LHC ACCESS SYSTEM Global Tests of the LHC

LHC ACCESS SYSTEM Global Tests of the LHC Access system Outcome P. Ninin, T. Petterson, S Grau, T. Ladzinski 18 th January 2008 18 January 2008 EDMS: 888085 1

LHC ACCESS SYSTEM Global tests • Organization: - Organized by Silvia Grau and Pierre Ninin - Performed by a large team of TS/ASE - In excellent collaboration with AB/OP (to help searching LHC, giving access, driving EIS-beam equipment, etc…) • From Monday 7 th to Sunday 13 th January 18 January 2008 EDMS: 888085 2

LHC ACCESS SYSTEM Main Test Objectives • Test the Access System when doing a patrol of all the LHC points • In « Access Mode » , test the correct system behaviour when one or more EIS-beam becomes unsafe • In « Beam Mode » , test the triggering of all the « EISBEAM» and external interfaces (BI, LBDS) in case of intrusion, door opening inside the LHC or Emergency stop. • Test the access delegation operation with the Experiments • In collaboration with AB/OP, prepare the LHC Access operation and train the users 18 January 2008 EDMS: 888085 3

LHC ACCESS SYSTEM Test Strategy Step 1: Factory acceptance Step 2: Site acceptance LHC 1 Step 3: Global acceptance LHC 1. 8 LHC 2 LHC 5 LHC 6 LHC 3 LHC 4 LHC 5 LHC 6 LHC 7 LHC 3 LHC 8 LHC 2 LHC 1. 8 18 January 2008 EDMS: 888085 LHC 1 4

LHC ACCESS SYSTEM Monday 7 th – Sunday 13 th January • Monday: - Move the system from “installation” to “operation” • Tuesday to Thursday: - Patrol the LHC • Thursday evening - Closing the LHC and move to BEAM mode • Friday - System test in Access mode • Saturday: - System test in Beam mode • Sunday: - Move the system back to “installation” mode 18 January 2008 EDMS: 888085 5

LHC ACCESS SYSTEM Patrolling the LHC • Simplified quick patrol of the 160 access sectors : - All patrol boxes have been rearmed and sectors quickly searched. • Seal (“Plombage”): - Emergency exit handles on all doors and access points • Teach the people to open doors the normal way • Emergency opening with the “emergency handle” is of course still possible - After the tests, the seals have been left on most doors • Patrol procedure - AB patrol procedure not used with the exception of LHC 8 - ATLAS real patrol procedure done (without SSA) - CMS patrol procedure attempt 18 January 2008 EDMS: 888085 6

LHC ACCESS SYSTEM Patrolling the LHC • Some people had to work in the sectors being searched: - List of authorised workers was provided before the tests - Workers were ready with the “biometry” and the dosimeter, no particular problem reported to access LHC in general mode - Access given to workers in restricted mode in the patrolled area • Access to experiments allowed throughout the whole week: - Grids removed at PX 24 and ULX 15 access points to allow access to ALICE and ATLAS during the tests. - One guardian was placed next to each missing grid to control access - Only one door forced between the ATLAS cavern and the service area (which had to be re-patrolled) - Atlas suspended work Saturday, due to the almost permanent BIW signals resonating…. - Alice representative in CCC during the tests to inform the experiment in case of real “evacuation” 18 January 2008 EDMS: 888085 7

LHC ACCESS SYSTEM Tests in access mode • “One or more EIS-beam becomes unsafe” • Initial state: - LHC fully searched, - “Access mode” - All EIS-beam are safe Eis-beam: Power converters: LHC 2, 3, 7, 8 18 k. V & 400 V cells: LHC 2, 3, 7, 8 TEDs: TI 2 & TI 8 Access Safety Blocks: LHC 3 External interfaces Beam Imminent Warning LHC beam dump Beam interlock system SSA SPS access system 18 January 2008 EDMS: 888085 8

LHC ACCESS SYSTEM Tests in access mode • Test case A: One EIS-beam becomes unsafe: - Result: • the system goes to access-off • Test case B: Two EIS-beam become unsafe: - Result: • the system goes to access-off • and BIW (Beam Imminent Warning) sirens start • Systematic tests done, for each EIS-beam, one by one • Tests were successful. • Remarks: - First time LHC went to “access-off”, after BIW : • evacuation siren in point 6, fire alarm and ODH alarms in point 6, and 3 minutes later AUG in point 8. • All false alarms, but resulting in two fire-brigade interventions. 18 January 2008 EDMS: 888085 9

