Roberto Saban & Mirko Pojer

Training the dipoles A.Werveij

What did we learn without beam in 2008 ?

Superconducting electrical circuits K-H.Meß

The Sector 34 Incident Ph.Lebrun

Calorimetric and electrical measurements and related software

N.Catalan-Lasheras

LHC Cryogenics: What did we learn from cool-down to first beams

S.Claudet

What else did we learn? M.Pojer

Roberto Saban & Mirko Pojer

2LHC Performance Workshop Chamonix - What did we learn without beam in 2008?

Training the dipoles

R.Saban Feb 24th , 2009

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.V

erw

eij

in sector 56




A.V

erw

eij

7000

8000

9000

10000

11000

12000

13000

0 20 40 60 80 100 120 140

Magnet number (ordered by 1st quench in SM18)

Cu

rren

t [A

]

ALS: 1st training quench SM18

ANS: 1st training quench SM18

NOE: 1st training quench SM18

ANS: Quenches sector 56

NOE: Quenches sector 56

7 TeV

6 TeV

in sector 56




A.V

erw

eij

in sector 56

Sector

Number of magnets Number of quenches to 7 TeV

ALS ANS NOE@ 6 TeV(±2)

@ 6.5 TeV

(±30%)Est. 1 Est. 2 Est. 3 Est. 4

1-2 49 96 9 0 4 22 41 40 49

2-3 56 60 38 1 8 23 97 92 130

3-4 56 65 33 1 8 21 87 83 116

4-5 46 46 62 2 12 22 145 136 198

5-6 28 42 84 1 15 21 190 178 262

6-7 57 36 61 2 12 20 142 133 194

7-8 54 40 60 2 12 14 140 132 192

8-1 64 24 66 2 13 19 151 142 208

Total 11 84 162 993 936 1349

Est. 1: Based on 115 MB’s that have been submitted to a thermal cycle in SM-18 (2008 before HWC, P. Xydi and A. Siemko)

Est. 2: Extrapolation from sector 5-6 data + estimate 1 for ALS & ANS

Est. 3: 2 quenches per NOE magnet + estimate 1 for ALS & ANS

Est. 4: 3 quenches per NOE magnet + estimate 1 for ALS & ANS

993 / 8 = 124124 / 3 = 41

5LHC Performance Workshop Chamonix - What did we learn without beam in 2008?R.Saban Feb 24th , 2009

K-H

.Meß

Superconducting electrical circuits

Harmless The third current lead shared by two circuits (Kirchoff’s Law) Short inside the dump resistor on the QF of Sector 45 Short to ground on the dump resistor on the QD of Sector 56 Missing resistor on one of the poles of an undulator which gave a

very exotic transfer function Leveling of the DFB to properly wet the superconducting cable

Potential inconvenience to operation Distribution of the conductors on cables sharing the same DFB but

on different circuits – symptom: apparent detraining The reference magnet puzzle – magnetization cycle

Potentially dangerous Symmetric quenches Transient spike when the dump switch opens due to the difference in

Eddy currents in the two apertures Quench back in corrector circuits inducing coupling on other circuits

… up to quenching the main magnets Pending: splice and voltage tap non-conformities

Unresolved MCBX and MCBY mystery: unexplainable transfer functions The hunch on the MCBYs Fast quench propagation observed on only the dipole circuits

Surprises


Mandate of the Task Force Establish the sequence of facts, based on

experimental measurements before incident, observations after incident and timing

Analyse and explain the development of events, in relation with design assumptions, manufacturing & test data and risk analyses performed

Recommend preventive and corrective actions for Sector 3-4 and others


Ph.L

ebru

nThe Sector 34 Incident

0

10

20

30

40

50

60

70

80

1.85

1.87

1.89

1.91

1.93

1.95

1.97

1.99

18:00 19:00 20:00 21:00 22:00

Valv

e op

enin

g [%

], cu

rren

t [kA

], CC

flow

[g/s

]

Cold

-mas

s tem

pera

tue

[K]

