5 PRE-TEST ANALYSIS OF LIFUS5/Mod3
Pre-test calculations are carried out with two main models: at first, attention is focused on chemical reaction model, after that another calculation is performed in order to investigate the thermodynamic and chemical models.
5.1 General criteria adopted for code models
The main SIMMER-III code options of reference calculations are listed hereafter and reported in [2]:
• The chemical model (namelist XSWC). In this namelist, FMOL was set to 0.75, FKCR was set to 0.1 (both are the default SIMMER values). For more details of the implementation of the chemical correlations between lithium-lead and water see [4].
• Inter-cell heat transfer applied between all the liquid components and solid particles, in vapor, in the structures, and between structures and liquid components (flag HMTOPT in the namelist XCNTL).
• Adjustment of vapor temperature in the two-phase cells with very small void fraction and instability in numerical calculation (flag EOSOPT in the namelist XCNTL).
• The properties of the lithium lead (XEOS and XTPP) are taken into account from CEA (Ref. [12]), while the properties of the lithium compound were simply set with the available information (Perry’s chemical engineers’ handbook) starting from the properties of the sodium compounds (Ref. [10]).
• All flags MXFOPT in the namelist XCNTL, including the drag coefficients, are set as the default SIMMER values.
• Since SIMMER-III code calculates the friction only in the mesh cells where the can wall structures are implemented, in the chemical interaction model calculations the friction in the injection line was neglected.
• The concentrated pressure drops due to geometrical discontinuities of LIFUS5/Mod3
facility are set at the orifice of the injector device, across the expansions and
restrictions in the injection line, across curves and direction changes, and through
valves and Coriolis mass flow meter (namelist XMXF).
5.2 Chemical interaction model
The aim of pre-test calculations is to design the experimental tests suitable for the assessment of the chemical model of SIMMER-III code. For this purpose, a well- established mass of water will be injected in reaction vessel S1_B.
5.2.1 Nodalization by SIMMER-III code
The LIFUS5/Mod3 facility is modelled as a reaction vessel, S1_B, an injection line, including a ½’’ injector and 3 penetrations, by which expansion, instrumentation and hydrogen extraction lines are represented. These penetrations are asymmetric with respect of vessel axis; therefore, their modelling was made by substitution area method.
The nodalization is obtained by 23 radial and 81 axial mesh cells. The facility domain is shown in Fig. 5.2. Colours distinguish the different fluids, as set at the beginning of the transient (t = 0 s). Therefore, the PbLi is represented in red, the water in blue, and the cover gas (initial amount of Hydrogen due to numerical constraint instead of Argon) in white. Rotating the 2D SIMMER-III domain along the axis of symmetry, the whole volumetric model is obtained, in which every cell is a toroidal volume with rectangular section.
Test section has been modelled as 3 fixed virtual walls; each space between one virtual wall and other represent a series of holes that will be present in the real test section and, as mentioned above, the sum of each area is equivalent to the substitution area implemented in SIMMER-III geometry. In the final configuration, thermocouples are fixed on 3 coaxial tubes attached to the holed plate (Fig. 5.1).
Fig. 5.1 – LIFUS5/Mod3 test section holed plate design.
The injection line cannot be coherently modelled in an axisymmetric domain. Therefore, to preserve the cylindrical shape of the injection tube, it was positioned vertically and coaxially with the model. This simplification does not respect the real geometry.
However, the total length and the cross-flow area are preserved. Friction losses and
pressure drops are simulated by orifice coefficients across geometrical discontinuities,
valves, and Coriolis flow meter (Fig. 5.4). Orifice coefficient of enlargement/constriction
flow rate and factors reported in datasheets using a spreadsheet. In Tab. 5.1 a summary of the orifice coefficients of injection line is reported for all components.
CELL COMPONENT ORIFICE COEFFICIENT
(1÷3,2) Curve 0.9
(1÷3,4) Valve 3.67
(1÷3,9) Curve 0.9
(1÷3,11) Curve 0.9
(1÷3,12) Constriction 0.01
(1÷3,15) Coriolis flow meter 0.76
(1÷3,16) Enlargement 0.25
(1÷3,20) Valve 3.67
(1÷3,33) Curve 0.9
(1,41) Injector constriction 0.46 (1,42) Injector enlargement 0.99
Tab. 5.1 – Summary of orifice coefficients of injection line
The reference mesh cells used for the analysis are reported in Fig. 5.3: those will be used for comparison with experimental data during post-tests calculations are (23,46) and (9,60) for pressure inside the reaction vessel S1_B in PbLi and Argon phase, (1,34) for pressure and temperature of the injected water, (1,1) for the pressure in the gas cylinder above the injection line.
The correspondence of main dimensions of LIFUS5/Mod3 facility preliminary design and SIMMER-III nodalization is reported in Tab. 5.2.
