Lessons Learned from Recent SSRL Operations Shanjie Xiao, James Liu - PowerPoint PPT Presentation

Lessons Learned from Recent SSRL Operations Shanjie Xiao, James Liu and Sayed Rokni SLAC National Accelerator Laboratory Radsynch 2017, Hsinchu, Taiwan

Overview Lessons learned from three incidents: • Excessive dose from gun chopper current leakage • Burning of BL12 burn through monitor • Alignment error of BL13 movable mask RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 2

Introduction: SSRL Layout SPEAR3: 3GeV 500mA Linac & Booster RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 3

Excessive Doses from Area TLDs 130+ area TLDs for SSRL Booster In April 2016, found quite a number of area TLDs read higher than before • Registered radiation Aug. 2015 – Jan. 2016 Lab Linac RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 4

Excessive Doses from Area TLDs 2 nd Floor of Bldg. 130 NA (NA) 0 (0) Dose in mrem 13 (0) 0 (0) 0 (0) 0 (0) 87 (19) 0 (0) 80 (19) 0 (0) Elevated dose Normal dose ## (##): Total (Neutron) Dose 1 st row for Aug. 2015 – Jan., 2016 103 (16) 2 nd row for Aug. 2014 – Feb., 2015 0 (0) Worst in Linac: 1.6 mSv  37 mSv • 0.8-1.0 mSv in some 2 nd floor offices (w/ • neutrons) Limit: 1 mSv for 2000 hours o Equivalent to 0.6-0.8 mSv for 2000 hours o occupancy • Occupants in the offices were relocated 5 immediately

Reasons of the Excessive Dose SSRL uses a thermionic gun and a chopper to select 10 Hz beams GT2 Chopper GT3 PM D COR3 Slit COR4 COR2 • Permanent magnet in chopper bends beam down as default status GT1 • Trigger pulses select 5-10 bunches from total of 3000 - Bunches that arrive early before the trigger are not kicked and strike below the slit - Bunches that arrive late after the trigger are kicked too hard COR1 thermionic and are lost above the slit Gun gun • The chopper does not exhibit perfect contrast as some fraction of the electrons before and after the chopper trigger can be transmitted through the chopper, and then be partially accelerated. RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 6

Reasons of the Excessive Dose • Typically the number of unwanted bunches leaking through the chopper is minimized during machine startup every year while at the same time optimizing the desired beam performance. • In the 2015 shutdown the gun cathode was replaced and in the process the alignment was unexpectedly changed. • When the desired beam was optimized, a small amount of leakage current was transmitted through the chopper. • Leak after trigger is much less since the chopper is Leak current when chopper is “off” only fired 60 times every 5 minutes. RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 7

Lessons Learned and Mitigations Careful study and control on radiation sources 1 st accelerate section Chopper Mitigations • Create a tuning procedure to reduce the leak: perform at the beginning of each run • Add shielding to shadow offices New shielding on this side • Add a radiation monitor beside Linac • Doses are reduced to normal RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 8

Introduction: SSRL Beam Line Front End Concrete Wall Injection Movable Fixed Synchrotron Hutch Stoppers Mask Mask Beams Grazing incidence, water-cooled mask for SR Burn-through monitor: bladder connected to ambient air Tungsten block for bremsstrahlung • For insertion, movable mask moves in first and then injection stoppers • For withdraw, injection stoppers move out first and then movable mask RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 9

BL12 Burn Through Monitor In November 2015, the BTM of the 1 st BL12 injection stopper was burned, vented SPEAR3 ring, and dumped stored beams • BTM system operated as designed - Burn while inserting the stopper moved into beams - Burn rapidly relative to insertion time - Beam was terminated before damaging stoppers - No personnel radiation exposure • Cause of failure investigated Normal beam center - Insertion command logic ok - Insertion timing tested - No root cause found Horizontally displaced as this front end serves a 10mrad • chicane and only one IVU was installed. Vertically displaced because burn through rapid relative • to insertion time resulting in melt as stopper first intercepted beam at the bottom of the stopper. RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 10

BL12 Burn Through Monitor: Lessons Learned Uncommanded state faults should be trapped • Introduce state fault trap logic If Movable mask not closed and injection stoppers not open  fault, dump beam - Avoid challenging BTM via introduction of MPS if feasible • A refractory metal plate attached to the leading edge of the upstream injection stopper featuring MPS monitored temperature sensors RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 11

BL13 Degraded Vacuum During Fall 2015 startup, vacuum at BL13 degraded when the line was closed (movable mask and injections stoppers inserted) • MM was in an angle larger than design since 2007 • MM was reduced to its design angle 1.2° in summer 2015 • Isolation valve after MM could be slowly heated (125°C over 2 hours) • Increase MM angle to 1.7° resolved the heating issue • Vented the line in summer 2016 - Isolation valve damaged Isolation Valve - Injection stopper BTM locally damaged (not burned) Injection Stopper RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 12

Cause: Movable Mask Alignment Error • MM was not fiducialized in 2007 as expected by design engineer such that when aligned the MM did not completely occlude the fixed mask aperture - Mask center should be aligned to vertical beam center - Aligned to mask tip • Larger pitch angle before 2015 coincidently compensated for fiducialization offset BL13 Movable Mask • Other beam lines are checked to be ok RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 13

Lessons Learned High power undulator beam line masks operate at small grazing angles calls more frequent alignment verification • Annual start up alignment checks for high power density undulator BL Improve alignment communication • More explicit documentation of feature positions rather than overall position information to ensure clear and complete communication of alignment results Increasing undulator power has pushed mask designs to operate at low intercept angles to maintain cyclic thermal stress at the endurance limit • The alignment tolerance risk is more significant than the cyclic stress risk. • Similar to other light sources, accept sub-endurance cycle lifetimes in exchange for better mask mis-alignment tolerance. RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 14

Summary Excessive dose due to gun chopper current leak • Need better understanding and control on radiation sources (Chopper OFF) BL12 burn through monitor • Need state trap to catch uncommanded faults BL13 movable mask alignment error • Need more frequent check on beam lines with small alignment tolerance • Need to improve alignment communication • Need to balance the alignment tolerance risk and cyclic stress risk This work is supported by the U.S. DOE, Office of Science, Office of Basic Energy Sciences, under Contract No. DE- AC02-76SF00515 RadSynch 2017, NSRRC, Hsinchu, April 19-21, 2017 15