dynamic simulation of high transients in a forced flow

21th sept. 2021 - Dynamic simulation of high transients in forced flow Supercritical helium loop for the sizing of an experimental set-up dedicated to the study of loss of vacuum – CHATS 2021 – S. SHOALA

DE LA RECHERCHE À L’INDUSTRIE

Commissariat à l’énergie atomique et aux énergies alternatives - www.cea.fr

Dynamic simulation of high transientsin a forced flow Supercritical heliumloop for the sizing of an experimentalset-up dedicated to the study of loss ofvacuum

21th september 2021

Sulayman SHOALA, Eric ERCOLANI, Jean Marc PONCET, Christine HOA, Frederic AYELA


Safety in cryogenics

2

Insulating vacuum failure is a major accident scenario- Large break on the Cryostat- Outside atmospheric air rushes into the vacuum space and condenses on cold surfaces- Very high heat load transferred to the cryogenic fluid

Cryogenics devices have to be protected against overpressure- By using safety valves or rupture disks

For instance, safety relief devices are sized by using a constant heat flux value :- Helium tank in diphasic discharge, without insulation: 3.8W/cm² - Helium tank in supercritical discharge, without insulation: 2W/cm²

Reference values of heat flux have just been measured for tank

Question: Which value of heat flux has to be considered for supercritical helium flowing in pipe?

Objective: Develop an experimental device to perform experimental measure of the heat flux

Iter’s Cryolines and cold boxes (iter.org)

CERN: [email protected] cold box (cds.cern.ch)


The HELIOS loop

3

HP1HP2

SCP

V1 V2V3 V4

V5

HELIUM BATH

REFRIGERATOR sHe 16 bar, 4,7KgHe 1,15 bar, 4,4K

HPBP

Test section

HELIOS is composed of two main parts:- 1 supercritical helium loop- 1 saturated helium bath

2 heat exchangers thermal coupling bath/loop

Bath connected to the refrigerator ( by HP & BP)- Cooling power available: 320 W

Why Helios is used ?- Supercritical helium loop is already available- Cold centrifugal pump imposes forced convection- Saturated helium bath Interface with refrigerator +

Thermal buffer

SupercriticalHe loop


The test section

4

A test section (TS) will be included in the helium loop of HELIOS

Vacuum loss in a restricted area thanks to a secondary vacuum vessel- Vacuum break with N2 at atmospheric pressure/temperature

Vertical upward supercritical helium flow

How to generate a « controlled » loss of insulating vacuum with HELIOS ?

Hypothesis for the sizing of the experiment :• ℎ𝑐𝑜𝑛𝑑𝑒𝑛𝑠𝑎𝑡𝑖𝑜𝑛 ≫ ℎ𝑓𝑜𝑟𝑐𝑒𝑑 𝑐𝑜𝑛𝑣𝑒𝑐𝑡𝑖𝑜𝑛

• Wall temperature imposed to 77K • Supercritical He flow: 20g/s• Dittus Boelter correlation

Heat load received by supercritical He ~ 2,7kW (4,25W/cm²)

Nitrogen tank1 bar – 300K

He 4 bar4,4 K

Secondary vacuum vessel

Pressure and temperature

sensors

Wall temperature sensors

𝐿=1𝑚

d = 20 𝑚𝑚

Diagram of the test section

D =300 𝑚𝑚

Quick opening valve


Sizing of the new HELIOS loop

5

Orders of magnitudes :- Estimated power transmitted to supercritical He in the test section : 2700W - Cooling power available : 320W

𝑷𝒔𝒆𝒄𝒕𝒊𝒐𝒏 ≫ 𝑷𝒓𝒆𝒇𝒓𝒊𝒈𝒆𝒓𝒂𝒕𝒐𝒓

Thermal buffer need to be sufficient to store the energy injected during the loss of vacuum

Last version of HELIOS :- Volume of the loop : ~30L- Volume of the bath : 310L

Not enough to keep the maximal pressure under the safety valve opening pressure (Pmax = 9 bar)

Sizing parameters :- Volume of the loop- Transit time between the test section and heat exchanger- Thermal exchange loop/bath- Volume of the bath (thermal buffer)

HELIOS has to be adapted to this new experiment


Geometry of the loop

6

The pressure increases due to the stored energy

𝑈𝑓 = 𝑈𝑖 +ሶ𝑞 ∗ Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡𝑚𝑙𝑜𝑜𝑝

Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡 =𝜌𝑉𝑇𝑆→𝐻𝑃2

ሶ𝑚

Volume distribution- Remove volume between TS and HP2 to reduced Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡- Cold volumes to be added : HP2 – SCP and HP1 – TS