LHC ACCESS SYSTEM Tests in beam mode • Test case: Move into « safe position » all « EIS-Beam» and external systems (BI, LBDS) in case of: - Intrusion, Door opening inside the LHC Emergency stop. • Initial state: - LHC fully searched and closed (leaving some people in, on purpose, for the tests) “beam mode” EIS-beam unsafe (18 KV cells “ON”, power converters “ON”, TEDs “OUT”, etc…) • A sample of all types of doors tested all around the LHC: - inter-site doors, sector doors, end of zone doors, trapdoors, ventilation doors, access point (both PAD and MAD), etc… • One real AUG done in LHC 7: OK • Outcome: Tests successful - All intrusions detected - Interlocks triggered all the EIS-Beam to move into SAFE position 18 January 2008 EDMS: 888085 10

LHC ACCESS SYSTEM EIS-Beam and external interfaces • Beam Imminent Warning when two “EIS-Beam” go unsafe: should actually be “Evacuation siren” rather than BIW • A power converter and its associated power distribution cell are currently considered as two independent EISBeams: should be considered as a single one • On several occasions the TEDs could not be moved. Still unclear why. • On several occasions: LASS saw TEDTI 8 “OUT” (actually “not IN”) although it was “IN”. Position Sensor issue? • Electrical problems with two power converters (warm magnets in point 3 and 7) – under investigation • LBDS: triggered as foreseen, timing of signals to be investigated • BIS: final system not installed 18 January 2008 EDMS: 888085 11

LHC ACCESS SYSTEM Test Results • Access system installation was completed just before Christmas • Project team is satisfied of the test outcome - 6 days extensive use of LASS and LACS - All LASS and LACS functionalities tested - Many lessons learned • Post-mortem data analysis still ongoing - Not easy due to the multiple data sources • Some external systems not in final configuration - Test of the interfaces (LBDS, PC, BIC, …) but not of the complete chain 18 January 2008 EDMS: 888085 12

LHC ACCESS SYSTEM LHC ACCESS CONTROL SYSTEM - LACS • Improvements seen in software since the LHC 7 and LHC 8 (TI 8 test) summer tests: - • • Human interface to LACS works fine but should be improved (slow response, could be more human friendly, occasional restart of the application) IHM Lass data representation to be controlled Video matrix migrated to a dedicated PC System diagnostics still have to be improved Token distributor issue: - problem under investigation: • Either common mode in key distributor • or evolution not in phase between HW & SW • Contractors and supplier contacted • Some minor non-conformities to be fixed • Access point software - Mechanical : equipment mounting (locks), PAD fine tuning… Information stickers User information and local IHM to be improved 18 January 2008 EDMS: 888085 13

LHC ACCESS SYSTEM LHC ACCESS SAFETY SYSTEM • The LHC Access Safety System behaves as specified • A few relays had to be changed, a few problems solved on line • Diagnostic module and interface with TI still to be developed • Homogenization to be done between the LASS software installed for TI 8 test (LHC 7 & 8) in November 06 and the version deployed on the other LHC sites (LASS V. 2) 18 January 2008 EDMS: 888085 14

LHC ACCESS SYSTEM Next stages • LASS - Once the key distributor issue is corrected, the LASS can be used as it stands today to ensure the safety of the personnel in the LHC. - Preparation of version 2 including all modifications/upgrades • required by AB/OP (patrol behaviour, evacuation instead of BIW in case of two EIS beam are Unsafe…) • Sectorization changes (new doors…) • Minor software upgrades • Ideally before the LHC start-up • Expect 6 months of development and testing of new software version • On site - Personnel permeability of the 160 access sectors – i. e. closing the small loopholes still open. • To be followed up by DSO and OP acceptance tests 18 January 2008 EDMS: 888085 15

LHC ACCESS SYSTEM Other lessons learned • Training for using the access system - All users : film prepared, soon integrated in the CERN training - Firemen : • How to enter the machine without loosing the patrol - procedures, training and keys… • Inform firemen before and during the tests 18 January 2008 EDMS: 888085 16

LHC ACCESS SYSTEM To be done • Test specific TI 8 loop as done for LHC 2 • Test delegation operation by the experiments • Test all the fire doors and crinolines when installed • Install Material Access Device personnel detection • Interface with the Atlas Safety System. 18 January 2008 EDMS: 888085 17

LHC ACCESS SYSTEM Magali’s conclusion (AB/OP) • Excellent collaboration between TS and AB • Overall a very successful and interesting week • To be followed up by DSO and OP acceptance tests 18 January 2008 EDMS: 888085 18