In order to evaluate the amount of hydrogen produced by the SIMMER-III implemented chemical model, BFCAL tool is used (Ref.[11]). This tool was developed to post-process the basefile of the SIMMER-III calculations permitting to obtain some specific parameters (such as the total mass, total energy, average temperature, location of the center of mass, etc.) in specified macro-regions. By using this tool, the total mass of the components at the start and at the end of the simulations can be calculated using the data results of the SIMMER-III calculations. The nodalization of the BFCAL tool is shown in Fig. 5.5, in which 9 different macro-regions are highlighted.
5.2.2 Boundary and Initial Conditions
The calculations start at t = 0 s, which represents the valve VP-SBL-06 opening.
Downstream the valve, the pressure is set at vacuum values to reproduce the
experimental procedures of the tests. Upstream the valve at the beginning of the
injection line, boundary conditions of continuous inflow and constant pressure (155
bar) are applied, to simulate the behavior of the Argon gas cylinder. A fixed amount of
water is imposed in the line, and is completely injected inside the reaction vessel S1_B
once the valve opens. The amount of water to be injected is represented by filling a
The injector break-up is simulated by the disappear of the virtual wall, which recreates the orifice. In order to overcome the limitations of SIMMER-III code, which does not allow to set the presence/absence of virtual walls basing on pressure conditions, the time rupture is calculated by the user in two steps: first, the calculation is run with the virtual wall closed in order to take into account the time at which the pressure reaches about 155 bar in cell (1,41). Then, the calculation is run again imposing the opening of the virtual wall at that time. The initial conditions of pressure and temperature in S1_B and in the injection line are set coherently with the data reported in the test matrix (Tab.
5.3). These assumptions imply that the water mass flow rate and the injection time are
not imposed but are calculated by SIMMER-III accordingly to pressure difference
between the injector and the reaction vessel.
Fig. 5.2 – SIMMER-III modeling of LIFUS5/Mod3 facility.
H
2extraction
Injection line
Reaction vessel Expansion
line
Instrumentation
line
Fig. 5.3: SIMMER-III modeling. Zoom on S1_B reaction vessel and reference mesh cells.
Fig. 5.4 – SIMMER-III modeling. Zoom on the injection line and orifice coefficients.
Position of orifice coefficients
PK(10,80) TLK1(5,45-47-
49) TLK1(4,45-47-
49)
PK(9,60)
PK(23,46)
TLK1(12,45-47-49)
Fig. 5.5 – SIMMER-III BFCAL nodalization of LIFUS5/Mod3 facility.
LIFUS5/Mod3 modeling LIFUS5/Mod3 preliminary design Volume parameter cells dimension Volume parameter dimension
S1_B h 38-57 0.52 S1_B h 0.59
D 1-23 0.267 D 0.257
V - 0.0292 V 0.0286
Free gas h 58-61 0.0682 Free gas V TBD*
D 1-23 0.267
V - 0.00382
Inj_device h 38-41 0.106 Inj_device h 0.08
D 1-3 0.01388 D 1/2’’
Dorifice SENSITIVITY Dorifice TBD*
Inj_line h 1-37 7.951 Inj_line L ~ 8.0
D 1-3 0.01388 D 1/2’’
*TBD accordingly with pre-test analyses
Tab. 5.2 – SIMMER-III LIFUS5/Mod3 model: correspondence of main dimensions.
5-6: injection line
3: injector
1: PbLi
7-8-9:
Extraction lines
4: cover
gas
5.2.3 Analysis of pre-test calculation
During the experimental campaign on LIFUS5/Mod3, 5 different tests are planning to be executed with fresh lithium lead in order to:
• Obtain reliable, reproducible, and detailed experimental data with well-known initial and boundary conditions by means of accurate instrumentations, to be compared with the numerical post-test calculations;
• Validate and assess the chemical model implemented in SIMMER-III Ver. 3F Mod. 0.1 by means of the measurement of hydrogen produced by the chemical reaction, the total injected mass of water, and temperature map;
• Improve the knowledge of physical behaviour and understanding the phenomenon involved during the interaction/reaction;
• Investigate the dynamic effects of energy release on the structures, and of the chemical reaction and hydrogen production;
After the results of the sensitivity analyses reported in Ref. [13], and considering the objectives discussed above, we concluded to perform 5 tests (reported in Tab. 4.1) to be chosen from the pre-tests reported in Tab. 5.3:
#
Injected water [g]
T H
2O [°C]
D orifice
[mm]
T PbLi [°C]
P inj [bar]
P cap rupture
[bar]
t cap rupture
[ms]
V gas [%]
1 100 300 4 330 155 161 28 27.34
2 50 300 4 330 155 178 25 27.34
3 150 300 4 330 155 155 37 27.34
4 200 300 4 330 155 150 39 27.34
5 250 300 4 330 155 157 43 27.34
6 100 285 4 330 155 182 31 27.34
7 100 325 4 330 155 158 32 27.34
8 100 300 1 330 155 150 28 27.34
9 100 300 2 330 155 150 28 27.34
10 100 300 8 330 155 151 28 27.34
Tab. 5.3 – Performed tests.