HP1HP2

SCP

V5Test section

Heatingሶ𝒒

Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡


Geometry of the loop

7

The pressure increases due to the stored energy

𝑈𝑓 = 𝑈𝑖 +ሶ𝑞 ∗ Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡𝑚𝑙𝑜𝑜𝑝

Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡 =𝜌𝑉𝑇𝑆→𝐻𝑃2

ሶ𝑚

Volume distribution- Remove volume between TS and HP2 to reduced Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡- Cold volumes to be added : HP2 – SCP and HP1 – TS

In the new loop design, the volume reaches 170L Δ𝑡𝑡𝑟𝑎𝑛𝑠𝑖𝑡

HP1HP2

SCP

V5Test section


A new heat exchanger has been added (HP3) :- Power injected in the TS >> Heat load used for the sizing of HP2- The new heat exchanger is placed between TS and HP2- 3 parallel copper coils immersed in the bath

The by-pass line :- Large part of the flow by-passes the TS- Improves the additional heat exchanger (HP3) performances- The cold circulator can be used on an extended operation field

The bath has been connected to the helium recovery system - Cold He gas can be evacuated from bath without passing through the

refrigerator

The bath volume was increased from 310L to 510L :- Space to include the new heat exchanger- Increased thermal storage

Additional modifications to the HELIOS loop

8

HP1HP2

SCPV1

V2

V3

V5

HELIUM BATH

REFRIGERATOR sHe 16 bar, 4,7KgHe 1,15 bar, 4,4KHPBP

Test section

HP3

V4

V6

V7V8

He recovery

ሶ𝒎𝒕𝒐𝒕 :80g/s

ሶ𝒎𝑩𝑷 :60g/s

ሶ𝒎𝑻𝑺 : 20g/s


Some illustrations

9

3,4m

2,4m3D view of the experimental setup

Photography of the 3 heat exchangers Photography of the bath


The CATHARE code

10

Code for Analysis of Thermal Hydraulics during Accident and for Reactor safety Evaluation

CATHARE : Reference tool for safety evaluation of pressurized water reactor, developed by CEA, FRAMATOME, EDF and IRSN since 1979

Devoted to thermal hydraulic simulation of multiphase flow transient at the system scale

2-fluids and 6 equations model :- 3 equations for each phase: mass, momentum and energy balance- 6 main hydraulic variables: 𝑃, 𝐻𝑙, 𝐻𝑔, 𝑎, 𝑉𝑙, 𝑉𝑔- Taking into account all the flow regimes, mass and heat transfer between each phase or fluid and structure ( ~200

closure laws )

The default fluid is two-phase water but other options are available:- Thermodynamic and transport properties of more than 100 fluids are available including helium (REFPROP database

developed by the NIST)


Dynamic modelling of the HELIOS facility

11

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8

ሶ𝒎𝒍𝒊𝒒 +

ሶ𝒎𝒈𝒂𝒔

imposed

Pext = 1,15 bar

HP1HP2

SCPV1

V2

V3

V5

HELIUM BATH

REFRIGERATOR sHe 16 bar, 4,7KgHe 1,15 bar, 4,4KHPBP

Test section

HP3

V4

V6

V7V8

He recovery

Pext = 1,1 bar

Some simplifications are made on the modelling, using relevant boundary conditions



12

The loop is modelled with 2 distinct circuits- Thermal coupling between bath and loop- Monophasic flow in the loop (3 bar, 4.4K)- 2 phase flow in the bath (1.1 bar, 4.3K)

Using closure law for water except for nucleate boiling (Kutateladze)

Heat exchanger sub-module

Pext = 1,15 bar

ሶ𝒎𝒍𝒊𝒒 +

ሶ𝒎𝒈𝒂𝒔

imposed

0D

1D

0D

0D compressor

1D pipe

1D heat exchanger

Comparison performed on a former experiment (cf article)

HP1HP2

SCP

V5Test section

HP3

V6

V7

0D valve HP1HP2

V1

HELIUM BATH

Path around HPHP3

V8

Pext = 1,1 bar



13

Transient scenario :- Δ𝑃𝑆𝐶𝑃 = 0,85 𝑏𝑎𝑟 (regulation on V7) / ሶ𝑚𝑆𝐶𝑃 = 75 𝑔/𝑠- ሶ𝑚𝑇𝑆 = 20𝑔/𝑠 (regulation on V5)- At t=50s : loss of vacuum

