ENERGY OF THE LHC AFTER LONG SHUTDOWN 1 (2013-2014)

21
E. Todesco ENERGY OF THE LHC AFTER LONG SHUTDOWN 1 (2013-2014) C. Lorin, E. Todesco and M. Bajko CERN, Geneva Switzerland With relevant inputs from colleagues L. Bottura, F. Bordry, G. d’Angelo, G. De Rijk, P. Fessia, K. Foraz, E. Nowak, A. Siemko, J. P. Tock, D. Tommasini, A. Verweij Measurements at SM18 of 3-4 magnets G. Dib, G. Deferne Production data of training at SM18 V. Chohan, the SM18 team and the Indian collaboration Chamonix, 8 th February 2011

description

Chamonix, 8 th February 2011. ENERGY OF THE LHC AFTER LONG SHUTDOWN 1 (2013-2014). C. Lorin , E. Todesco and M. Bajko CERN, Geneva Switzerland With relevant inputs from colleagues L. Bottura , F. Bordry , G. d’Angelo , G. De Rijk , P. Fessia , K. Foraz , - PowerPoint PPT Presentation

Transcript of ENERGY OF THE LHC AFTER LONG SHUTDOWN 1 (2013-2014)

Page 1: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco

ENERGY OF THE LHC AFTER LONG SHUTDOWN 1 (2013-2014)

C. Lorin, E. Todesco and M. BajkoCERN, Geneva Switzerland

With relevant inputs from colleagues

L. Bottura, F. Bordry, G. d’Angelo, G. De Rijk, P. Fessia, K. Foraz,

E. Nowak, A. Siemko, J. P. Tock, D. Tommasini, A. VerweijMeasurements at SM18 of 3-4 magnets

G. Dib, G. DeferneProduction data of training at SM18

V. Chohan, the SM18 team and the Indian collaboration

Chamonix, 8th February 2011

Page 2: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 2

CONTENTS

Estimates to reach 6-6.5 TeVMagnets measured during productionMagnets (mainly of 5-6) training during hardware commissioningMagnets tested in 2009-2011 (from 3-4 incident)

Models for 7 TeV

Risks related to quench heater weakness

Strategy

Page 3: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 3

ESTIMATES TO REACH 6-6.5 TeV:1 - SCALING OF PRODUCTION DATA

First approach: take the measurements after thermal cycle, compute the quench probability and apply to the LHC

[P. Pugnat, A. Siemko, IEEE Trans. Appl. Supercond. 17 (2007) 1091.]

About 10% sampling (119 magnets tested out of 1232)

«After a TC performed on ∼11.5 % of MB […] ∼25 % of MB required at least one training quench to reach the nominal field equal to 8.33 T […] the number of

quenches that may occur during the first powering cycles up to nominal field is around 330»

Probably biased?

According to the specification, magnets which had «bad performance» in virgin state went under a test after thermal cycleEstimate lowered to 200 total quenches to account for this effect, thus giving 25 quenches/sector

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 330

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 200-330

Page 4: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 4

ESTIMATES TO REACH 6-6.5 TeV:1 - SCALING OF PRODUCTION DATA

Simplest approach: take the measurements after thermal cycle, compute the quench probability and apply to the LHC [P. Pugnat, A. Siemko, IEEE Trans. Appl. Supercond. 17 (2007) 1091.]

Second analysis performed at the end of the production

[E. Todesco et al., Chamonix 2009, C. Lorin, E. Todesco, CERN ATS Report 2010-164]

Large spread induced by the inclusion/exclusion of bad casesNeed of training to get to 6.5 TeV

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 200-330

Scaling of production data after thermal cycle 2008 0 0 50-80 230-550

Page 5: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 5

ESTIMATES TO REACH 6-6.5 TeV:2 - MONTECARLO OF PRODUCTION DATA

Take the data before thermal cycle, use the correlations in performance before and after thermal cycle with a MonteCarlo [B. Bellesia, N. Catalan Lasheras, E. Todesco, Chamonix 2009]

Contrary to previous scaling, we also use data in virgin conditionsSimilar result, but first evidence of a weakness of Firm3 present in the production data

8

9

10

11

12

0 20 40 60 80

Cur

rent

(kA

)

First quench number

Firm1

Firm2

Firm3

Firm2 HC

Firm3 HC

Sector 5-6: Montecarlo vs hardware commissioning

7 TeV

6.5 TeV

Page 6: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 6

ESTIMATES TO REACH 6-6.5 TeV:3 - TRAINING IN THE TUNNEL IN 2008

Hardware commissioning data in 2008 (training in the LHC)

All sectors reached 5 TeV, seven reached 5.5 TeVTwo sectors trained up to 6 TeV with few quenchesSector 5-6 pushed up to 6.6 TeV with 30 quenchesSurprise: nearly all quenches from Firm3 magnets

Scaling of these data More pessimistic w.r.t production data – sign of degradation

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 200-330Scaling of production data after thermal cycle 2008 0 0 50-80 230-550

Scaling of hardware commissioning 2008 10 30 100 ?