For sake of completeness, the results of one case are analyzed hereafter, which is considered the “reference” case, while in the following sections, the results of the other tests are reported, as sensitivities in respect to the “reference” calculation.
5.2.3.1 Initial condition results
The SIMMER-III initial conditions at the time of valve opening (t=0 s for the pre-test
calculations) are reported in Tab. 5.4.
# Parameter Unit Design Value CALC value
1 Abs. pressure in S1_B bar 1 1
2 PbLi temperature in S1_B °C 330 330
3 Injected water temperature °C 300 300
4 Abs. pressure in injection line bar 0.01 0.01
5 Gas cylinder pressure bar 155 155
6 Free gas volume m
3TBD 5.57E-3
Tab. 5.4 – L5M3_Case#1: comparison between design and calculated values at the initial conditions.
5.2.3.2 Reference calculation results
The “reference” input deck is labeled “#1”, as reported in Tab. 5.3. The related time trends and the resulting sequence of events are reported from Fig. 5.6 to Fig. 5.23. The transient can be divided into four different phenomenological phases (Fig. 5.12):
Phase 1 [from onset of valve opening to 28 ms]: water injection line pressurization.
As soon as the valve VP-SBL-06 opens, water starts to flow and to pressurize the pipeline upstream the injection cap. The start of the transient (t = 0 s) is selected as the time of the valve opening. A constant time trend of pressure is imposed at the beginning of the injection line to simulate the constant inflow of Argon gas from the cylinder through the line. The design of the test specifies that the cap should be ruptured at the reference pressure of 155 bar, therefore the calculation is set by the disappearing of the virtual wall which simulates the injector orifice. The time rupture is calculated by two- steps: first, the calculation is run with the virtual wall closed in order to take into account the time at which in cell (1,41) the pressure reaches about 155 bar. Then, the calculation is run again imposing the opening of the virtual wall at that time. The thermal-hydraulic conditions of water (pressure and temperature) at the start of the injection calculated by SIMMER-III code are listed in Tab. 5.5 and illustrated from Fig.
5.6 to Fig. 5.10.
# Conditions Value Note
1 Pressure at the injector 160.83 bar
The thermal-hydraulic conditions are calculated by the code @ t = 28 ms after SoT
at the mesh cell (1,41)
2 Temperature of injected water 324.31°C
3 Void fraction of injected water 0.94 [-]
4 Time of injector rupture 28 ms
Tab. 5.5 – L5M3_Case#1: thermal-hydraulic conditions of water at the injection time.
Phase 2 [from 28 to 36 ms]: coolant flashing and first pressure peak.
The water injection and flashing in the melt of the reaction vessel causes a sudden steep pressure peak. The value of the pressure peak reaches 26.9 bar in in the reference cells of PbLi zone (23,46), then decreases slightly. In cover gas zone (9,60) and hydrogen extraction line (10,80) pressure undergo only a small increase, as shown in Fig. 5.11 and in Fig. 5.12.
The calculated mass flow rate of the injected water (Fig. 5.13) present a spike at 1.42 kg/s, then decreases due to the pressurization of the reaction vessel. Nevertheless, water is continuously injected in this phase, indeed the mass flow rate increases again.
The maximum amount of injected water is 0.0105 kg of water.
During phase 2, the hydrogen generated is still negligible (according to code results, the maximum value during this phase is 0.56 g (Fig. 5.13) but the value increases during the transient, reaching the equilibrium at the end of phase 3.
Fig. 5.20 to Fig. 5.22 show the PbLi temperature in S1_B reaction vessel in different radial and axial position, accordingly to the design of the test section which will be installed inside the vessel to support the thermocouples. The results calculated by the code confirm that the chemical reaction in this phase is still negligible. Indeed, the temperature are more affected by the water cooling effect than the chemical reaction.
The PbLi temperature values are calculated by the code to be 235°C, which is the minimum value that SIMMER-III code can evaluate for the PbLi in liquid phase, being its melting temperature.
Phase 3 [from 36 ms to 600 ms]: pressurization due to water and gas injection and hydrogen generation, up to pressure equilibrium.
This phase can be further divided into two sub-phases, 1) up to 200 ms, characterized by the water injection and hydrogen production, and 2) up to the end of the phase, characterized by the continuous gas injection up to the pressure equilibrium.
In order to assure that all the water will be injected in the reaction vessel, the design of the tests specifies that a continuous gas flow is injected for all the duration of the experiment in the reaction tank S1_B. This procedure affects the pressure transient in the reaction vessel, which is not anymore driven by the water injection, flashing and chemical reaction, but it permits to exactly evaluate the amount of injected water (redundancy in its evaluation by means of Coriolis flow meter, DP meters, initial amount of water in the line), and therefore to validate the SIMMER-III chemical model.
During the first sub-phase, the pressure in S1_B reaction vessel increases (Fig. 5.11 and
Fig. 5.12), driven by the water injected, the water evaporation in the zones where the
chemical reaction leads to increase the temperatures above the saturation temperature,
and the hydrogen generation. In the meantime, the pressure in the cover gas region
increases. The last trend shows a peak and a low oscillating behavior due to the piston
effect of the injected gas. The effect of gas compression in the system is more noticeable
considering the pressure in the hydrogen extraction line. Due to the position of the
time, coherently with the pressure drops and the geometrical dimensions of the line.
