Predicting Residual Radiation using Beam Loss Data – Early Results

Bruce C. Brown10 March 2010

Main Injector/Recycler Group Meeting

Bruce C. Brown -- MI/RR Group Meeting 2

Personal Note

Early operation of the Main Injector was at modest intensities and very high transmission (Think Tevatron, not the Booster or Main Ring) but in 2004, we were preparing for NuMI operation at about x10 intensity. I have been seeking since then to understand losses: mechanisms, residual radiation, monitoring.

Can we relate: protons lost to Residual Radiation, to BLM readings, to radiation safety limitations?

Can we control the loss locations?3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 3

Today’s Outline

• Primary Goal – Predict Residual Radiation• How to relate Residual Radiation to BLM’s• Some Results• Preliminary Conclusions• Issues with BLM data in Datalogger• Proposed additional efforts – Data cleanup for BLM – faulty readings– How much loss is important?– Lost Protons vs. BLM readings– Exploring BLM pedestal with I129 or I130

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 4

Predicting Residual Radiation

We need to predict residual radiation so that:

We can plan tunnel access work Will we need time for cool down? Can we predict cool down for shutdown? We know long term trends Will the current loss rate make work difficult?

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 5

Relating Residual Radiation to BLM Data Logs of LI(T) for losses - since 11Oct06 Residual Radiation at Bar-Coded location since 10Oct05 We access the tunnel after about 1 hour or more. We find

that the lifetimes of interest are about 2.5 hours. We will store sums of LI(T) for 10 minute intervals

Create half life weighted integrals (I is isotope, TM is measurement time).

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 6

Relating Residual Radiation to BLM Assume geometry of loss is constant:

Constant pattern for creating ionization in BLM Constant pattern for creating residual radiation at bar-coded locations

Measured Residual Radiation is linear in weighted BLM Sums

Note that we are dividing by the half life so the coefficient relates the residual radiation to the weighted loss rate . This is a linear fit so we can perform it with existing data. However, if long half life isotopes were produced well before the BLM data record, we will need to add a term to take care of it. We have a plan.

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 7

Console Application I130

Guan Wu is developing a program to implement a fitting procedure:

It reads data from the data logger tables, applies some cuts (more to be developed) and sums them to a convenient set of ‘Quanta’ and stores the results in Sybase.

It plots data logger or Quanta results for examination It provides a fit as shown above.

He is doing the work so I will do the talking !! 3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 8

Fit Results at V105 Corrector

3/10/2010

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 12100000000

100

200

300

400

500

600

700

800

V105 Residual Radiation vs. Integral of LI105

MeasFit

Linux Time (sec)

Resi

dual

Rad

iatio

n at

V10

5 (m

R/hr

)

Bruce C. Brown -- MI/RR Group Meeting 9

Fit Results at H104,Q104D

3/10/2010

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 12100000000

20

40

60

80

100

120

Q104D Residual Radiation vs. Integral of LI105

MeasFit

Resi

dual

Rad

iatio

n at

Q

104D

(mR/

hr) LIfetime Coeffi-

cient312 days 0.10395.591 days 0.08612.5789 hrs 0.8271

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 12100000000

5

10

15

20

25

30

35

40

H104 Residual Radiation vs. Integral of LI104

MeasFit

Resi

dual

Rad

iatio

n at

H10

4 (m

R/hr

)

LIfetime Coefficient312 days 0.03125.591 days 0.04372.5789 hrs 0.2565

Bruce C. Brown -- MI/RR Group Meeting 10

Fit Results at V621 Corrector

3/10/2010

10400000001060000000108000000011000000001120000000114000000011600000001180000000120000000012200000000

20

40

60

80

100

120

140

160

180

V621 Residual Radiation vs. Integrals of LI621

MeasFit

Resi

dual

Rad

iatio

n at

V62

1(m

R/hr

)

Bruce C. Brown -- MI/RR Group Meeting 11

Results at 610We get useful results even at low loss

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 12100000000

5

10

15

20

25

H610 Residual Radiation vs. Integral of LI610

MeasFit

Resi

dual

Rad

iatio

n at

H61

0 (m

R/hr

)

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 1210000000012345678

Q610D Residual Radiation vs. Integral of LI610

Series1Series3

Resi

dual

Rad

iatio

n at

Q

610D

(mR/

hr)

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 12

Residual Radiation Fit at LAM52A1Previously at 3000 mR/hr now 300 mR/hr

3/10/2010

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 12100000000

500

1000

1500

2000

2500

3000

3500

LAM52A1 Residual Radiation fit to LI522A Integrals

MeasFit

Unix Time (seconds)

Resi

dual

Rad

iatio

n at

LA

M52

A1

(mR/

hr)

Bruce C. Brown -- MI/RR Group Meeting 13

Residual Radiation at Primary Collimator – Q230D

3/10/2010

11200000001130000000114000000011500000001160000000117000000011800000001190000000120000000012100000000

