio
3 o ca
Se 3 rL
ns or L 2 n so
oc a
4 Se
tion 5 1
(a)
4
56
55
54
53
52
Temperature, C
51
50
49
48
47
46
45
44 5
1 4
on
2 ati
3 oc
Se 3
o rL
ns or L 2 ns
oc a
4 Se
tion 5 1
(b)
159
54.4
54.2
54
Temperature, C
53.8
53.6
31
53.4 2
4
53.2 Heaters
53
52.8
0 1 2 3 4 5 6
Thermocouple location
(c)
51 24 3
50
25 4
49
48
N 5
47 A
46
45 27 6
44 6
1 5
n 1 2 3 4 5 6
2 4 atio
3 oc
Se 3 rL
ns o 4
2 n so
r Lo
cat 5 Se
ion 6 1
(d)
160
Heater shroud inlet
Thermocouples 33 to 38
56
55 33 1
54
53 34 2
52
Temperature, C
51 35 3
50
36 4
49
48
37 5
47
46
38 6
45
44 6 1 2 3 4 5 6
1 5
n
2 4 tio
3 ca
3 Lo
Se
ns o 4 or
2 ns
r Lo
cat 5 Se
ion 6 1
(e)
Figure 61 (a) to (e) Represents local temperature field within the experimental setup operating
with Lt=2022mm, Pt=3.33KW, at t=300min
number
161
3
74
73
72
71
70
69
Temperature, C
68
67
66
65
64
63
62
61 5
1 4
on
2
3 c ati
3 Lo
Se
ns or
or L 2 ns
o cat
4 Se
ion 5 1
(a)
4
74
73
72
71
70
69
Temperature, C
68
67
66
65
64
63
62
61 5
1 4
on
2 ati
3 oc
Se 3 rL
ns or L 2 n so
o cat
4 Se
ion 5 1
(b)
162
72.8
72.6
72.4
Temperature, C
72.2
72
3
1
71.8 2
4
71.6
Heaters
71.4
71.2
0 1 2 3 4 5 6
Thermocouple Location
(c)
68
67 N 4
66 A
65
64 26 5
63
62 27 6
61 6
1 5
on
1 2 3 4 5 6
2 4
c ati
3 3 Lo
Se r
ns o 4
2 n so
r Lo
cat 5 Se
ion 6 1
(d)
163
Heater shroud inlet
Thermocouples 33 to 38
74
73 33 1
72
71 34 2
70
69 35 3
Temperature, C
68
67 36 4
66
65 37 5
64
63 38 6
62
61 6 1 2 3 4 5 6
1 5
2 on
4 ati
3 3 L oc
Se
ns 4 or
or L 2 ns
o cat 5 Se
ion 6 1
(e)
Figure 62 (a) to (d) Represents local temperature field within the experimental setup operating
with Lt=2022mm, Pt=3.33KW, at t=700min
In the view of above results from Figure 47 to Figure 62 for three different heater power input
levels of 3.33, 6.66, 10KW. From Figure 47, Figure 54 and Figure 56 it can be perceive that very
low thermal stratification is observed in the free board water volume for different operating
heater powers. As infer from local temperature distribution for various time instances is shown in
Figure 51, Figure 52, and Figure 53 for heater power of 10KW and Figure 60, Figure 61 and
Figure 62 for heater power of 3.33KW, highlights the existence of buoyant water plume
circulation within the tank. It is imperative to note that only two temperature measuring planes
(wire mounted with 5 thermocouples) remain submerged in the water. From local temperature
164
field variation during initial phase of heat-up, specific to heater power of 10KW (Figure 51 to
Figure 53), it can be infer that temperature variation is slightly asymmetric or skewed in nature,
this enhances the probability of the existence of single expanding natural circulation plume rising
in the centre portion of the setup. Possibly this rising plume is then bifurcate into two loops,
descending downwards along the walls. With the time progression (t=250 min), temperature of
both the planes varies in the same phase (both rise and fall simultaneously) and is slightly
asymmetric with respect to the centre, this observation upheld the assumption of rising buoyant
plume in the central region of the setup, and possible decent of the fluid along the walls forming
single natural convection loop within the setup. Similar phenomena is also observed for t=350
min.
From local temperature field variation during initial phase of heat-up, specific to heater
power of 3.33KW (Figure 60 to Figure 62) it can be infer that temperature variation is
asymmetric or skewed in nature but is in phase (Both rise and fall simultaneously), this enhances
the probability of the existence of single expanding natural circulation plume rising in the off
centric portion (1st quadrant approximately) of the setup. Possibly this rising plume is then
descending downward along the walls. With the time progression (t=300 min), Out of phase
temperature variation is observed in the central region of the setup, this observation enhance the
probability two distinct natural convection regions with possibly multiple internal recirculation
loops. With time progression (t=700min) skewed out of phase temperature profile is observed,
this observation enhance the probability two opposite circulating natural convection.
