Electron linacs: From the laboratory to the factory floor - PowerPoint Presentation

Slide1

Electron linacs:From the laboratory to the factory floor

CLIC Workshop

CERN

David Brown, Mevex Corporation

February 2014

Slide2

Electron linacs – workhorses in many fields

Cross-linking/curing

Medical therapy

Industrial imaging/inspection

Security applications

Medical device sterilization

Gemstone treatment

Semi-conductor irradiation

Mining applications (GAA/PAA)

Medical isotope production

Vaccine production

Curing of composite materials at operating temperature

Food irradiation for safety and shelf-life extension

Quarantine/Phytosanitary treatments for fruits

Slide3

A bit of information about Mevex

Incorporated in 1987

Privately held, family company

Organic growth / Self-financing

40 employees total: Canada, Sweden, Belgium, Thailand, France

Core Technology:Accelerator structuresPeak surface field strengths up to 100MV/mCompact S-Band structures (30MV/m average – unloaded)High power industrial linacs (15MV/m average – unloaded)Pulsed power and RF systemsControls and monitoringRadiation calculations and safety systems

Slide4

Mevex installed base summary…

Applications

Acc

Mod

Beams

Energy

Gemstones

2

2

1

22

Contract irradiation

1

1

1

10

Medical product research

1

1

1

5

Medical product sterilization

1

1

1

10

Medical product sterilization

1

1

1

5

Medical product sterilization

2

2

2

10

Medical product sterilization

6

3

6

5

Gemstones

2

2

1

22

Medical Therapy

6

0

6

6

Semiconductor irradiation

1

1

1

10

Contract irradiation

2

1

1

10

Contract irradiation

2

1

1

10

Medical product research

1

1

1

5

Contract irradiation

2

2

2

10

Food treatment (pathogen control and shelf life extension)

1

1

1

10

Gemstones, isotopes, semiconductors

2

2

1

20

Medical sterilization

1

0

1

10

Medical isotope production

3

3

1

35

Medical sterilization

2

2

1

10

Medical sterilization

1

1

1

10

 

40

28

32

 

Slide5

Gradients – To repair or to replace a section…. That is the question

Conditioning effort is proportional to gradient (to the n

th

power).

Conditioning effort is also related to required “missing pulse tolerance”.

“High gradient” S-Band: Pulse duration 2-4 usec.30MV/m takes 5 day bakeout at 400C and 2-5 days on RF test stand.Cannot be re-gunned/repaired in the field“Low gradient” S-Band:Pulse duration 8 – 16 usec.15MV/m takes no bakeout and 24 hours RF conditioningPlanned maintenance activities mean approximately 24 hours down.Catastrophic failures can be repaired but may take up to 2 weeks and may require a bakeout at 180C.

Slide6

Post-conditioning performanceMedical guides can be quickly (and fairly easily) replaced.

Medical guides typically require low breakdown/pulse/m (less than 10

-12

)

Conditioning to these gradients and breakdown rates is “easily” achievable.

This BDR requirement applies to certain “real-time” security applications.Industrial guides and their scanning systems are typically “fixtures”.Changing them is a big dealIndustrial guides can typically tolerate higher breakdown/pulse/mBreakdown rates may be in the range of 10-5 BD/pulse/m immediately following a pump down. Conditioning happens “on-the-fly” while the machine is making money.BDR drops during operation for approximately 7-10 days following pump-down.Conditioning to these gradients and breakdown rates is “easily” achievable.

Slide7

Industrialization….

Low production rate

Easy customization by application

Must be easy to understand and repair.

Industrial safety equipment.

Industrial PLC and HMIDistributed I/OModular-ized softwareConnector-izedRevision control

Slide8

Our next frontier – High energy, power, and reliability

Gemstones

Semiconductors

Medical isotope production

Moly-99 / Tc99m

I-123Cu-67Etc….Driving sub-critical assembliesPhoto-fissionHeatElectricity

Isotopes

Nuclear waste

This is long for us: (3 x 1.2m) 10,000 times shorter than CLIC

Slide9

Isotope production: A work in progress

The availability of high flux reactors for the production of medical isotopes caused panic several years ago.

Several Canadian groups received funding to do pilot-scale testing of alternatives.

Cyclotrons were built to directly produce Tc-99m from enriched Mo-100.

A

linac facility was funded to produce Mo-99 from natural Moly and enriched Mo-100.NRC did early calculations, target configurations, testing, and separation experiments.The Canadian Light Source coordinated the funding proposal and implementationThe pilot-scale linac was produced by Mevex and installed at the Canadian Light Source.35MeV1.2mA average current (average beam power 40kW)3 standing wave sections, 1.2m each3 klystronsS-Band – 2998MHz

Slide10

Isotope production: Production machine requirements

Parameters/overview:

35-50 MeV

3 – 5 mA average current (100 – 200kW average beam power)

3 - 5 standing wave sections, 1.4m each

3 -5 klystronsS-Band – 2998MHz“low gradient” 15MV/m averageHigh reliabilityPerforming service/maintenance activities in areas that have been activatedShut-downs are expensive ($1000’s per hour)Down-time causes scheduling/logistics problems… long time to recover.