Page 7: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 7

ESTIMATES TO REACH 6-6.5 TeV:4 - INCIDENT MAGNETS TESTED IN 2009-2011

32 magnets measured, coming from 3-4 incidentLow sampling (15 from Firm1 – 11 from Firm2 – 6 from Firm3)Could be biased if the magnets suffered from incident«For these magnets, coils have not been touched, at most the electrical connections were cured» [P. Fessia]

Unfortunately, only 6 from Firm3

Scaling of these dataConsistent with production data

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 200-330Scaling of production data after thermal cycle 2008 0 0 50-80 230-550

Scaling of hardware commissioning 2008 10 30 100 ?Test of 3-4 magnets recovered after incident 2011 0 0 0 330

Page 8: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 8

CONTENTS

Estimates to reach 6-6.5 TeVMagnets measured during productionMagnets (mainly of 5-6) training during hardware commissioningMagnets tested in 2009-2011 (spares and refugees from 3-4)

Models for 7 TeV

Risks related to quench heater weakness

Strategy

Page 9: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 9

MODELS FOR 7 TEV: QUENCH «CREEP»

A log fit was proposed for fitting training data in the tunnel

Is the same type of log t dependence is used for creep

(Also used for flux creep in Tevatron data of chromaticity decay)Not physical in the limit t

900 to 1300 quenches to reach 7 TeV

8500

9000

9500

10000

10500

11000

11500

12000

12500

13000

1 10 100 1000

Quench number

Qu

ench

cu

rren

t [A

]

5.5 TeV

7 TeV

6.5 TeV

6 TeV

S45

S56 in SM-18

S56

S78

190

A. Verweij, Chamonix 2009

CERN ATS 2009-001 (2009) 25

nBAnI log)(

Page 10: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 10

MODELS FOR 7 TEV: SLIP-STICK MODEL

Slip-stick model gives a functional equation for training

Equilibrium between superconductor margin and energy dissipated by friction [P. P. Granieri, C. Lorin, E. Todesco, IEEE Trans. Appl. Supercond. 21 (2011) 3555]

x

yez

O

Fr

Energy

CurrentIssI2I1 I3

E1E2

I4

E3 Ef = α(I4-I14)

Ef = α(I4-I24)

Em

Fr

x

Δx)( );()( IIIIfIE ssssmm

41

41 )( IIIE f

4IxFE rf

8

9

10

11

12

13

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Cur

rent

(kA

)

Number of quenches per magnet

Firm2 LHC data

Power 10 fit

Page 11: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 11

MODELS FOR 7 TEV: SLIP-STICK MODEL

The slip-stick model gives a functional equation for training

This model shown that asymptotic behaviour reached exponentially

One parameter is freed to have a better fit of initial part of virgin data

And extrapolation is taken on 5-6 data (2008 training in the tunnel)This gives an estimate of 110 quenches per octant for 7 TeV (as lower limit of from log fit)

)/exp(1)( qss nnInI

Method Year 6 TeV 6.25 TeV 6.5 TeV 7 TeVScaling of production data after thermal cycle 2006 200-330Scaling of production data after thermal cycle 2008 0 0 50-80 230-550

Scaling/extrapolation of hardware commissioning 2008 10 30 100 900-1300Test of 3-4 magnets recovered after incident 2011 0 0 0 330

Page 12: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 12

LOG VERSUS EXP

Both a log fit and an exp fit work … isn’t it strange ?It may happen over certain range

I show it for chroma decay which has similar exp fit

The double exp fit is used today in FiDeL for keeping chroma constant at LHC injection

0

5

10

15

20

0 10000 20000 30000

Chr

oma

deca

y

Time (s)

0

5

10

15

20

300 3000 30000

Chr

oma

deca

y

Time (s)

Double exp used to fit chroma decay at injection The same data in log scale !