Presence of buoyant water plumes will enhance fluid mixing within the pool. This
phenomenon includes transfer of heat, mass and momentum between buoyant plumes and
stagnant fluid zones within the SFP, local erosion of thermally stratified zones, and interaction of
165
adjacent thermal plumes Buoyant plume circulation will significantly affects the prediction of
liquid local boundary conditions such as surface temperature and surface replenishment rate near
the air water interface for an estimation of evaporative mass loss.Due to un-availability of large
overboard water column depth above the fuel rack simulator (simulating later stages of phase 1
accident in SFP), system quicken attains pseudo steady state. It is principal to note the presence
of the vapour plume column in the unfilled section of the setup. This accumulation is mainly
observed when pool water level is low and pool water temperature is above 70oC. The prominent
reason for this accumulation is the close nature of the pool due to the presence of side walls, only
replenishment of this vapour column from the top surface is possible. During Low air circulation
physically intact containment space under ESBO conditions), replenishment of this vapour
column is not very often therefore with time progression this vapour column gets nearly
saturated. This phenomenon limits the molecular diffusion based mass transfer of water molecule
through air-water interface. Similar phenomena are also observed when containment atmosphere
gets nearly saturated with steam under LOCA conditions. It is principal to note that as the
Pressure Transmitter, initial fall in the evaporation flux is due to the thermal expansion of the
From Figure 49,Figure 50, Figure 58 and Figure 59, it can be infer that the evaporation flux
varies non-linearly with fluid temperature and rapid increases in the evaporative mass flux is
observed beyond pool temperature of 70oC for input heater power of 10KW and 6.66KW. From
Figure 58 and Figure 59, nearly linear evaporation flux variation with time is observed only for
166
input heater power of 3.33KW. No high mass flux region is observed for input heater power of
3.33KW.
5.8 SUMMARY
This chapter covers the experimental work for flow and heat transfer assessment of SFP. During
the course six set of experiments with two different liquid levels and three different heater
powers are performed. Procedure for conducting experiment is similar for all set of experiments.
This chapter includes the data generated during the course of experiments. Processed measured
data in the form of plots of local temperature field within the experiment setup, level fall plots,
evaporative mass flux plots are used for developing the understanding of flow field evolution
with time within the experimental setup. Non dimensional numbers specific to NC scaling such
as Richardson number, Stanton numbers are estimated for the P=3.33KW and Lt= 2767mm case.
whereas deviation in the estimated Stanton number and RELAP5 predictions is observed. This
deviation is due to the over estimation of heat transfer coefficient in the heater shroud section by
RELAP5. RELAP5 algorithm use constant Nusselts number (Nu=4.36, q”=constant) specific to
laminar flow regime in forced convection scenarios for evaluating heat transfer coefficient
whereas in the experimental setup heat transfer is due to natural convection. Circulation velocity
and temperature difference across the heater shroud assembly estimated in the experimental
setup are found to be consistent with the RELAP5 simulation results. Thermal stratification is
not observed for the experiments conducted at low and high heater power. Local temperature
distribution within the setup confirms the existence of circulating buoyant hot water plume. This
buoyant hot water plume circulations are strong enough for a good mixing of pool water.
Evolution from single NC circulation loop into multiple NC circulation loops is also observed.
167
CHAPTER-6. ASSESSMENT OF VARIOUS EVAPORATION
CORRELATIONS AGAINST EXPERIMENTAL DATA
In the pertaining literature on evaporation models, M. Quinn Brewster [14] reviewed many
models like Mc Adams, Shahs, Horizontal natural convection model (HNC), Stagnant film
model, Hugos model etc. Based on this comparison, their assessment is that the HNC model and
Hugo’s model come closest to modeling the actual flow. Bruce R. Hugo, William C. Kinsel [12]
high temperature evaporative mass flux data generated by the Boelter et al [15] is fitted in this
model for the estimation of calibration constants. Boelter et al [15] generate evaporative mass
flux versus temperature data from a small vertical electrically heated cylinder filled with water in
the stagnant atmosphere. Hugos refined his model by using evaporative mass flux data bases on
the makeup water requirements from the SFP at the Columbia Generating Station nuclear power
plant (CGS). Figure 63 enumerates the comparison between estimated evaporation fluxes by
stagnant film model, HNC model against the Boelter experiment data. Comparison results
highlights the applicability of HNC model to predict the actual evaporative mass flux data
168
Figure 63 SF model, HNC model comparison with Boelter experiment data [Source: Ref [14]]
169
Figure 64 compares various literature available evaporation models with the HNC model. From
Figure 64 it can be observed that Hugo’s model estimated evaporation mass flux is in close
agreement with the predictions of HNC model. It is important to note that Hugo’s model has
been refined by using evaporative mass flux data bases on the makeup water requirements from
the SFP at the Columbia Generating Station nuclear power plant (CGS), therefore as per the
pertinent literature on high flux evaporation models applicable to NPP-SFP; Hugo’s model is the
Figure 65 to Figure 68 compares the Hugo’s model predictions against the present high
temperature evaporation mass flux data generated during the experimentation. Figure 65 and
Figure 66 represents the evaporative mass flux measured for the scaled decay power of 10KW
and 2780mm liquid level case. Figure 67 and Figure 68 represents the evaporative mass flux
measured for the scaled decay power of 10KW and 2000mm liquid level case.
