Observations on possible factors affecting LAr purity on the 50kton TPC from the viewpoint of the internal walls

Alejandro S. Díaz

Chemical Engineering Department, Surface Analysis and Fluid Interaction Laboratory (ASIF), U of Concepción, Chile

First FLARE Collaboration Meeting, Fermilab, Chile, November 4-6, 2003.

SummarySummary

In the present work, it is presented a first approach to answer some In the present work, it is presented a first approach to answer some questions about the construction of a big scale LAr detector, under questions about the construction of a big scale LAr detector, under the viewpoint of the internal surfaces’ influence:the viewpoint of the internal surfaces’ influence:

-What factors may affect scaling up a detector to a very large tank?

-What auxiliary experiments/measurements are necessary to provide an ironclad proof that it will have the necessary purity?

-How well is the whole process understood?

ICARUS standard procedure vs FLARE 50ktonICARUS standard procedure vs FLARE 50kton

ICARUS

HNO3 wash Degassing of Assembly Hot air bakeout? N2 Purge Ar Purge LAr fillingWater rinse small parts (120 ºC)Drying

FLARE

HNO3 wash Degassing of Assembly Evacuation to LN2 Cooling LAr fillingWater rinse small parts 10-3 mbarDrying

Outgassing Outgassing

No vacuum pre-degassing of the detector:

-Not evacuating the vessel will influence the residual oxygenin the chamber volume when purging

-It has been observed the purification of the liquid argon via sorption of impurities on the walls, under the condition of previousdegassing.

-It will affect all internal walls' ability to adsorb impurities.One still can expect impurities to adsorb on the walls due to the low temperatures but gas exposed parts will slowly outgas (due to the lower total concentration of the bulk phase)

Outgassing, ICARUS dataOutgassing, ICARUS data

Outgassing, calculated total massOutgassing, calculated total mass

Total outgassed mass, molecular regime

0

1

2

3

4

5

6

7

8

9

10

50 70 90 110 130 150 170 190Time (h)

Ou

tgas

sed

mas

s (m

ol)

First order fit to T300 Evacuation run data (1) Isothermal evacuation assumed.

Tank volume: 260 m3Total Estimated exposed surface: 2000m2Temperature: 293KThird stage: p = 2E-4 mbar to 1E-4 mbarTurbo Pumping started at 10E-1 mbarPumping speed: 2000 l s-1Negligible external leaks assumed

Estimated from T300 data

Asymptotic value: 9.2 mole

Outgassing, total mass by componentOutgassing, total mass by component

Oxygen Nitrogen WaterActive volume: 0.10 0.27 2.32 kgTank surface: 0.02 0.04 0.38 kg

0.12 0.31 2.70 kg total

Outgassables:

-ICARUS run tests have shown that the main adsorbed component on internal surfaces is water (90% mass fraction).Hence, the need of a low temperature bake.

-Conservative rough estimates of impurity content of the 50kton calculated extrapolating T600 outgas data (caveant):

-If adsorbed water desorbs, it will imply a change in the load of the purifying systems.

Outgassing factorsOutgassing factors

- Adsorption enthalphy is negative, enthropy is positive, but depends on chemical potential (which depends on concentration)

higher concentration on surface = higher barrier to overcome for adsorption

-It is not possible to say (a priori) which factor will be more weighting (P and T vs delta C) (But P and T could win, imho).

-Some data on the Ar-H2O-N2-O2 system liquid-adsorption equilibrium in each exposed material should be needed.

-The possibility of a low temperature bakeout of the entire detector assembly should be evaluated.Hot air, 120ºC or similar.

adsadsads STHG

Fluid dynamics in theFluid dynamics in the vessel:vessel:

-A smaller vessel (T600) could be easier to homogeneize in the fluid composition.But Residence Times Distribution expected to be close to a countinuously stirred tank (Para, Tang). How close? ----> RTD

Fluid dynamics in theFluid dynamics in the vessel:vessel:RTD will affect the gas purging time and LAr recirculation time required to obtain a desired purity.

-Fluid speed distribution will determine the mass transfer process between the walls and the bulk liquid (boundary layer, eddy currents), given the slow thermal diffusion of oxygen in LAr (D=2·10-5 cm2s-1).

C

C*

F

2/12/1

Ux

5)(Ref)x(

*

z CCD

zCD

F

Fluid dynamics in theFluid dynamics in the vesselvessel:

-More detailed simulations of the inside-tank hydrodynamics could be done to clarify this.

What auxiliary experiments/measurements are necessaryWhat auxiliary experiments/measurements are necessary ??

A lab scale experiment, (e.g. 50L) should have as objectives:

-Study the new procedure influence on the Argon purity and consequently, on the purification system load-Quantify how the adsorption capacity is affected by the atm O2-N2

exposure stages for a given material.

Strong proof of feasibility will be given by the construction of a pilot scale system, with comparable hydrodynamic and thermal behavior, exposed surfaces S/V ratio and chemical composition.

ExperimentsExperiments

A small volume cryostat, ~1 m3, with a purity monitor, purification system (filter beds, cryopumps, piping).

Vessel with similar material surface proportions and similar (geometrical) surface to volume ratio to the projected detector.

Do we know well what is happening?Do we know well what is happening?

It is necessary to analyze three aspects:

C

C*

F

C*=kC C*?

Transport Interphase SLE Surface

Adsorption Isotherms

Data needed

-Surface Area

-CompositionHydrodynamics

Conclusions

-The main difference between FLARE and ICARUS will be the (non) outgassing of the walls (from the viewpoint of walls and LAr purity )

-The amount of outgassables can be quantified extrapolating ICARUS data.

-Equilibrium data of the H20-Ar-N2-O2 system is needed. (Thermo or real-world)

-A “low temp bake” of the detector assembly would help to create a high purity environment

-Hydrodynamics determine walls’ mass transfer.

-Internal walls behavior is an important factor for the working of the detector

LAr purification via adsorption on the walls

Argon purity behavior in a lab scale 10-4 mbar outgassed vessel, without purification systems (Bettini et al, 1991).

Outgassing, total mass by componentOutgassing, total mass by component

)1(·Sg·xAm in

iiitog

Assuming that liquid submerged surfaces will not outgas, an estimate of the outgassable material can be obtained as:

Where the sub index i indicates the material ‘i’,

At is the total exposed area [m2], xi is the fraction of material i on the overall area,

2

2i

i m

m

At

Ax

i is the roughness of the material i,

Area lGeometrica

Area Surfacei

2

2

m

m, Sgi is the specific mass capacity

of material i in

2m

mol,i is the submerged fraction of the material i.

area submerged

area exposed

2

2

m

m

The parameters At , xi and i are obtained directly from construction plans data, i from a surfacecharacterization of the used materials, Sgi can be obtained by fitting degas data from the materials.