Page 13: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 13

SUMMARY OF ESTIMATES

Final comparison of different estimates

0

200

400

600

800

1000

5.75 6 6.25 6.5 6.75 7

Num

ber

of q

uenc

hes

in th

e L

HC

Energy (TeV)

Scaling of production after thermal cycle

Scaling/extr. on hardware commissioning

Scaling on magnets from 3-4

Page 14: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 14

SUMMARY OF ESTIMATES

Large error associated to the estimatesFor scaling up to 6.5 TeV, low samplingFor 7 TeV the model and the extrapolation, adding to the sampling

We have to guess an average probability of quenching

It could look like a binomial

… but it is notI am still convinced the error is proportional to the 1/n

Discretization errors are already large

0 quenches could mean 0.5 quenches per magnet … this would give 50 quenches to get to 6.5 TeV in 3-4 magnets

kk ppk

npnkF )1(),;(

ppnkFn

),;(1

n

pppnkF

n

)1(),;(

1

Tested Quench Tested Quench Tested Quench Tested Quench

Firm1 15 0% 15 0% 15 0% 15 20%Firm2 11 0% 11 0% 11 0% 11 27%Firm3 6 0% 6 0% 6 0% 6 33%

Expected quenches in LHC

6 TeV 6.25 TeV 6.5 TeV 7 TeV

0 0 0 331

Page 15: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 15

SUMMARY OF ESTIMATES

Final comparison (zoom of previous data)Different models give an idea of the associated errors

0

20

40

60

80

100

120

140

5.75 6 6.25 6.5 6.75 7

Num

ber

of q

uenc

hes

in th

e L

HC

Energy (TeV)

Scaling of production after thermal cycle

Scaling/extr. on hardware commissioning

Scaling on magnets from 3-4

Page 16: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 16

CONTENTS

Estimates to reach 6-6.5 TeVMagnets measured during productionMagnets (mainly of 5-6) training during hardware commissioningMagnets tested in 2009-2011 (spares and refugees from 3-4)

Models for 7 TeV

Risks related to quench heater weakness

Strategy

Page 17: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 17

RISKS

7 dipoles from Firm2 have non conformity in the quench heaters [from J. P. Tock]

3 at the level of the IFS, no need of disassembly – just a warm up4 inside the magnet, they will be replaced in 2013 shutdown

In 2010, about 100 magnet quenches at current 2 kA and 6 kA [from A. Siemko, this Chamonix]

Plus, since 2008, quench heaters have been fired ~6000 times in the tunnel at 0 A [from E. Nowak]

Firing per magnet ranging from 3 to 17, average 5

Now voltage reduced from 900 V to 200 V during hardware commissioning

During a magnet training quench, on average 4 magnets quench [from A. Verwej]

So, if we make 100 training quenches we have 400 quench heater firing, still well in the shadow

Page 18: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 18

CONTENTS

Estimates to reach 6-6.5 TeVMagnets measured during productionMagnets (mainly of 5-6) training during hardware commissioningMagnets tested in 2009-2011 (spares and refugees from 3-4)

Models for 7 TeV

Risks related to quench heater weakness

Strategy

Page 19: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 19

STRATEGY

Push the four initial sectors to 6.5 TeV (Nov 2013- Mar 2014)

See if our estimates are valid: 50 quenches expected in four sectors

Fix 6.5 TeV as LHC energy and push the other four sectors there

If more quenches are needed, fix 6.25 TeV as LHC energy and push there the other four sectors (10 quenches expected in four sectors)

Best guess of 2013-2014 shutdown schedule [K. Foraz]

Page 20: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 20

CONCLUSIONS

Magnets coming from 3-4 do not show degradation of performanceOur best estimates to train the LHC (with large errors)

30 quenches to reach 6.25 TeV 100 quenches to reach 6.5 TeV

Two quenches/day 2 to 5 days of training per sectorWith 100 quenches one expects 400 quench heater firings

The planTry to reach 6.5 TeV in four sectors in March 2014Based on that experience, we decide if to go at 6.5 TeV or step back to 6.25 TeV in March 2014

This still leaves three months for MonteCarlo

Page 21: ENERGY OF THE LHC  AFTER LONG SHUTDOWN 1 (2013-2014)

E. Todesco Energy of the LHC after 2013-2014 long shut-down - 21

Back-up

Quadrupoles performance:SM18:

MQ as good as Firm1 MB

2008 hardware commissioning:« During HWC they show no need of retraining, reachingnominal with a few quenches »]

8

9

10

11

12

13

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

Cur

rent

(kA

)

Number of quenches per magnet

Firm1Firm2Firm3MQ

8

9

10

11

12

0.0 0.5 1.0 1.5 2.0

Cur

rent

(kA

)

Quench per magnet

MQ, virgin

MQ, hardware commissioning

7 TeV = 215 T/m

MQ