170
25
20 Experimental kg/m2 Hr
Evaporation flux, kg/m2hr
Hugoes, kg/m2 hr
15
10
0
16 116 216 316 416 516 616 716
Time, min
Figure 65 Comparison of Hugo’s model against present experimental data for 10KW heater
power and 2794mm initial liquid level
25
Experimental 10KW
20
Hugoes
Evaporation Flux, kg/m2hr
15
10
0
0 10 20 30 40 50 60 70 80 90 100
Temperature, C
Figure 66 Comparison of Hugo’s model against present experimental data (evaporation flux
versus average top plane temperature) for 10 KW heater power and 2794mm initial liquid level
171
40
35 Experimental
kg/m2hr
30
Evaporation flux, kg/m2hr
Hugoes, kg/m2 hr
25
20
15
10
0
76 126 176 226 276 326 376 426
Time, min
Figure 67 Comparison of Hugo’s model against experimental data for 10KW heater power and
2046mm initial liquid level
40
35
Experimental
10KW
30
Evaporation flux, Kg/m2hr
Hugoes corelation
25
20
15
10
0
40 50 60 70 80 90 100
Temperature, C
Figure 68 Comparison of Hugo’s model against present experimental data (Evaporation flux
versus average top plane temperature) for 10KW heater power and 2046mm initial liquid level
172
Figure 65 to Figure 68 enumerates the shortcoming of Hugo’s model to predict the evaporation
mass flux correctly for high heater input power cases. Heater input power of 10KW represents
the peak decay power that can experienced by a rack during fresh core dump scenario.
Significant deviations are observed in between the estimated and measured value of evaporation
equation, based on the pertaining literature [16], the use of the analytical solution to Fick’s
equation and boundary layer theory is a typical approach for modelling evaporation. Previously
published evaporation correlations have been either purely empirical or are based on the process
Hugo’s model is developed on diffusion theory as a foundation and its parameters are then
refined by using evaporative mass flux data bases on the makeup water requirements from the
SFP at the Columbia Generating Station nuclear power plant (CGS). Hugo’s correlation required
surface temperature and partial pressure of water on the surface and in free stream as inputs.
Effect of the flow development and heat transfer mechanism within the pool is not captured in
the correlation. It is important to note that air- water interface temperature is strongly depends on
the pool dynamics and decay power of the stored fuel. Based on the limited applicability
information and shortcomings of Hugo’s correlation to predict evaporation mass flux under high
power scenarios, development of two correlations for distinct power and initial liquid levels has
been taken up in this present work. On the basis of experimental observations and phenomena
involved, two different approaches namely power based and diffusion based, were employed for
173
6.1.1 POWER BASED APPROACH
Several principal variables considered in the present development of evaporation correlation are
following:
2. As per the experimental observations, measured value of the evaporation rate is higher
than the values predicted by the molecular diffusion based correlations. This discrepancy
is due to the presence other controlling factors such as surface renewal rate, surface
turbulence (chaotic nature) etc. Thermal plume movement is responsible for the
that the thermal plume movement is implicitly linked with the heater power. To account
this augmentation of evaporation rate due to many implicit factors indirectly linked with
dimensional decay power under high decay power scenario. Here Qtherotical is the
From the data generated during1st, 4th, 5th, 6th set of experimental runs it is inferred that the decay
power of the fuel strongly determines the surface conditions such as temperature and surface
replenishment rates on the air water interface which ultimately affect the evaporation rates.
Regression analysis of experimental evaporation data generated for different power and liquid
levels under various ambient conditions highlights the strong effect of non-dimensional decay
174
power on the evaporation rates. It is principal to note that the phenomena involved in the 1st set
of experimental run is different from the 4th 5th and 6th set. In latter cases height of overboard
water column is low (2000mm) as compared to the former case (2780 mm). Due to the low
liquid column height and presence of significant depth of vapour column in latter stages of
experiment, system is expected attains pseudo steady state. In the 1st set of experiment the air
water interface is close to the physical end of the experimental setup therefore there is no build-
up of vapour column. Due to the availability of significant overboard water depth, heat-up of the
system is slow and in latter stages of experiment, system is expected to attain pseudo steady
state. Figure 69 to Figure 72 confirms the existence of pseudo steady state in the latter stages of
experiment. Regression analysis technique is employed to develop the best possible correlation
representing experiment data. Comparison of power based correlation with the diffusion based
Table 17 Comparison of power based evaporation correlation with the Hugo’s correlation.
Sr. Hugo’s Correlation Developed Correlation Under High
Decay heat power = qc,u .W.D Decay heat power = qc,u .W.D
qr,u is radiative heat loss, qc,s evaporative heat loss Where qc,u is decay heat under quasi
175
P1,s)
of water in air.
conditions.