LBALA_24R3_TT821.POSST




LBALB_24R3_TT821.POSST

LBALB_26R3_TT821.POSST

LBBLA_24R3_TT821.POSST




LBBLD_25R3_TT821.POSST

LBBLD_27R3_TT821.POSST

LQASB_23R3_TT821.POSST

LQOAA_25R3_TT821.POSST

LQOBA_24R3_TT821.POSST

LQOBA_26R3_TT821.POSST

QRLAA_25R3_CV910.POSST

QRLAB_23R3_CV910.POSST

QURCA_4_FT201.POSST

RPTE.UA43.RB.A34:I_MEAS

-400

-200

0

200

400

600

800

1000

8400 8450 8500 8550 8600 8650 8700 8750

Current [A]

Vo

lta

ge

[mV

]

0

100

200

300

400

500

600

700

Te

mp

[K

]

V_meas (Timber)

V_meas (QPS data)

Voltage simulated [mV]

T_max simulated [K]

Arjan Verweij, TE-MPE

Temperature drift during the 7 kA current flat top (15 Sep 2008)

Measured versus simulated incident with 220 n joint and bad contact with U-profile and wedge

No electrical contact between wedge and U-profile with the bus on at least one side of the joint

No bonding at joint with the U-profile and the wedge


Ph.L

ebru


The current decay from 8700 A

Energy MJ %

Stored in the magnets 595.0 100

Dissipated in UJ33 71.0 12

Dissipated in UA43 104.8 18

Dissipated in cold mass 144.4 24

Dissipated in electrical arcs 274.8 46


Ph.L

ebru


PTQVQV QV QVQV SVSV

Cold-massVacuum vesselLine ECold support postWarm JackCompensator/BellowsVacuum barrier

Q D D QD D D QD D D QD D D QD

0

2000

4000

6000

8000

10000

12000

Time

Curr

ent (

A)

0

5

10

15

20

25

P H

e ve

ssel

(bar

), P

beam

(mba

r), P

insu

latio

n (b

ar)

Ins Vac A19R3

Ins Vac A23R3

Ins Vac A27R3

I DCCT

I dump 3

I dump 4

P CM Q19-21.R3

P CM Q23-25.R3

P CM Q27-29.R3

Vac beam 23R3.R

Vac beam 23R3.B

Vac beam 31R3.R

Vac beam 31R3.B


Ph.L

ebru


• Calorimetric measurements• Electrical measurements on

«suspect» cells/subsectors powered at limited current

• During further power tests, track temperature evolution in normalized conditions

• Modify quench detection system to include interconnects and bus bar splices

• Consider option to measure currents in 13 kA circuits at both ends of sector and detect differentials

• Review possible improvement of mechanical clamping of interconnects and gradually implement whenever possible

Prevention of initial faultRecommendations

Mitigation of consequences• Increase number/size of relief

devices on cryostat vacuum vessels

• Review number, size & position of pressure relief devices on beam vacuum system

• Review closure logic of beam vacuum sector valves

• Consider possibility of triggered opening of quench relief valves below set pressure

• Consider general firing of quench heaters

• Reinforce external anchoring at locations of vacuum barriers

• Reexamine personnel underground access rules

• Review location of AUG in tunnel and protection from blast

• Review recorded signals, recording frequency and time stamping coherence among different systems


N.C

ata

lan-L

ash

era

sCalorimetric and electrical measurements

Sector 12

7 kA 7 kA 7 kA 9.3 kA+40 mK

-10 mK

Rela

tive t

em

p

1 or 2 hour flat tops

7 kA 7 kA 8.5 kA 7 kA

Rela

tive t

em

p

Sector 23 Sector 34 Sector 45

Sector 56 Sector 67 Sector 78 Sector 81

+40 mK

-10 mK

Rela

tive t

em

p

• All the current plateaux were scrutinized for suspect temperature increase

• Unstable conditions and dynamic temperature control prevent accurate calculations.