Slide11

(CNS Workshop Dec-09)

11

Tc-99m:

140

keV

-ray, 6 hr half life Used for 90 % of nuclear medicine imaging

Canada – about 5500 procedures per day

Ottawa Hospital – about 15 cameras

Slide12

(CNS Workshop Dec-09)

12

Mo-99 via U-235 fission:

Mo-99 at peak of fission mass distribution

~ 6 % of fissions yield Mo-99

Half life of 66 hrs

Slide13

(CNS Workshop Dec-09)

13

An alternative route:

Photonuclear reaction on Mo-100

Natural Mo about 10 % Mo-100

Available at enrichments of > 95 %

Known for more than 40 years

Slide14

(CNS Workshop Dec-09)

14

Work at Idaho National Laboratory:

Late 1990’s

Worked through technical, economic details

Suggested single 15 kW accelerator for Florida

Each target about 15 g (1 cm by 2 cm)

Mo-100 consumption measured in µg

“Goats” are “milked” for their Tc-99m

Slide15

(CNS Workshop Dec-09)

15

Key enabling technologies:

High-power electron

accelerators

Separator for low specific

activity

Mo-100 enrichment > 95 %

Slide16

(CNS Workshop Dec-09)

16

One estimate:

Canadian

requirements (33M people):

430 six-day Ci of Mo-99 per weekAssume reactor model: need 2500 Ci of Mo-99 per week at end-of-beam

Need to produce 360

Ci

of Mo-99 per day

From INL study, 14 kW beam yields 25

Ci

after 24 hrs

Single 100 kW machine capable of producing about 180

Ci

in 24 hours

(From US NRC study – world production)

Slide17

(CNS Workshop Dec-09)

17

For Canada, 5,500 Tc-99m procedures per day

Each procedure requires 10 - 30

mCi

; thus

110

Ci

/day of Tc-99m

Every 24 hrs, can elute ~100 % of remaining Mo-99 activity

So need to replace 22

Ci

/ day of Mo-99

From

EoB

to delivery can be less than 1 t

1/2

(~ 3 days)

Conclude 44

Ci

/day, EoB, should be adequate

Another estimate:

These estimates differ by a factor of 8

Largely because of “six-day curie”

Slide18

(CNS Workshop Dec-09)

18

Mo-100 estimates:

Enriched to > 99 %: $2,000 per gram (~$600/g for large quantities)

Material will be recycled

Each day, irradiate two 15 g targets to yield 180

Ci

each

Recycle time set by decay: 10

mCi

can be handled with modest shielding: need 40 days

Need (2 x 15) [g/day] x 40 [days] = 1200 g of Mo target material: 2.4 M$

Nine cycles per year: losses per cycle expected to be small: suppose 4 %

Then need 430 g per year to replace Mo-100 losses

Slide19

(CNS Workshop Dec-09)

19

Capital

Cost (k$)

Two

35

MeV

, 100 kW accelerators, each 7 M$

14000

Building

, infrastructure, 3500

ft

2

, $

1000/ft

2

3500

Hot

cells

3000

Mo-100

2400

Laboratory equipment

200

Total capital

23100

Facility costs – two 100 kW machines in a single location:

Assumptions:

Both machines run 24 hours/day, 5 days a week

Targets will be processed on site, yielding

molybdate

ready for the

separator

Using “six-day curie” concept, but from

EoB

to shipping should be less than two days

Slide20

(CNS Workshop Dec-09)

20

Variable

Cost (k$)

Cost

of capital, 20 %

4620

Operator salaries (8 operators, 80 k$ each)

640

Supervisory, scientific salaries (head, two engineers, physicist, 120 k$ each)

480

Utilities, 2 MW, at 13 cents/kW-hr

1600

Target processing (two technicians, 80 k$ each)

160

Replacement Mo-100 (9 cycles/year, 4 % loss per cycle)

800

Accelerator maintenance and repairs (10 % of capital)

1400

Shipping (50 units per day, 260 days per year, $50 per unit)

650

Total variable

10350

Yearly output of Mo-99, 360

Ci

/day,

EoB

, 260 days per year

94000

Ci

Yearly output of six-day curies of Mo-99

21 000

Ci

Yearly output of Tc-99m, for five

milkings

62000

Ci

Separator costs from 1.5 to 5.0 ¢/

mCi

Unit cost of Tc-99m

~25 ¢/

mCi

Present customer cost about 100 ¢/

mCi

Slide21

(CNS Workshop Dec-09)

21

I-123:

159

keV

-ray, 13 hr half life

Several charged particle reactions can be used

Xe-124 (p,

pn

) Xe-123 gives best purity

Need 15 to 30

MeV

protons; enriched Xe-124

Typical dose costs $700, versus $20 for Tc-99m

Can also use Xe-124 (

, n) Xe-123

Slide22

(CNS Workshop Dec-09)

22

Oganesyan

et al

, Dubna, USSR, 199025 MeV, 0.3 kW

Measured 20

mCi

per hour for 10 g target

Scaling:

10 hr irradiation, x 10

100 kW beam, x 330

In 10 g, expect 66

Ci

Pluses:

Separation very easy

Gas is easily recycled

Minuses:

Half life of 13 hrs

Gas easily lost

Slide23

(CNS Workshop Dec-09)

23

35MeV

, 100kW

Linac

facility requirements (Single Unit)

Description

Value

Total power consumption (peak/maximum)

800kW

Total power consumption (typical operation)

650kW

Facility chilled water temperature

8C to 15C

Facility chilled water flow rate

360 liters/min

Facility chiller, heat removal capacity (recommended)

800kW

Ozone extraction fan - VFD control

3kW

Electrical conversion efficiency (AC to beam power)

Approximately 15-20%

Slide24

(CNS Workshop Dec-09)

24

A

ccelerator

cluster – 4

Linacs, 35MeV, 100kW each

Slide25

Thanks and acknowledgements:

Mark de

J

ong, The Canadian Light Source

Carl Ross, National Research Council, Canada

Walter Davies, National Research Council, CanadaJim Harvey, Northstar Medical Radioisotopes LLCChris Saunders, Prairie Isotope Production EnterprisePeter Brown, Mevex Corporation