Yambient)
and
coefficient, W/m2K
coefficient
176
Cs is average sensible heat of air water
45
40 Experimental kg/m2hr
Hugoes, kg/m2 hr
35
Evaporation Flux, kg/m2hr
25 Exp-10%
20
15
10
0
77 127 177 227 277 327 377 427
Time, min
Figure 69 Comparison of developed correlation against Hugo’s model and experimental data
(P=10KW, initial liquid level 2046mm)
177
45
Experimental kg/m2hr
40
Hugoes, kg/m2 hr
35
Evaporation flux, Kg/m2hr
15
10
0
40 50 60 70 80 90
Temperature, C
(a)
32
Experimental evaporation flux, kg/m2hr
27
22
17
Experimental vs Corelation
10%
12
-10%
45 degree
7
2
2 7 12 17 22 27 32
Evaporation flux, Power based corelation, kg/m2hr
(b)
Figure 70 (a) Comparison of developed correlation (Evaporation flux versus average top plane
temperature) against Hugo’s model and experimental data (P=10KW, initial liquid level
2046mm) (b) Parity plot between experimental data versus correlation predictions
178
30
Experimental kg/m2 Hr
Hugoes, kg/m2 hr
25
Developed corelation
Evaporation flux, kg/m2hr
20 Experimental +10per
Experimental -10%
15
10
0
0 100 200 300 400 500 600 700
Time, min
Figure 71 Comparing developed correlation against Hugo’s model and experimental data
(P=10KW, initial liquid level 2794mm)
179
30
Experimental 10KW
25 Developed corelation
Evaporation Flux, kg/m2hr
Hugoes
20
Experimental +10%
15 Experimental -10%
10
0
29 39 49 59 69 79 89 99
Temperature, C
(a)
30
Experimental evaporation flux, kg/m2hr
25
20
15
Experimental vs Corelation
10 P=10KW,2794mm
45 degree
5 10%
-10%
0
0 5 10 15 20 25 30
Evaporation flux, power based corelation, kg/m2hr
(b)
Figure 72 (a) Comparison of developed correlation (Evaporation flux versus average top plane
temperature) against Hugo’s model and experimental data (P=10KW, initial liquid level
2794mm), (b) Parity plot between experimental data versus correlation predictions
180
Figure 69 to Figure 72 benchmarks the evaporation mass flux predictions of developed
correlation against the predictions of Hugo’s correlation applied under similar operating
conditions and experimentally measured evaporation mass flux data. As evident from the above
comparison, developed correlation is able to predict the evaporation flux more precisely as
developed correlation against the evaporation mass flux data (Figure 69 and Figure 70 (a))
highlights the good agreement between both (close to -10% range) for 2046mm initial liquid
level experimental run. Parity plot in between experimental evaporation mass flux data and
power based correlation predictions is shown in Figure 70 (b). From Figure 70 (b) it can be infer
that predictions of power based correlation are within 10% range compared to the experimental
results except for initial heat-up phase. Initial deviation is due to the use of average relative
humidity value in the correlation, whereas during the course of experiment, relative humidity
value is continuously varying with time. Better fit during the initial heat up phase can be
achieved by using piece wise linear approach During later phase of heat up, better fit is observed
due to increase in vapour pressure of water with increase in surface temperature at the air water
interface, this weakens the effect of relative humidity term in the power based correlation
Whereas, benchmarking of developed correlation against the evaporation mass flux data (Figure
71 and Figure 72 (a)) highlights the good agreement between both (within -10% range) for
2794mm initial liquid level experimental run only after 400min (nearly 70oC pool temperature).
As observed earlier for P=10KW and initial liquid level of 2046mm case, initial deviation is due
to the use of average relative humidity value in the correlation, whereas during the course of
experiment, relative humidity value is continuously varying with time. Better fit during the initial
heat up phase can be achieved by using piece wise linear approach during later phase of heat up,
181
better fit is observed due to increase in vapour pressure of water with increase in surface
temperature at the air water interface, and this weakens the effect of relative humidity term in the
power based correlation. From Figure 72 (b) it can be infer that power based correlation predicts
the evaporative mass flux within 10% range compared to the experimental results. It is principal
to note that as heating of the SFP is a complex transient phenomenon involves energy addition
via decay heat of the stored fuel assemblies and energy dissipation by various mechanisms such
mass flux correlation is only possible by assuming pseudo steady state of the pool. It is noted that
for low liquid levels in a pool (such as nearly 2000mm initial liquid level in experimental run 4,
5 and 6) pool overboard water attains pseudo steady state quite quickly as compared to the high
pool level cases (such as nearly 2780 mm initial liquid level in experimental run 1, 2 and 3).
Similar behaviour can be observed from Figure 71 and Figure 72 for the 2794 mm liquid level
case pool may only attains pseudo steady state beyond 400 min (Pool temperature of 70oC )
whereas for initial 2046mm liquid level case pools quickly attains quasi steady state. Therefore
the applicability of the power based developed correlation is limited to pseudo steady state
region.
Several principal variables considered in the present development of evaporation correlation are
following:
182
vapour pressure evaluated at surface temperature, PSat(Tω) is the vapour pressure
surface temperature, PSat(Tω) is the vapour pressure evaluated at free stream conditions.
From the data generated during 2nd and 3rd set of experimental runs, Regression analysis of
experimental evaporation data generated for different power, same liquid level under various
ambient conditions highlights the strong effect of surface temperature on the evaporation rates
whereas the effect of decay power ratio is observed as weak under these conditions. It is
principal to note that under 2nd and 3rd experimental runs, depth of overboard water column is
kept significant whereas heater power is kept low. This situation simulates the inception of phase
1 ESBO accident scenario in the SFP. As the effect of the decay power ratio is weak on the
evaporation rate, the experimental setup is expected to be in the transient heating throughout the
time. Under this scenario diffusion based Mass transfer mechanism governs the evaporation rate.
Regression technique is used for the development of diffusion based evaporation correlation.
