Overview TDR Objectives and Run1 2 experience Mark

Overview TDR Objectives and Run-1, 2 experience Mark Lancaster : University of Manchester/UCL 9 May 2019 : g-2 Operations Review

Aims as stated in TDR Statistical uncertainty and systematic uncertainty both 100 ppb. With same value as BNL would give 6. 9�� discrepancy wrt to SM. 100 ppb (stat. ) requires a dataset x 18. 5 that of BNL and with 15% systematic runs dataset x 21 that of BNL. “. . it’s extremely sensitive to new physics. It’s still running, but if I were to put my money on something that would signal new physics, it’s the g-2 experiment at Fermilab. I think it’s really fascinating. . . ” 2 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Significance vs error budget for BNL value TDR Run-1 3 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

If measure 1 sigma below BNL value Need O(10) BNL and syst. < 175 ppb to get to 5�� SM discrepancy 4 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Robust predictions Statistical uncertainty predictions are on a robust footing. 11% of BNL Extrapolating run-1 data to x 18. 5 BNL predicts 101 ppb stat. error vs 100 ppb in TDR. 5 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Commissioning : Jun 2017 – Mar 2018 Run-1 PHYSICS Run-0 Commissioning Run-1 Commissioning 5 orders of magnitude improvement in muon flux TDR envisaged 200 days of commissioning (June-17, Nov-17 Mar-18) 6 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Accelerator Performance 7 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Injection / storage Inflector Magnet μ+ Q 1 7 cm Q 4 K 1 K 2 K 3 7. 1 m cancels 1. 45 T main field x 3 (40 -55 k. V) radially centers beam Kicker magnets 15 -28 k. V Q 2 vertically focusses beam Focussing quadrupoles Q 3 Collimators 8 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Multi-disciplinary experiment Traditional HEP/NP 9 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Collaboration Muon g-2 Collaboration 201 collaborators at 33 institutes in 7 countries 57% US, 43% non-US. FNAL is 17% of the collaboration. 10 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Operations Org Chart 25% non-US, 20% FNAL, 55% US institutes in operational roles 11 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

ESH One experiment liaison and close communication with PPD ESH Oct-20 : Work Pause Nov-13 : HPI Report Nov-14 : MC 1 Access resumed Nov-27 : MC 1 High Bay Access resumed Dec-28 : Resumed kicker/quad work Jan-28 : Magnet turned on Resumption of shutdown work delayed by 6 -8 weeks. 12 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience Oct-19

ESH Two new training courses : MC 1 -HB, MC 1 -MR Improved work planning / oversight. 13 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Improved role articulation Run-2 aim: Double Run-1 dataset x 3. 3 BNL (raw) 14 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-1 Raw data: x 2 BNL but several different quad/kicker settings. Resulted in 7 datasets with approx. x 1. 4 BNL. Run-2 aim: fewer datasets with constant conditions 15 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-1 On average 35% of e+/fill postulated in the TDR (1, 100 e+/fill) 16 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-1 to Run-2 Run-1 issues affecting integrated stats (& systematics/ease of analysis) - kick was too low - kicker had significant downtime - quad sparks - magnet downtime due to cryo issues Such that fraction of days with > 100 M e+ (TDR: 800 M) was 57%. The run-1/2 shutdown work aimed to address these issues 17 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Field Performance More uniform and stable than Run-1 RMS Dipole=13. 8 ppm (Run-2), 16. 3 ppm (Run-1) Insulation added has reduced temperature dependence 18 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Field Performance Run-2 Run-1 19 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-1 to Run-2 Cryo : expected downtime now 2. 5 days/month on average. Kicker running at 142 k. V. Run-1 average was 125 k. V. Aim : 165 k. V. Ran briefly at 160 k. V. 20 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

e+ / fill Run-2: 1. 35 x Run-1 from DQM This despite a 19% reduction in intensity due to running the Li lens with 10% lower current 21 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

e+ / day 5% BNL/day When accelerator down for more than 8 hours Run-2: 1. 52 x Run-1 per day 22 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

TDR vs Run-1 vs Run-2 TDR Run-1 Run-2 Mean e+/fill 1100 400 539 Total # days prodn. running 200 106 38 --->60 800 M 181 M 274 M 430 M 483 M 50% 62% e+/day/avg e+/day/max Uptime 72% Downtime sources: g-2 systems, accelerator, cryo, g-2 trolley runs TDR Uptime: x 0. 9 g-2 systems, x 0. 85 accelerator, x 0. 94 trolley 23 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-1 to Run-2 Combined uptime of kicker/quads is now 91%. 24 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

