TRITIUM REMOVAL BY LASER HEATINGC.H. Skinner, C. A. Gentile, G. Guttadora, A. Carpe,

S. Langish, K.M. Young, M. Nishi(b), W. Shu(b), N. Bekris(c)

(a)Princeton Plasma Physics Lab, Princeton NJ 08543, USA(b)Tritium Engineering Laboratory, JAERI, Ibaraki, Japan

(c) Tritium Laboratory, Karlsruhe, FRG.

Tritium removal from plasma facing components is a serious challenge facing next step magnetic fusion devices that use carbon plasma facing components. The long term tritium inventory for ITER-FEAT is limited to about 350 g, mainly due to safety considerations. It is potentially possible that the inventory limit could be reached after a few weeks operation, requiring tritium removal before plasma operations can continue. Techniques for tritium removal have been demonstrated in the laboratory, and on tokamaks but they are slow and generally involve oxidation which will decondition the vessel walls (requiring additional time devoted wall conditioning) and generate undesirably large quantities of HTO.

A novel laser heating technique has recently been used to remove tritium from carbon tiles that had been exposed to tritium plasmas in TFTR. A continuous wave Nd laser operates at powers up to 300 watts. The beam is directed by galvonometer driven scanning mirrors and focussed on the tile surface. The surface temperature is measured by an optical pyrometer. The tritium released is measured by a ionization chamber and surface tritium measured by an open walled ion chamber. Any changes in the laser irradiated surface are monitored with a microscope. To date tritium has been released in air and argon atmospheres and surface temperatures up to 2,300 C have been achieved. We will present measurements of the removal of tritium as a function of the laser intensity, and scan rate. Potential implementation of this method in a next step fusion device will be discussed.

Presented at 6th International Conference on Tritium Science and Technology, Tsukuba, Japan Nov. 11-16th

• Next decade offers prospect of construction of next-step DT burning tokamak(s).

• Plasma material interactions will scale up orders of magnitude with increase in stored energy and pulse duration (bigger change than core plasma parameters).

• Tritium retention in machines with carbon plasma facing components will become significant constraint in plasma operations.

• Techniques for rapid efficient removal of tritium are needed.

0

100

200

300

400

500

600

0 50 100 150 200

Number of ITER pulses (~400 s/each)

10 g-T/pulse

5 g-T/pulse

1 g-T/pulse

JET DTE-1equivalent

ITER-FEAT Modelling predictions

Brooks et al.: 2-5 g-T/pulse

Precautionary operating limit for

mobilisable in-vessel T inventory

ITER plans to install CFC divertor with option to switch to more reactor relevant all-W armoured targets prior to D-T operation.

Change depends on:

• frequency and severity of disruptions,

• success achieved in mitigating the effects of T co-deposition.

ITER

• In TFTR, several weeks were needed for tritium removal after only 10-15 min of cumulative DT plasmas

– Future reactors with carbon plasma facing components need T removal rate >> retention rate

• Heating is proven method to release tritium but heating vacuum vessel to required temperatures (~350 C) is expensive.

• Present candidate process involves oxidation, requiring lengthy machine re-conditioning and expensive DTO processing.

• But – most tritium is codeposited on the

surface – only surface needs to be heated.

• Modelling showed lasers could provide

required heating.

400500600700800900

1000110012001300140015001600170018001900200021002200230024002500260027002800

0 0.01 0.02 0.03 0.04 0.05Time (s)

Temperature vs. time at different depths intopyrolitic perp. under 3,000 w/cm2 for 20 ms.

(a) Pyro perp.

surface

20µ

50µ

100µ

200µ

H E A T P U L S E

Heat transfer modelling shows a multi –kw/cm2 flux for ≈ 20 ms heats a 50 micron co-deposited layer to 1,000-2,000 K, appropriate for tritium release

10 cubes cut from TFTR tiles exposed to DT plasmas and irradiated w/laser in Ar atmosphere.

Microscope images taken before and after laser irradiation.

Vary raster pattern, laser power, laser focus, scan speed, atmosphere (air/Ar)...

Measure temperature, change in surface appearance, tritium release....

ScanningMirrors

LaserBeam

PYROMETER

Lensesfuzzy | shinyappearance

line 1line 2

5

3

6

example:: CUBE 2B 6-zone raster pattern.line spacing 0.5 mm = pyro. view 0.7 mm

4

Videos

Temperature ?3,4 @ 91W,6,5 @ 242W

Complete flexibility in raster patterns Nd:Yag laser, continuous wave 325 watt.

Computer programable laser scanning unit

Fast, high spatial resolution pyrometer.

Digital microscope, still or video capability

Tritium measured ion chambers & Differential Sampler.

Computer: Laser Control & Data Acquisition

Q Mark Scan Head

TFTRtile

Vacuum ChamberRemoved for photo

Nd laser power only 6 w to avoid camera damage (300 w available)

TFTR DT tile cube KC17 2E in air at 200 mm/s.