From this time on, the pressure in the system is due only to the continuous gas injection, up to the system pressure stabilization (155 bar).
At the end of the phase 3, the results of total injected water and hydrogen generation are 0.1062 kg (Fig. 5.13 and Fig. 5.14) and 8.93 g (Fig. 5.13 and Fig. 5.15), respectively. The result of hydrogen production agrees with the theoretical stoichiometric values, being between the maximum (11.80 g) and minimum (5.90 g) values obtained by the chemical reaction considered (LiOH or Li
2O products). All the water, in liquid and vapor phases, is consumed at the end of the phase 3 (Fig. 5.18 and Fig. 5.19).
The procedure of such validation tests leads to change the calculated hydrogen evaluation. Indeed, SIMMER-III constraints do not permit to assign two different non- condensable gases. Therefore, the amount of hydrogen calculated by the code is affected by the amount of gas injected during the transient (Argon gas). The mass flow rate of hydrogen (Fig. 5.16) is then evaluated, time-integrated and detracted from the total mass of hydrogen.
Considering the temperatures (from Fig. 5.20 to Fig. 5.22), it is worth underlining that:
1. the mesh cells in which the temperatures are evaluated by the code are chosen to be coherent with the position of the thermocouples will be installed in the reaction vessel S1_B;
2. the unavoidable lumped approximation of the code modeling implies averaged but punctual conditions in the mesh cell, being SIMMER-III a system code. Moreover, the needed assumption and simplification of the 2-D axisymmetric modeling generally leads to lower temperature values (the peak temperature are local hot spot occurring at the interface between contact surfaces of the liquids).
During phase 3, the effect of the chemical reaction is not anymore negligible. Indeed, the temperatures increase, reaching a peak maximum value of 445°C at 120 ms, when the water is still injected. The highest values are calculated by the code in the top level of the reaction vessel, coherently with past experimental tests (Ref. [14]). The chemical effect is visible up to about 600 ms, when the temperature trends show another increase up to 370 °C. From this time on, the temperatures remain at their initial values.
Phase 4 [from 600 ms to EoT]: system pressure stabilization.
The phase 4 is characterized by the stabilization of the pressure in the system (Fig. 5.6).
At this moment, the pressures in the injector and in the reaction vessel are equalized, therefore the gas injection is stopped (Fig. 5.16). On the contrary, due to pressure stabilization, PbLi in the reaction vessel S1_B drops down into the injection line, as possible to note from Fig. 5.17 and Fig. 5.23.
The material composition and temperature maps during the calculated transient are
shown in Fig. 5.23. These screenshots highlight and confirm that the temperature peaks
of PbLi are hot spots at the interface between the two fluids, and between liquid metal
and cover gas at the end of the transient. More important is to underline where the
the remain volume, the inertia of the PbLi and the mixing phenomena do not lead to a significant increase in temperature.
# Phase Time span [ms]
1 Water injection line pressurization 0 – 28
2 Coolant flashing and first pressure peak 28 – 36
3 Reaction vessel pressurization 36 – 600
4 System pressure stabilization 600 – EoT
Tab. 5.6 – L5M3_Case#1: phenomenological analysis.
Fig. 5.6 – L5M3_Case#1: calculated pressure (PK) in the injection line at different positions (zoom 0 – 1.0 s).
Fig. 5.7 – L5M3_Case#1: pressure (PK) and temperature of Argon gas (TGK) between the gas cylinder and the beginning of the injection line imposed as BIC (zoom 0 – 100 ms).
Fig. 5.8 – L5M3_Case#1: calculated pressure (PK) and temperature of water (TLK3) in the injection line (zoom 0 – 100 ms).
Injector breaks-up
Injector breaks-up
Fig. 5.9– L5M3_Case#1: calculated pressure (PK) and temperature of water (TLK3) at the injector device (zoom 0 – 100 ms).
Fig. 5.10 – L5M3_Case#1: calculated volume fraction of water (ALPLK3) and vapor (ALPGK) at the injector device. (Zoom 0 – 100 ms).
Injector breaks-up
Injector breaks-up
Fig. 5.11 – L5M3_Case#1: pressure trends in S1_B reaction vessel.
Fig. 5.12 – L5M3_Case#1: pressure trends in S1_B reaction vessel. (Zoom 0 – 1.0 s).
Phase 4 Phase 3b
Phase 3a Phase 1
Phase 2
Fig. 5.13 – L5M3_Case#1: water mass flow rate and hydrogen production. (Zoom 0 – 1.0 s).
Fig. 5.14 – L5M3_Case#1: integral water mass flow rate.
Fig. 5.15 – L5M3_Case#1: calculated hydrogen generation (MASFPG) in different macro-regions (S1_B, Cover Gas, Extraction lines).