Table 18 compares the diffusion based developed correlation with the Hugo’s correlation.
183
Table 18 Comparison of diffusion based evaporation correlation with Hugo’s correlation
Sr. Hugo’s Correlation Developed Correlation Under High
Where qc,u is decay heat under quasi steady state, Where qc,u is decay heat under quasi
qr,u is radiative heat loss, qc,s evaporative heat loss , steady state, qr,u is radiative heat loss,
184
25
Experimental 6.66KW
20
Evaporation flux, kg/m2hr
10
0
52.5 57.5 62.5 67.5 72.5 77.5 82.5 87.5 92.5 97.5
Pool Temperature, oC
Figure 73 Comparing developed correlation (Evaporation flux versus average top plane
temperature) against experimental data (P=6.66 KW, initial liquid level 2763 mm)
7 Experimental, 3.33KW
6
Evaporation flux, kg/m2hr
0
-4 6 16 26 36 46 56 66 76
Pool Temperature, oC
(a)
185
5
Experimental vs
Experimental evaporation flux, kg/m2hr
4.5 corelation
45%
4
3.5 10%
3
-10%
2.5
1.5
1.5 2 2.5 3 3.5 4 4.5 5
Evaporation flux, diffusion based corelation, kg/m2hr
(b)
Figure 74 (a) Comparing developed correlation (Evaporation flux versus average top plane
temperature) against experimental data (P=3.33 KW, initial liquid level 2767mm, (b) Parity plot
between experimental data versus correlation predictions
Figure 73 and Figure 74 benchmarks the diffusion bases developed correlation with the
experimental data obtains during test runs. Developed correlation predications are in good
agreement with the experimental data. From Figure 73 specific to experimental run with
Power=6.66 KW and initial liquid level 2763 mm, predictions of developed correlation are found
to be in agreement with the experimental results beyond pool temperature of 80oC. This initial
deviation is due to the use of average relative humidity value in the correlation, whereas during
the course of experiment, relative humidity value is continuously varying with time. Better fit
during the initial heat up phase can be achieved by using piece wise linear approach. During later
phase of heat up, better fit is observed due to increase in vapour pressure of water with increase
in surface temperature at the air water interface, this weakens the effect of relative humidity term
186
in the correlation. From Figure 74 (b), specific to experimental run with Power 3.33 KW and
initial liquid level 2767 mm. It is observed that developed correlation predicts the evaporation
flux within 10% range compared to the experimental results right from the beginning of heat up.
Deviation in the later phase of heat up observed is due to the use of average relative humidity
value in the correlation, whereas during the course of experiment (extended beyond 2 days in this
case), relative humidity value is significantly varies with time due climate variation because of
raining. It is principal to note that the scaled heater power of 3.33KW represents the uniformly
distributed actual decay power of 17.14MW (Full filled Actual At- Reactor SFP VVER1000, 582
FAs). Therefore in principal diffusion based evaporation correlations predict the evaporation rate
during the inception of phase-1 accident under ESBO conditions in SFP when initial overboard
water column is significant. During the latter stages of accident when overboard water column
drops to low level (Lt/Lf< 2) application of power based evaporation correlation is advisable.
Application of power based evaporation correlations is also advisable in emergency full core
dump scenarios; under this scenario SFP may experience maximum thermal loading.
Comparison of power based and diffusion based evaporation correlation is tabulated in Table 19.
Figure 75 to Figure 84 compares the estimated results from these two approaches against the
experimental data generated for different operating conditions of the experimental setup.
187
Table 19 Comparison of Power based and Diffusion based correlations
Sr. Developed Correlation Under Developed Correlation Under High
Where qc,u is decay heat under quasi steady Where qc,u is decay heat under quasi
state, qr,u is radiative heat loss, qc,s evaporative steady state, qr,u is radiative heat loss, qc,s
mass flux.