N.C

ata

lan-L

ash

era


0

0.2

0.4

0.6

0.8

1

1.2

1.83

1.84

1.85

1.86

1.87

1.88

1.89

0 1 2 3 4

I / I p

late

au [-

]

Tem

pera

ture

[K]

Time [hour]

TemperatureCurrent

DU

(DT)

Assessment of the baseline slope (valve opening

mismatch)the temperature increase during

powering plateauthe internal energy variation (J/kg)the deposited energy assuming a

mass of 26 l/m of He

Before heating With heating

ΔU [J/kg] -1.1 78

M [kg] 823

ΔU [kJ] -0.92 64.2

t [s] 2880 6600

W [W] -0.3 9.7

ΔW [W] 10

31R1 31R6

15R1 23R331R7

0

5

10

15

20

25

-50 -40 -30 -20 -10 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200

# o

f su

b-se

cto

rs [-]

Sub-sector resistance variation w/r to baseline [nW]

Dipole (5 sectors) Quadrupole (4 sectors)

NNXN: Confirmed by electrical measurements

~ 2 W @ 7 kA

The new powering procedures will demand mandatory calorimetric and electrical tests in ALL sectors at the beginning of the next LHC powering campaign


N.C

ata

lan-L

ash

era


0.7mV/7kA=100nOhm0.7mV*7kA=4.9W

Snapshot on 03.09.08 : 0.85mV*8.4kA=7.1W

Calorimetric measurements spotted a suspect region in Sector 12

0 2000 4000 6000 8000 10000 12000 14000-0.20

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40D1_L-Ua

Current (A)

Vo

lta

ge

(m

V)

Error Bars:±2σ/√n

1.6 nOhms

105 nOhms

an inter-pole splice resistance of 105 nOhm in magnet 2334 (B16R1)


S.C

laudetCryogenics: from cool-down to first beams

≈15K/d x 13 d

≈8 K/d x 8 d

≈12 K/d x 4 d

“25 days for 20K”

+ 3 days for Filling

+ 2 days for 1.9K

30 days

(4 to 5 wks)


S.C


Distribution Line Distribution Line

Magnet Cryostats Magnet Cryostats

Cold Compressorbox

Warm CompressorStation

Shaft

Su

rface

Cavern

Tun

nel

LHC Sector (3.3 km) LHC Sector (3.3 km)

New 4.5 K refrigerator 1.8 K Refrigeration Unit


Cold Box

Interconnection BoxCold Compressor

box


1.8 K Refrigeration Unit Ex-LEP 4.5 K Refrigerator

Upper Cold Box


Lower Cold Box

• Running two sectors with one cryoplant was tested during powering

• Not valid for large transients, but an interesting alternative for low beam loads. It is a validated fall-back scenario if serious problems with a refrigerator


S.C


Non conformities were detected, workarounds found during the run and consolidations actions ongoing level gauges on stand alone magnets, DFBLC heat load, valves on current leads, heat loads on the triplet, etc.

Stability was achieved stable services, global refrigeration mastered, 15mbar established, DFB, current lead and beam screen cooling loops stabilised

Recovery from quenches or failures a lot of experience gained and results obtained

Towards more stable services electricity, cooling water controls, insulation vacuum

Getting ready for round the clock operation


M.P

oje

rWhat else did we learn?

The ProceduresThe Tools which implemented the procedures and assisted

the operators during the execution and the analysis

SequencerPost-Mortem browserPIC supervisionQPS supervisionPowering to Nominal (P2N)Databases… others were developed during the hc

efficiency, automation and no compromise

Some teething problems were identified and were corrected or are being corrected remote resets, communications problems, timing, coherence between time stamps, etc.


M.P

oje

rWhat else did we learn?

Hardware problems

600A-10V LHC type power converters QPS requires a smooth change in current, otherwise it trips

LHC 600A-10V: 0V-crossing distortionThe power converters generate some distortion when crossing through zero voltage with current in the load QPS trips

600 A-10 V Crowbar Issue Some PCs don’t have the crowbar and are not safe under certain conditions

Reduced dI/dt and d2I/dt2 which could limit operation and physics

Frequent clogging of water filters installed on cables and converters lines

flow reduction and stop of the converter

ECR to add crowbar on those circuits

A decision was taken to change all filters around the machine from the present 50 μ to 100 μ mesh


Conclusions

Many surprises: training & sector 34

During the commissioning the equipment owners and the operation crews gathered the experience which they expected to re-commission, run and debug the equipment

The incident in sector 34 requires modifications of the hardware, upgrade of protection systems and the development of additional test procedures

While during most of the commissioning campaign the observations matched what was expected; in a few cases however, they revealed non-conformities some of which remain to be understood, followed and corrected or coped with

Roberto Saban & Mirko Pojer

Documents

Transcript of Roberto Saban & Mirko Pojer