100

200

300

400

500

600

700

Q230D Residual Radiation fit to LI230 Integral Loss

MeasFit

Unix Time (seconds)

Resi

dual

Rad

iatio

n at

Q23

0D (m

R/hr

)

LIfetime Coefficient312 days 0.2145.591 days 0.20412.5789 hrs 0.4257

Bruce C. Brown -- MI/RR Group Meeting 14

Residual Radiation at V309Last corrector in Collimation Region

3/10/2010

11200000001130000000114000000011500000001160000000117000000011800000001190000000120000000012100000000

100

200

300

400

500

600

700

800

900

V309 Residual Radiation fit to LI309 Integral Loss

MeasFit

Unix Time (seconds)

Resi

dual

Rad

iatio

n at

V30

9 (m

R/hr

)

LIfetime Coefficient312 days 0.37525.591 days 0.53852.5789 hrs 0.5227

Bruce C. Brown -- MI/RR Group Meeting 15

Residual Radiation at V109Why the poor fit to recent data?

3/10/2010

1070000000 1090000000 1110000000 1130000000 1150000000 1170000000 1190000000 1210000000

-10

0

10

20

30

40

50

60

V109 Residual Radiation vs. Integrals of LI109

MeasFit

Resi

dual

Rad

iatio

n at

V10

9 (m

R/hr

)

tau E

312.3 day 0.1241

5.590 dat 1.7324

2.579 hr -2.626

Chisqr 2.183

Bruce C. Brown -- MI/RR Group Meeting 16

Various Fit Results using 3 Isotopes

3/10/2010

Location BLM 312.3 day 5.591 day 2.5789 hr chisqr312.3 5.591 2.5789

V105 LI105 1.0843 1.3421 0.9249 14.653V621 LI621 0.373 0.3916 0.2069 3.576V109 LI109 0.1241 1.7324 -2.626 2.183 Q230D LI230 0.214 0.2041 0.4257 6.668 V309 LI309 0.3752 0.5385 0.5227 15.132 LAM52A1 LI522A 0.236 0.2738 -0.0172 153.681 H104 LI104 0.0312 0.0437 0.2565 0.518 Q104DS LI104 0.1039 0.0861 0.8271 1.799 V609 LI609 3.2885 1.328 2.5358 4.733 S408 LI408 1.8541 1.1733 1.5016 3.269

Bruce C. Brown -- MI/RR Group Meeting 17

Preliminary Results

• The Proposed Fit seems to provide a useful description of the residual radiation based on the BLM integrals.

• Coefficient of 1E+4 implies 1 mRad/sec of BLM loss yields 10 mR/hr on typical object in the tunnel

• Using three isotopes provides a useful description for present purposes. Will explore alternative options later. Software handles 5 now.

• Better review of the data logger information may produce better fits.

• Should be able to provide cool down curve prediction for shutdown planning.

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 18

Issues with BLM results in Datalogger

• We know that Main Injector Cycles without a prepare for beam will result in storing the same result again and again. If this was due to a high loss abort, the data may show high loss (fix code ready to tweak).

• Present 600 second sums have a few unphysical points. Will have to explore how to remove them. At present simply cut after reviewing data.

• The pedestal is not a zero so the integrals may have added or missing Rads. Pedestal of 2 Rad may create 10 mR/hr residual radiation.

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 19

Lost Proton Measurement

The Beam Loss Monitors have a measured calibration in Rads. This relates the ionization detected to the radiation field which produced that ionization

The loss geometry varies around the ring but fortunately the regular placement of BLM detectors (above beam height on the wall at the downstream end of each quadrupole) makes the characteristic response to lost protons somewhat regular

Using the integral around the ring, a nominal response is 0.66 Rads/1E12 Protons lost at 8 GeV. Extinction scans produce a number about half of this.

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 20

Today’s Outline

• Primary Goal – Predict Residual Radiation• How to relate Residual Radiation to BLM’s• Some Results• Preliminary Conclusions• Issues with BLM data in Datalogger• Proposed additional efforts – Data cleanup for BLM – faulty readings– How much loss is important?– Lost Protons vs. BLM readings– Exploring BLM pedestal with I129 or I130

3/10/2010

Bruce C. Brown -- MI/RR Group Meeting 21

S408 Residual Radiation

3/10/2010

1060000000 1080000000 1100000000 1120000000 1140000000 1160000000 1180000000 1200000000 12200000000

20

40

60

80

100

120

S408 Residual Radiation fit to Integral of LI408

MeasFit

Unix Time (seconds)

Resi

dual

Rad

iatio

n at

S40

8 (m

R/hr

)

Bruce C. Brown -- MI/RR Group Meeting 22

Residual Radiation at V609Note old radiation levels here

3/10/2010

1060000000 1080000000 1100000000 1120000000 1140000000 1160000000 1180000000 12000000000

20

40

60

80

100

120

V609 Residual Radiation fit to LI609 Loss Integral

MeasFit

Linux Time (seconds)

Resi

dual

Rad

iatio

n at

V60

9 (m

R/hr

)