P1,e)/(P-P1,s)
P1,s)
and
188
Le is Lewis number, Le =Sc/Pr , For air-water
vapour mixtures Le =1
W/m2K
heat or Low liquid level (Lt/Lf< 2)case decay heat or High liquid level (Lt/Lf>
2.8)case
(Power based approach)
P1,s) P1,e)/(P-P1,s)
189
P1,s) P1,e)/(P-P1,s)
25
Experimental kg/m2 Hr
10
0
17 117 217 317 417 517 617 717
Time, min
Figure 75 Comparing developed correlations with experimental data (P=10KW, initial liquid
level 2794mm)
190
25
Experimental kg/m2 Hr
20
Diffusion based
Evaporation Flux, kg/m2hr
Corelation
Power based Corelation
15
10
0
0 10 20 30 40 50 60 70 80 90 100
Temperature, C
Figure 76 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=10KW, initial liquid level 2794mm)
25
Experimental
20 Power based corelation
Evaporation Flux, kg/m2hr
15
10
0
52 152 252 352 452 552 652 752 852 952 1052
Time, min
Figure 77 Comparing developed correlations with experimental data (P=6.66KW, initial liquid
level 2763mm)
191
25
20 Experimental
Evaporation Flux, kg/m2hr
10
0
52 57 62 67 72 77 82 87 92 97
Temperature, C
Figure 78 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=6.66KW, initial liquid level 2763mm)
12
10 Experimental
Evaporation Flux, kg/m2hr
0
500 1000 1500 2000 2500 3000
Time,min
Figure 79 Comparing developed correlations with experimental data (P=3.33KW, initial liquid
level 2767mm)
192
12
10 Experimental
0
29 39 49 59 69 79 89
Temperature, C
Figure 80 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=3.33KW, initial liquid level 2767mm)
40
35 Experimental
kg/m2hr
30
Evaporation flux, kg/m2hr
Power based
25 corelation
20 Diffusion based
Corelation
15
10
0
78 128 178 228 278 328 378 428
Time, min
Figure 81 Comparing developed correlations with experimental data (P=10KW, initial liquid
level 2046mm)
193
40
35 Experimental
kg/m2hr
30
Power based
corelation
25
Diffusion based
20 Corelation
15
10
0
40 50 60 70 80 90
Figure 82 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=10KW, initial liquid level 2046mm)
25
Experimental
20 Power based
Evaporation Flux, kg/m2hr
corelation
Diffusion based
15 corelation
10
0
0 200 400 600 800 1000 1200 1400 1600 1800
Time, min
Figure 83 Comparing developed correlations with experimental data (P=3.33KW, initial liquid
level 2022mm)
194
25
20
Evaporation Flux, kg/m2hr
Experimental
15 Power based
corelation
Diffusion based
10 corelation
0
30 40 50 60 70 80 90 100
Temperature, C
Figure 84 Comparing developed correlations (Evaporation flux versus average top plane
temperature) with experimental data (P=3.33KW, initial liquid level 2022mm)
No of Experimental correlation
setup
1 P=10KW, Initial liquid Full core dump in a single compartment Power based
2 P=10KW, Initial liquid Full core dump in a single compartment Power based
195
3 P=3.33KW, Initial Fully exhausted SFP with average decay Power based
4 P=3.33KW, Initial Fully exhausted SFP with average decay Diffusion based
5 P=6.66KW, Initial Fully exhausted SFP with average decay Diffusion based
6.2 SUMMARY
This chapter covers literature on evaporation models specific to NPP SFP applications in a
condensed form. M. Quinn Brewster [14] reviewed many models like Mc Adams, Shahs,
Horizontal natural convection model (HNC), Stagnant film model, Hugo’s model etc. Based on
this comparison, their assessment is that the HNC model and Hugo’s model come closest to
modelling the actual flow. ). Due to nearly explicit nature of Hugo’s model in solving for pool
temperatures, predictions of Hugo’s model are compared with the experimental data for P=
10KW, Lt= 2780mm, and P=10KW, Lt=2000mm cases. Comparison of results highlights the
shortcoming of Hugo’s model to predict the evaporation mass flux correctly for high heater input
power cases. To overcome this observed limitation, a new model of evaporation from warm SFP
is developed in this chapter. Developed evaporation model is applicable from low operating
temperature (T = 50oC) (Low mass transfer regime) up to high operating temperature (T > 90 oC)
196
(High mass transfer regime). Two different empirical correlations are developed that is nearly
explicit in solving for pool temperatures. Empirical correlations are developed for the estimation
of surface evaporation flux with two different approaches. They are diffusion based approach
(similar to Hugo’s correlation) and decay power based approach, each specific to SFP conditions
during the course of accident. Predictions of power based correlation are found to be more
consistent with the experimental data for high decay heat and low liquid level scenarios whereas
for high liquid level and low power scenarios predictions of diffusion based correlation are found
197
CHAPTER-7. POST-TEST CFD SIMULATIONS
Nuclear Reactor Safety (NRS) is less well developed but is rapidly accelerating. The need to use
CFD arises because many traditional reactor system codes (such as RELAP5) are based on a
network of 1-D or 0-D volumes. As observed from previous sections that the flow in components
such as upper and lower plenum, downcomer and within fuel assemblies are highly three
dimensional in nature. Natural circulation, mixing and stratification (if any) within the SFP is
also essentially 3-D in nature, and representing such complex flows by pseudo 1-D
approximations may not just be oversimplified but could even be misleading, resulting in
erroneous judgments being made. This section covers the Post-test CFD simulation of the
7.2 OBJECTIVE
2. Thermal hydraulic assessment of the experimental setup operating under given operating
conditions.
ANSYS WORKBENCH Version 19.2 with inbuilt pre-processor and post processor is employed
to initiate the CFD simulations. CFD solver, in particular ANSYS FLUENT is used for
performing entire CFD related activities. ANSYS FLUENT solves equations of mass,
198
momentum and energy using control volume approach. ANSYS FLUENT is a widely accepted
(38)
[ ] (39)
[ ] [ ] (40)
In the present analysis, the interaction of natural circulation plumes and their thermal deposition
development of thermal profiles within the experimental setup which ultimately influence the
considered in the analysis. The schematic view and dimensions of computational grid is shown
in Figure 85 and Figure 86 structured grid with mesh size of 0.1mm is generated by using
199
ANSYS MESH (Pre-processor). In principal 2 million total numbers of quad elements is
generated by pre-processor.
200
(a) (b)
Figure 86 Computational Domain (a) Overall 2D-StructuredMesh, (b) Magnified view of heater
Boundary conditions used in the case setup are tabulated in Table 22.