Fig. 5.16 – L5M3_Case#1: evaluation of gas (BFLUXG) and hydrogen mass flow rate through the injector.
Fig. 5.17 – L5M3_Case#1: PbLi mass flow rate (BFLUXF) through the injector.
Fig. 5.18 – L5M3_Case#1: mass of water (MASN1) in different macro-regions.
Fig. 5.19 – L5M3_Case#1: mass of vapor (MASN2) in different macro-regions.
Fig. 5.20 – L5M3_Case#1: PbLi temperature (TLK1) at different radial position and bottom level.
(Zoom on 0 – 1 s).
Fig. 5.21 – L5M3_Case#1: PbLi temperature (TLK1) at different radial position and middle level.
(Zoom on 0 – 1 s).
Fig. 5.22 – L5M3_Case#1: PbLi temperature (TLK1) at different radial position and top level.
(Zoom on 0 – 1 s).
Time = 111 ms
Overall nodalization Zoom on S1_B reaction vessel Time = 361 ms
Overall nodalization Zoom on S1_B reaction vessel
Time = 589 ms
Time = 1 s
Overall nodalization Zoom on S1_B reaction vessel Time = 5 s
Overall nodalization Zoom on S1_B reaction vessel Time = 12 s
Overall nodalization Zoom on S1_B reaction vessel
5.2.4 Sensitivity calculations
5.2.4.1 Influence of different water mass injection
The first set of sensitivities are performed changing the amount of injected water. The results are compared with those obtained for the Case#1. The amount of injected water calculated by SIMMER-III does not exactly correspond to the values defined in the Test Matrix (Tab. 5.3) due to different density implemented in the code. The calculated pressure transient, mass flow rate of injected water, hydrogen produced by the chemical reaction, and maximum PbLi temperatures in the region that will be instrumented by thermocouples, are reported from Fig. 5.24 to Fig. 5.28.
The total amount of injected water influences:
• The mass flow rate trends, as a result of the SIMMER-III code modeling and the initial and boundary conditions imposed in the calculations (Fig. 5.26 and Fig. 5.27).
• The mass flow rate peak, instead, is not affected by the amount of injected water but only by the pressure difference between injection line and reaction vessel S1_B, coherently with the boundary conditions set in the calculations (see sect. 5.2.2).
• The hydrogen produced by the chemical reaction. Indeed, more water is injected, more hydrogen is generated, coherently with the stoichiometric theoretical value (Fig. 5.26 and Fig. 5.27).
• The value of the first pressure peak due to the water flashing is not affected by the amount of injected water, but only by the pressure of injector device breaking-up and the thermal-hydraulic conditions of the injected water (Fig. 5.24 and Fig. 5.25).
• The amount of the injected water, and therefore the enhance of the chemical reaction results also in the calculated PbLi temperatures (Fig. 5.28), which show higher peaks.
However, as already discussed in sect. 5.2.3.2, these temperatures are local hot spot calculated by the code at the interface between PbLi and water.
The main results of relevant parameters characterizing the sensitivity analyses are
reported in Tab. 5.7
Parameter Unit #1 #2 #3 #4 #5
P @ inj break bar 161 178 155 150 157
P peak bar 26.9 29.1 28.8 25.4 25.7
Mass flow rate peak kg/s 1.42 1.16 1.24 1.46 1.40 Injected water kg 0.106 0.054 0.162 0.216 0.269 Hydrogen production g 8.9 4.6 13.6 18.1 22.7 T PbLi max (local hot
spot)
°C 445 391 460 435 475
Tab. 5.7 – Sensitivity analyses: influence of amount of injected water - main parameters characterizing the calculations.
Fig. 5.24 – Sensitivity analyses: influence of amount of injected water - pressure trends in S1_B reaction vessel.
Fig. 5.25 – Sensitivity analyses: influence of amount of injected water - pressure trends in S1_B reaction vessel (focus on 0 – 1 s).
Fig. 5.26 – Sensitivity analyses: influence of amount of injected water - water mass flow rate at the injector device and hydrogen production.
Fig. 5.27 – Sensitivity analyses: influence of amount of injected water - water mass flow rate at the injector device and hydrogen production (focus on 0 – 1 s).
Fig. 5.28 – Sensitivity analyses: influence of amount of injected water – maximum PbLi temperature trends (focus on 0 – 2 s).
5.2.4.2 Influence of water temperature
The calculations #6 and #7 are performed changing the temperature of the injected water. The results are compared with those obtained for the Case#1. The calculated pressure transient, mass flow rate of injected water, hydrogen produced by the chemical reaction, and maximum PbLi temperatures in the region that will be instrumented by thermocouples, are reported from Fig. 5.29 to Fig. 5.32.
The temperature of the injected water influences:
• The value of the first pressure peak due to the water flashing, but not the pressure increase rate, neither the timing of the pressurization phase (Fig. 5.29 and Fig. 5.30); however, it is worth underlining that the first pressure peak might be affected also by the pressure of injector device breaking-up.