Loss of vacuum : TS’s external wall temperature imposed to 77K

Bath isolated from the refrigerator, discharge in the He recovery- V1 close- V8 fully open

0

20

40

60

80

100

0 50 100 150 200 250 300 350

Tem

per

atu

re [

K]

Time [s]

External temperature TS

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8

ሶ𝒎𝒍𝒊𝒒 = 𝟎

ሶ𝒎𝒈𝒂𝒔 = 𝟎Pext = 1,1 bar

Regulation only during the first 50s



14

Maximal power transmitted to the fluid : 2780 W Take few seconds to reach its maximum

- Mass flow decreases in the section due to the expansion of the heated helium

0

500

1000

1500

2000

2500

3000

0 50 100 150 200

Po

we

r [W

]

Time [s]

Power transmitted to the fluid

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8



ሶ𝑸𝑻𝑺



15

P2 is the maximal pressure in the loop P2 << Pmax (9bar)

2

2,5

3

3,5

4

4,5

5

5,5

6

6,5

0 50 100 150 200 250 300 350

Pre

ssu

re [

bar

]

Time [s]

Pressure evolution at the Supercritical Centrifugal pump Inlet/Outlet

P1

P2

P1 P2

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8





16

No reversed flow is observed CATHARE predicts a similar inlet and outlet mass flow It seems possible to obtain an enthalpy balance on the TS with only one flow meter placed at the upstream of the section

ሶ𝒎𝟏

ሶ𝒎𝟐

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8



0

5

10

15

20

25

30

0 50 100 150 200 250 300 350

Mas

s fl

ow

[g/

s]

Time [s]

Test section inlet/oulet He mass flow

ṁ1

ṁ2



17

For a saturated bath with constant pressure :- ሶ𝒎𝒗𝒂𝒑 is proportionnal to the power received by the bath

Only ሶ𝒎𝟑 can be measured : ሶ𝒎𝒗𝒂𝒑 = ሶ𝒎𝟑/(𝟏 −𝝆𝒗

𝝆𝑳)

Due to pressure loss (V8), the flow to reach its nominal value after a delay (𝑃𝑏𝑎𝑡ℎ increases from 1,1 to 1,17 bar)

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8



0

20

40

60

80

100

120

140

0 50 100 150 200 250 300 350

Mas

s fl

ow

[g/

s]

Time [s]

He mass flow ṁ3 evacuated from the bath ሶ𝒎𝟑

ሶ𝒎𝒗𝒂𝒑



18

ሶ𝑚3 can be used to calculate heat power received by the bath

ሶ𝑄𝑣𝑎𝑝 = ሶ𝑄𝑇𝑆 + ሶ𝑄𝑆𝐶𝑃 + ሶ𝑄𝐿𝑜𝑠𝑠 = ሶ𝑚3 ⋅𝐻𝐿

1−𝜌𝑣𝜌𝑙

= ሶ𝑚𝑣𝑎𝑝 ⋅ 𝐻𝐿

To know the heat flux transmitted at the test section, all additional energy sources must be quantified

HP1HP2

SCP

V1

V5

HELIUM BATH

Test section

HP3

V6

V7V8


ሶ𝒎𝒗𝒂𝒑 = 𝟎Pext = 1,1 bar

0

500

1000

1500

2000

2500

3000

3500

0 50 100 150 200 250 300 350

Po

we

r [W

]

Time [s]

Total exchange power(HP1+HP2+HP3)Q ̇_vap

ሶ𝒎𝟑

ሶ𝑸𝒗𝒂𝒑

ሶ𝑸𝑻𝑺

ሶ𝑸𝑺𝑪𝑷

ሶ𝑸𝑳𝒐𝒔𝒔


Conclusion

19

A new design has been defined for the experimental device called HELIOS, in order to perform loss of vacuum on a dedicated test section and carry out experimental measures of heat flux

CATHARE has been used to modelled HELIOS with its new design - The thermal hydraulic behaviour of the loop has been investigated- Sizing criteria is respected with the new design- Experimental methods to establish the heat flux transmitted the fluid with sufficiently low incertitude has been

investigated

The experimental set up is under construction :- Bath and the new heat exchanger are in place- Construction of the loop is expected to begin in October

First experimental campaign expected in 2022

Experimental result will be compared with CATHARE


Commissariat à l’énergie atomique et aux énergies alternatives - www.cea.fr

Thank you for your attention

dynamic simulation of high transients in a forced flow

Documents