201
Table 22 Boundary conditions
1 Heat transfer coefficient at the air water 1000W/m2K, Average value estimated by
2 Heat transfer coefficient at the side 5W/m2K, From Churchill and Chu
walls correlation
Fluid and material properties used in the case setup are tabulated in Table 23.
ρ=8030 Kg/m3
k=16.27 W/m K
Cp=502.48 J/kg K
202
7.5.4 TURBULENCE MODEL- No turbulence model is selected, laminar flow is considered in
the analysis.
7.5.5 ASSUMPTIONS
in the analysis, whereas heat loss due to surface evaporation is modeled by using average
2. Decrease in the fluid level due to the surface evaporation is not modeled in the analysis.
5. Heat transfer due to natural convection of air above the air-water interface is ignored in
the analysis.
1 Continuity 0.001
2 x-velocity 0.0001
3 y-velocity 0.0001
4 Energy 10-6
1 Scheme PISO
203
2 Time Transient
Spatial Discretization
2 Pressure Standard
Continuity
x velocity
y velocity
Energy
7.6 RESULTS
Temperature profiles (Area weighted average temperatures) measured at four different locations
204
Plane 1
400 mm
Plane 2
400 mm
Plane 3
400 mm
Plane 4
400 mm
205
18200 18400 18600 18800 19000 19200 19400 19600 19800 sec
18200 18400 18600 18800 19000 19200 19400 19600 19800 sec
206
18200 18400 18600 18800 19000 19200 19400 19600 19800 sec
18200 18400 18600 18800 19000 19200 19400 19600 19800 sec
207
Figure 93 Temperature contours at t=329 min, Pt=10KW
208
7.7 DISSCUSSION ON RESULTS
volume.
shrouds) is also contribute in the heat dissipation from the Fuel assemnbies.
3. Heat loss from the air-water inerface and side walls facilitate the setup of natural
convetion loops.
4. Heat loss from the air water interface is dominant with respect to the wall losses,
5. Weak natural convective flow is only observed within the heater shroud
assembly. No natural convection loops are observed in the bulk overboard water
6. During the initial heatup transient, dominant mode of heat transfer within the
over board water volume is conduction only. Natural convection loops may
7. Prediction of heatup rate by CFD technique i.e 1.8oC for 30min of flow time is
in proximity with the experiment observations of 2oC per 30min of run time.
Insignificant thermal stratification is predicted in the pool by CFD techniques during the heat up
transient. Natural convective flow is only observed within the heater shroud assembly. No
209
natural convection loops are observed in the bulk overboard water volume during the heat-up
transient i.e dominant mode of heat transfer within the overboard water volume is conduction
within the fluid. It is observed that 2D- CFD model is not able to simulate the natural convection
phenomena in the overboard volume. Due to computational limitations a relatively simple CFD
model is used in the present work. However a 3-dimensional CFD model of the experimental
setup comprises of heater shroud section modeled as a porous body coupled with hydraulic
resistance coefficients of heater shroud assembly, required volumetric heat generation and
variable heat transfer coefficient with respect to temperature as a boundary conditions for
removing heat due to surface evaporation may capture the development of natural convection
7.9 SUMMARY
This chapter covers the post-test CFD simulation of the experimental setup operating at 10KW of
heater power. Simlified two dimensional geometric model with actual experimental setup
domain is meshed by using ANSYS workbench meshing tool. Nearly 2 million structured mesh
with 0.1mm cell size is generated. Fluent case is setup as volumetric heat generation source in
the heater section and average heat transfer cofficient on the top surface as a boundary condition.
Entire fluid domain is defined as water with boussinesq approximation for density estimation,
rest other thermophysical proterties are assumed to be constant and solid domain is defined as
stainless steel (SS316). Entire case setup is initialised at 50oC as a startup temperature. It is
observed that negligible thermal stratification is predicted in the pool by CFD techniques during
the heat up transient. Natural convective flow is only observed within the heater shroud
assembly. No natural convection loops are observed in the bulk overboard water volume during
210
the heat-up transient i.e dominant mode of heat transfer within the overboard water volume is
conduction within the fluid. It is observed that 2D- CFD model is not able to simulate the natural
convection phenomena in the overboard volume. Limitations of CFD model and possible
211
CHAPTER-8. CONCLUSIONS AND RECOMMENDATIONS
This thesis provides a detailed insight of thermal hydraulic phenomena occurring within the
NPP-SFP during loss of coolant accident. Following are the conclusions which can be derived
1. Empirical correlations are developed for the estimation of evaporative mass flux from warm
NPP-SFP air water interface under prolonged loss of cooling accidental scenario. Based on
observations, two different modeling approaches i.e., diffusion based approach (similar to
Hugo’s correlation) and decay power based approach, has been developed for different states
2. Predictions of power based correlation are found to be more consistent with the experimental
data for high decay heat and low liquid level scenarios, whereas for high liquid level and low
power scenarios predictions of diffusion based correlation are found to be in line with the
experimental data.
3. Thermal stratification is not observed for the experiments conducted at low and high heater
power. Local temperature distribution within the setup confirms the existence of circulating
buoyant hot water plume. This buoyant hot water plume circulations are strong enough for a
4. Significant thermal stratification in the overboard volume is predicted by the RELAP5 code.
Hence reduction of number of control volumes for RELAP5 model will result in good
mixing. This will decrease the thermal stratification in the overboard volume.