• The peak and the trend of the mass flow rate (Fig. 5.31), the total amount of injected water, and therefore the hydrogen produced by the chemical reaction (Fig. 5.31) due to the tests procedure are slightly influenced because the water density changes accordingly with its temperature.
• The temperature of the PbLi inside the reaction vessel S1_B (Fig. 5.32) is influenced by the initial water temperature.
The main results of relevant parameters characterizing the sensitivity analyses are reported in Tab. 5.8.
Parameter Unit #1 #6 #7
T PbLi @ SoT °C 330 330 330
T H2O @ SoT °C 300 285 325
P @ inj break bar 161 182 158
P peak bar 26.9 25.2 28.3
Mass flow rate spike kg/s 1.42 1.34 1.50
Injected water kg 0.106 0.110 0.101
Hydrogen production g 8.9 9.1 8.4
T PbLi max (local hot spot) °C 445 398 420
Tab. 5.8 – Sensitivity analyses: influence of temperature - main parameters characterizing the calculations.
Fig. 5.29 – Sensitivity analyses: influence of water temperature - pressure trends in S1_B reaction vessel.
Fig. 5.30 – Sensitivity analyses: influence of water temperature - pressure trends in S1_B reaction vessel (focus on 0 – 1 s).
Fig. 5.31 – Sensitivity analyses: influence of water temperature - water mass flow rate at the injector device and hydrogen production (focus on 0 – 1 s).
Fig. 5.32 – Sensitivity analyses: influence of water temperature - selected temperature trends (focus on 0 – 2 s).
5.2.4.3 Influence of the dimension of the orifice diameter
One of the parameters which mostly affects the simulation transients is the orifice diameter. Calculations #8, #9 and #10 are performed changing this dimension, in particular considering 1, 2 and 8 mm, respectively. The results are compared with those obtained for the Case#1. The calculated pressure transient, mass flow rate of injected water, hydrogen produced by the chemical reaction, and maximum PbLi temperatures in the region that will be instrumented by thermocouples, are reported from Fig. 5.33 to Fig. 5.36.
The dimension of the orifice diameter influences:
• The value of the first pressure peak due to the water flashing, and the pressure increase rate, up to the pressure of injection, and the timing of the pressurization phase (Fig. 5.33);
• The peak and the trend of the mass flow rate, largely different (Fig. 5.34);
• The total amount of injected water (Fig. 5.35), even though considering the integral value function of time, there is no difference;
• The hydrogen produced by the chemical reaction, particularly the rate of production, which depends on the injected water mass flow rate (Fig. 5.34). The total amount of hydrogen generation instead, which depends on the injected water mass, is not affected by the orifice diameter;
• The temperature of the PbLi inside the reaction vessel S1_B (Fig. 5.36). However, this behavior is influenced also by modeling and the choice of mesh cells. Indeed, greater is the orifice diameter, higher is the maximum temperature, but for a diameter equal to 8 mm, the maximum PbLi temperature is lower (in the mesh cells I=5-12, J=45-49 which model the instrumentation region). This trend is due to the water which is rapidly injected into the reaction vessel S1_B and interacts with PbLi in higher positions (outside the selected mesh cells).
The main results of relevant parameters characterizing the sensitivity analyses are
reported in Tab. 5.9.
Parameter Unit #1 #8 #9 #10
D orifice mm 4 1 2 8
P peak @ inj break bar 161 150 150 151
P peak bar 26.9 8.3 12.3 47.7
Mass flow rate peak kg/s 1.42 0.10 0.38 4.52
Injected water kg 0.106 0.105 0.106 0.107
Hydrogen production g 8.9 8.9 8.9 9.2
T PbLi max (local hot spot) °C 445 375 464 366
Tab. 5.9 – Sensitivity analyses: influence of orifice diameter - main parameters characterizing the calculations.
Fig. 5.33– Sensitivity analyses: influence of orifice diameter - pressure trends in S1_B reaction vessel (focus on 0 - 2.5 s).
Fig. 5.34 – Sensitivity analyses: influence of orifice diameter - water mass flow rate at the injector device and hydrogen production (focus on 0 – 3 s).
Fig. 5.35 – Sensitivity analyses: influence of orifice diameter – injected water (focus on 0- 2.5 s).
Fig. 5.36 – Sensitivity analyses: influence of orifice diameter – maximum PbLi temperature trends (focus on 0 – 2 s).
5.2.5 Summary of outcomes from first pre-test series The main outcomes can be summarized as follows:
• Mass of injected water influences the hydrogen production and the PbLi temperature. The maximum temperature and hydrogen production are reached with 269 g of injected water. They are 475°C and 22.7 g, respectively.
• The calculated mass of H2 ranges from 4.6 g to 22.7 g which correspond to a concentration of 70 % to 91 % respect to the initial mass of gas (1.99 g of Ar);
• The maximum mass flow rate is calculated as 4.52 kg/s in the case 8mm of diameter;
• The diameter of the orifice is correlated to the first pressure peak: a larger diameter corresponds to a higher peak. The maximum pressure is reached with an orifice of 8 mm and is 47.7 bar.