7/8” cube cut from TFTR tritiated tile inside chamber.

(KC17 2E)

Co-deposit gets hotter as it is less thermally conductive.

Laser 242 w, 2000 mm/s, ‘soft’ focus

0

500

1000

1500

2000

2.15 2.155 2.16 2.165time s

co-deposit side1770 C

cut side1080 C

cube KC22 6E

Thermal response:

Cube KC17-2B: Insert shows laser interaction attenuated by ND5 filter. Note much stronger interaction with codeposit on left.

Laser power 242 w, 50 mm/s Temperature: 1841 C left, 1181 C rightT release: 2.1mCi left, 1.1mCi right

≈10 mm

laser power

zero

29 w

91 w

242 w

cube KC17 3C

scan speed 50mm/s

soft focus, 4 zones:

Microscope x1

Zoom 7mm wide 45 deg view before laser

note surface granularity

Zoom 7mm wide 45 deg after 242 w

significant ablation

cube KC22 6E, laser power 242 w, soft focus, max temp. 1770 C,scan speed 1000 mm/s, 10 ms > 500 C,

images before &

after 18 mCi release

x1

x4

Model predicts heat penetrates ≈ 50 micron in 10 ms

Heat penetration depth with 1000 mm/s scan speed matches co-deposited layer

Tritium measured by FemptoTech ion chamber in closed loop system

Surface tritium on cube surface measured with open wall ion chamber,

Chamber contamination measured with swabs,

CUBEthermo-coupleVacuum Pump

Barytronpressure

gauge

Ar/air input

circulatingpump

flowmeter

60 micronfilter

Ion Chamber

15 micronfilter

Nd laser

Differential AtmosphericTritium Sampler (DAT)used for one expt.

Pdcatalyst

mol.sieve

mol.sieve

HTO trapped HTtrapped

Irradiate left 1/2 of cube KC22 1E in air atmosphere, pump & purge,then fill with Ar, and irradiate right 1/2. Laser 242 w, 50 mm/s, soft focus.

First measure tritium with ion chamber in closed loop, then add molecular sieve envelopes in DAT to determine HTO/HT mix

AtmospheremCi

(Femptotech)mCi HTO

(DAT)mCi HT(DAT)

air 8.3 4.8 0.7

argon 6.9 0.99 not measured

Not possible to measure HT in DAT without oxygen.

RGA shows Ar @ 99%, - possibly trapped H2O in tiles

- Note: not an issue with tiles inside operating tokamak.

Femptotech electrically calibrated by manufacturer only

Swabs show ~ microCi contamination of chamber.

~ 20 ms heating to > ~ 1500 C gives good tritium release with minimal change in surface (yellow area)

- how much tritium is left behind ?

■

■

■

■

■■

■

■

■

■

■

0

1

2

3

4

5

6

0 1000 2000degrees C

Tritium release vs. temperature

0

1

2

3

4

5

6

0 50 100 150 200 250m

Ci

duration > 500 C (msec)

Tritium release vs. scan speed

Cube 6E Laser Bake in air

0

10

20

30

40

1 11 21 31 41 51 61 71 81 91

time < 45 mins >

mCi

0

100

200

300

400

500

600

degrees C

Temp. ->

T released230 - 450 C < -

laser power 40 w, 80 w , 100 w, 80 w

• First Nd laser scan...

• Then Nd laser bake:

– rotate cube so that cut side faces laser, remove 1 lens to defocus laser. 100 w stationary laser beam heats cube to > 400 C for 40 mins in air to oxidise codeposit.

Two experiments so far:

• 1) ‘soft’ focus: – 46 % of total tritium released by scan

with almost no effect on surface

• 2) ‘hard focus’ - cube @ focal plane– 84% of total tritium released with

minor changes on surface.

Conclude:major part of co-deposited tritium can be released by scanning laser.

• Time needed to scan ?– 30 MJ required to heat top 100µm of 50

m2 area. - corresponds to output of 3kW laser for only 3 hours !

• Nd laser can be coupled via fiberoptic

• Potential for oxygen free tritium release in operating tokamak

– avoid deconditioning plasma facing surfaces

– avoid HTO generation(HTO is 10,000x more hazardous than T2 and very expensive to reprocess)

• Tritium removal by laser heating demonstrated.– no oxygen to decondition PFC’s – no HTO to process

• Method scalable to next-step device• Further optimization planned

* * * BONUS from Nd laser work * * *

Heating by continuous wave laser mimics heat loads in

transient off-normal events in tokamaks.

Opens new technique for studying key issues for

next step devices:

erosion by brittle destruction.

particulate (dust) generation.

Preprint: PPPL reports 3603, 3604 available from http://www.pppl.gov/pub_report/