212
5. Insignificant thermal stratification is predicted in the pool by CFD techniques during the heat
up transient. Natural convective flow is only observed within the heater shroud assembly. No
natural convection loops are observed in the bulk overboard water volume during the heat-up
transient. During the heat up phase, dominant mode of heat transfer within the overboard
water volume is conduction only. It is observed that 2D- CFD model is not able to simulate
6. Due to computational limitations a relatively simple CFD model is used in the present work.
However a 3-dimensional CFD model of the experimental setup comprises of heater shroud
section modeled as a porous body coupled with hydraulic resistance coefficients of heater
shroud assembly, required volumetric heat generation and variable heat transfer coefficient
with respect to temperature as a boundary conditions for removing heat due to surface
evaporation may capture the development of natural convection loops within the system.
The work presented in this thesis is confined up to initial phase of loss of cooling accident in
NPP-SFP (Phase-1 accident). This work can be extended further covering fuel uncovery (Phase-
2 accident) and fuel oxidation (Phase 3 accident) phase during loss of cooling accident in NPP-
SFP. This study can be further extended for the assessment of fuel cooling and structural
integrity during reflooding exercise within the SFP. Rigorous CFD numerical experiments can be
undertaken for simulating the natural convection flow circulation phenomena within the NPP-
SFP.
213
REFERENCES
[1] www.nrc.gov
[2] Organisation for Economic Co-operation and Development. (August 25, 2017).
PIRT on Spent Fuel Pools under loss of cooling and loss of coolant accident
Spent Fuel Pools in case of loss of heat removal due to water leakage. 8th
Thermodynamics.
[6] Yanagi, C., Et al., 2012. Evaluation of heat loss and water temperature in a spent
[7] Hung, T-C., et al., 2013.The development of three dimensional transient CFD
model for predicting cooling ability of spent fuel pools. Applied Thermal
Engineering, pp 496-504.
[8] Kataoka et al., 1992. Experiments on convection heat transfer along a vertical flat
2014. Prediction of temperature and water level in a spent fuel pit during loss of
214
[10] Tadashi FUJI, Yoshiyuki KATAOKA and Michio MURASE., 1995. Evaporation
[11] Noureddine Boukadida, Sassi Ben Nasrallah,. January 2000. Mass and heat
[12] Bruce R. Hugo, William C. Kinsel,. February 2014. Predicting Evaporation Rates
from spent nuclear fuel storage pools. International nuclear safety journal.
[13] Mirza Mohammed Shah., January 2014. Methods for calculation of evaporation
rates by natural convection air flow with applications to spent fuel pools.
[15] L.M.K Boelter, H.S. Gordon, J.R. Griffin,. 1946. Free evaporation into air of
water from a free horizontal quiet surface. Industrial and engineering chemistry.
[16] Doctor of Philosophy thesis by Bruce Robert Hugo. December 2015. Modeling
[17] P. K. Vijayan, H. Austregesilo (1994)., Scaling Laws For Single Phase Natural
Circulation Loops.
[18] H.H. Bau and K.E. Torrance, Transient and steady behavior of open,
symmetrically heated, free convection loop, Int J. Heat Mass Transfer 24 (1981)
215
597-609
[19] M. Ishii and I. Kataoka. Similarity Analysis And Scaling Criteria For LWR’S
Convection Shutdown Heat Removal System Testing. Decay Heat Removal and
Convection Liquid Metal Fast Breeder Reactor Shutdown Heat Removal Test
ANSI/ANS-5.1-2005
216
ANNEXURE
Consider the statistical modeling between the dependent and one independent variable. When
there is only one independent variable in the linear regression model, the model is generally
termed as simple linear regression model. When there are more than one independent variables
in the model, then the linear model is termed as the multiple linear regression model.
Let y denotes the dependent (or study) variable that is linearly related to k independent (or
This is called as the multiple linear regression model. The parameters are the regression
coefficients associated with X1,X2,…,Xk respectively and is the random error component
reflecting the difference between the observed and fitted linear relationship. There can be various
reasons for such difference, e.g., joint effect of those variables not included in the model, random
factors which cannot be accounted in the model etc. The jth regression coefficient represents the
expected change in y per unit change in jth independent variable Xj .As , model is said
to be linear when it is linear in parameters. In such a case should not depend on β's.
specific to power based approach is subjected to fit in the experimental data by means of linear
considering this equation equivalent to Y = a.X1m.X2n, taking log on both sides, ln(Y)= ln (a) +
217
m ln (X1)+ n.ln (X2), this equation resembles with Y =X'1β1+X'2 β2. Similarly for diffusion based
data by means of linear multi variable linear regression tool. As m"[Kg/m2.hr] = a ( )m.ln((P-
P1,e)/(P-P1,s) considering this equation equivalent toY = a.X1m.X2n, taking log on both sides,
ln(Y)= ln (a) + m ln (X1)+ n.ln (X2), this equation resembles with Y' =X'1β1+X'2 β2.
interface
Used Multivariable regression analysis for
estimating unknowns of Y' =X'1β1+X'2 β2+X'3β3
218
Compiled data for Lt= 2022 mm, P= 6.66 KW
219