• The most relevant parameters chosen for the definition of test matrix are the temperature of water (from 285°C to 325°C) and the mass of injected water (from 50 g to 150 g). At least at first the orifice diameter will be kept to 4 mm, as in the reference case.
Final pressure in S1_B reaction vessel is not correlated to the considered parameters but
is affected by the need of injecting all water in the reaction vessel, thus keeping the
argon gas and the injection valve opened up to the pressure equalization (i.e. imposed in
the calculation).
5.3 Reaction/interaction model
This section regards the pre-tests related the second series of experiments, which will be executed to assess the thermodynamic and the chemical models of SIMMER-III. In this case, the injection line has been modified in order to consider the friction and the form losses. Friction losses are accounted adding a so called “can wall” and considering test section as physics, not only as a virtual wall. In this case injection line completely filled with water, and the injected mass of water is not known a priori. This procedure is similar to what was done in LIFUS5 past experiments.
5.3.1 Nodalization of LIFUS5/Mod3 facility
The SIMMER nodalization of the facility is essentially the same of the previous cases and is reported in Tab. 5.10. A grid of 23 radial cells and 81 axial cells has modelled. Injection line is modelled as in LIFUS5/Mod3 facility, which is a ½” Swagelock pipe, with internal diameter of 8 mm and external diameter of 12.85 mm. In this case the injection line is completely filled with water, because in this case the aim of simulation is not the validation against hydrogen production, but we are interested to the studying of the entire pressure/temperature transient. Therefore, a Coriolis flow meter is installed in the line, to measure the mass of water injected in the facility. Can wall is defined with the parameters ARCWIB and ASMINB(7) in SIM05 input file; ARCWIB represent a surface area per unit volume of right can wall and is calculated as
𝐴𝑅𝐶𝑊𝐼𝐵 = 2𝜋𝑟𝐿
𝜋(𝑅
2−𝑟
2)𝐿 = 316.45 𝑚
−1Where r is the internal radius of the injection line, R the external radius and L the length of injection line. ASMINB(7) is the volumetric fraction of the structural material. The facility domain is shown in Fig. 5.2. Black colour represents can wall, located in the injection line, and test section, placed in PbLi reaction vessel. Test section is simulated as cladding material, through the parameters ASMINB(4), set as 1, which represents the volumetric fraction, and the parameter RPINIB, set as 0.1, representing the intact fuel pin outer radius. Friction losses and pressure drops are the same of the previous cases and reported in Tab. 5.1, with the addition of those caused by the can wall.
The correspondence of main dimensions of LIFUS5/Mod3 facility preliminary design and SIMMER-III nodalization is reported in Tab. 5.10.
BFCAL tool is also used (Ref.[11]). The nodalization of the BFCAL tool is shown in Fig.
5.5.
5.3.2 Boundary and Initial Conditions
Boundary and initial conditions are the same of chemical interaction model and are
reported in Sec. 5.2.2.
Fig. 5.37 – L5M3_Case#1bis: nodalization of facility.
Can wall
Physical test section
PK(6,71)
LIFUS5/Mod3 modeling LIFUS5/Mod3 preliminary design Volume parameter cells dimension Volume parameter dimension
S1_B h 38-57 0.52 S1_B h 0.59
D 1-23 0.267 D 0.257
V - 0.0292 V 0.0286
Free gas h 58-61 0.0682 Free gas V TBD*
D 1-23 0.267
V - 0.00382
Inj_device h 38-41 0.106 Inj_device h 0.08
D 1-3 0.01388 D 1/2’’
Dorifice 4 Dorifice TBD*
Inj_line h 1-37 7.951 Inj_line L ~ 8.0
D 1-3 0.01285 D 1/2’’
*TBD accordingly with pre-test analyses
Tab. 5.10 - LIFUS5/Mod3 nodalization: Case#1bis.
5.3.3 Analysis of pre-test calculation
Time trends of parameters are very similar to those described in reference case L5M3_Case#1. In this case the transient can be subdivided into four main phases (as in the previous set of pre-tests). For this reason, only the main differences will be highlighted in the following description.
Phase 1 [from onset of valve opening to 572 ms]: water injection line pressurization.
This phase is almost identical to those of chemical interaction model. There is a little delay in injector breaks-up, due to the different pressurization time to overcome the rupture pressure. This fact can be explained considering the pressure drops owing to the can wall. In Tab. 5.11 main properties of injected water are reported.
# Conditions Value Note
1 Pressure at the injector 150 bar
The thermal-hydraulic conditions are calculated by
the code @ t = 572 ms after SoT at the mesh cell (1,41)
2 Temperature of injected water 309°C
3 Void fraction of injected water 0.737 [-]
4 Time of injector rupture 572 ms
Tab. 5.11 - L5M3_Case#1bis: thermal-hydraulic conditions of water at the injection time.
Phase 2 [from 572 to 630 ms]: coolant flashing and first pressure peak.
In this phase, a sudden depressurization in injection line can be observed (Fig. 5.38, Fig.
5.39 and Fig. 5.40). This behavior can be justified considering again the distributed pressure drops on the can wall. Temperature of water shows a maximum value in correspondence of impact with injection device (Fig. 5.40).
Pressure reaches a maximum value of 28.6 bar in the reference cells of PbLi zone (23,46). The pressure trends in the cover gas zone (9,60) and S3 expansion line (6,71) are shown in Fig. 5.42 and in Fig. 5.43.
The calculated mass flow rate of the injected water (Fig. 5.44) presents a spike at 1 kg/s.
PbLi temperature in S1_B reaction vessel in different radial and axial position are reported in Fig. 5.50 and Fig. 5.52, accordingly to the design of the test section which will be installed inside the vessel to support the thermocouples.
Phase 3 [from 630 ms to 1300 ms]: pressurization due to water and gas injection and hydrogen generation, up to pressure equilibrium.
In Fig. 5.43 is possible to observe that the pressure in cell (6,71), corresponding to the rupture disk on the S3 dump line, has an oscillatory trend: this can be explained considering the “piston effect” of the PbLi alloy on the cover gas. Furthermore, the pressure reaches a value just slightly higher than 100 bar, therefore much lower than rupture disk breaking pressure (190 bar at 400°C). During this phase PbLi temperature reaches a peak value of 555°C in the test section zone. In Fig. 5.53 are reported the temperature maps during the transient.
Phase 4 [from 1300 ms to EoT]: system pressure stabilization.
This phase is longer than in L5M3_Case#1, in fact, even if the pressure reaches a stable value, injection of water continues until about 4 s from the SoT (Fig. 5.44 and Fig. 5.45), and hydrogen is produced as a consequence (Fig. 5.44 and Fig. 5.46). At the end of the transient the mass of produced hydrogen is 31.3 g and mass of injected water is 0.371 kg, as reported in Tab. 5.13. In this phase, PbLi might drop down into the injection device, as depicted in Fig. 5.47. Mass of vapor and water are completely consumed at about 7s from SoT (Fig. 5.48 and Fig. 5.49).
# Phase Time span [ms]
1 Water injection line pressurization 0 – 572
2 Coolant flashing and first pressure peak 572 – 630
3 Reaction vessel pressurization 630 – 1300
4 System pressure stabilization 1300 – EoT
Tab. 5.12 - L5M3_Case#1bis: phenomenological analysis.
Parameter Unit #1
T PbLi @ SoT °C 330
T H2O @ SoT °C 300
P @ inj break bar 150
P peak bar 28.6
Mass flow rate spike kg/s 1
Injected water kg 0.371
Hydrogen production g 31.3 T PbLi max (local hot spot) °C 555
Tab. 5.13 - L5M3_Case#1bis: main parameters characterizing the calculations.
Fig. 5.38 - L5M3_Case#1bis: calculated pressure (PK) in the injection line at different positions (Zoom 0 – 5 s).
Fig. 5.39 - L5M3_Case#1bis: calculated pressure (PK) and temperature of water (TLK3) in the injection line (Zoom 0 - 2 s).
Fig. 5.40 - L5M3_Case#1bis: calculated pressure (PK) and temperature of water (TLK3) at the injector device. (Zoom 0 - 2 s).
Injector breaks-up
Injector breaks-up
Fig. 5.41 - L5M3_Case#1bis: calculated volume fraction of water (ALPLK3) and vapor (ALPGK) at the injector device (Zoom 0 – 2 s).
Fig. 5.42 - L5M3_Case#1bis: pressure trends in S1_B reaction vessel.
Injector breaks-up
Fig. 5.43 - L5M3_Case#1bis: pressure trends in S1_B reaction vessel (Zoom 0.4 – 1.5 s)
Fig. 5.44 - L5M3_Case#1bis: water mass flow rate and hydrogen production (zoom 0 – 6 s).
Phase 1 Phase 2 Phase 3
Phase 4
Fig. 5.45 – L5M3_Case#1bis: integral water mass flow rate.
Fig. 5.46 - L5M3_Case#1bis: calculated hydrogen generation (MASFPG) in different macro-regions (S1_B, Cover Gas, Extraction lines).
Fig. 5.47 - L5M3_Case#1bis: PbLi mass flow rate (BFLUXF) through the injector.
Fig. 5.48 - L5M3_Case#1bis: mass of water (MASN1) in different macro-regions (zoom 0 - 7 s).
Fig. 5.49 - L5M3_Case#1bis: mass of vapor (MASN2) in different macro-regions (zoom 0 - 7 s).
Fig. 5.50 - L5M3_Case#1bis: PbLi temperature (TLK1) at different radial position and bottom level (Zoom on 0 – 3 s).
Fig. 5.51 - L5M3_Case#1bis: PbLi temperature (TLK1) at different radial position and middle level (Zoom on 0 – 3 s).
Fig. 5.52 - L5M3_Case#1bis: PbLi temperature (TLK1) at different radial position and top level (Zoom on 0 – 3 s).