We have a reduced per fill intensity wrt TDR e+/fill TDR Effect Factor Wedges 1. 06 Li Lens 0. 81 Kick (142 vs 155 k. V) 0. 84 Quads (18 vs 28 k. V) 0. 92 Actual beamline apertures 0. 8 TOTAL 0. 53 1100 RUN-2 Predicted RUN-2 Actual *includes systematic runs where rate was lower. 582 539 +/- 58* This is well understood both in terms of g-2 systems and accelerator 25 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

And a reduced uptime wrt TDR g-2 systems TDR Run-2 0. 9 0. 77 0. 92 (DAQ), 0. 91 (kicker/quad), 0. 92 (cryo) MI Cycles 1. 4 vs 1. 33 sec Trolley Runs 0. 95 0. 94 Testbeam Users 0. 94 0. 91 Accelerator uptime 0. 85 0. 80 TOTAL 0. 72 0. 5 TDR expectation: 1, 100 e+/fill and 72% uptime 800 M/day Predicted Run-2 : x 0. 49 (e+/fill) and x 0. 69 (uptime) 270 M/day Run-2 actual : 274 M/day (3. 2% BNL/day). Run-2 peak: 5. 6% BNL/day 26 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Testbeam Rate reduced for 10 secs from every 60. 27 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

AD Uptime Run-2 : 80% g-2 AD AD-study g-2 AD 28 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

AD overtime Downtime in last 30 days approx. 147 hrs. Time to fix problems approx. 54 hrs. O/T restriction has reduced accelerator uptime for g-2 from 92% to 80%. 29 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

What can we reasonably expect to improve in Run-3/4 Accelerator uptime to be closer to 90% not 80% assuming overtime restrictions don’t exist in Run-3/4. Modest g-2 DAQ improvements: x 0. 92 x 0. 95. Optimisation of upstream wedges: x 1. 06 x 1. 15 Moves average BNL/day from 3. 2% to 4. 0%. Other potential improvements: - temperature control (critical for field & calorimeter gain) - install new inflector: x 1. 4 - increase quad / kicker voltage : x 1. 1 ? ? - faster switching PS (mitigates testbeam) : x 1. 05 30 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-2 Extrapolations For Run-2 assuming 3. 6% BNL per day (overtime restriction lifted) x 2. 2 BNL 5/14 running starts Saturday 2. 3 BNL x 3. 5 x 2. 25 Run-1 We would have met our goal without 5/14 running. 31 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Run-3 / 4 extrapolations Ru n 3 Ru n 4 Assume 4%/BNL day i. e. AD overtime restriction + modest DAQ/wedge improvements. N. B. Peak Run-2 is 5. 6% day. Gets very close to 21. 5 BNL goal And reach goal if new inflector deployed in Run-3/4 shutdown 32 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience Run-3 start Oct-1 and ends early : May-15 Run-4 shared with Mu 2 e 6 months g-2 & 3 months Mu 2 e

Possible Run-4 with �� -4 n u � ) (� R Ru n-3 (�+ )� Run-4 could alternatively accumulate x 8. 5 the BNL �� - sample e. g. if becomes systematics limited with �� + or �� + result > 5�� Requires extra work in Run-3/4 shutdown - Cryo pumps - Kicker standoffs New inflector would give x 12 BNL �� sample. 33 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

2018 Review (Inflector) Recommendations Add a list of specialized/long-lead time parts to the appropriate procedures and obtain/order them in advance to assure no schedule delays. Finalize negotiations of the required personnel resources with Fermilab and involve g-2 institutional technical resources wherever they can be used. Negotiations concluded with resources being prioritised for the kicker work and the inflector work was discontinued in favour of the kicker work. Update procedures and JHAs based on the review committee comments. Under the expected tight schedule requirements, continue to assure that all safety rules and regulations are followed closely, and all safety documentation is up to date. 34 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Computing With anticipated increase in data recorded and in light of experience there areas where SCD expertise could help expedite the analysis. We need twice this Wai ting Qu to b e qu eued eue d jo bs 1, 400 grid slot quota Using HPC for simulation - 15 mins of data presently takes 1 week to simulate ART/SQL interface Access to external grid resources 2 Gb footprint Running jobs Code optimization/speed. 35 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience

Summary We are now operating smoothly : accumulating 1 BNL per month at rate x 1. 5 that of run-1. Run-1/2 shutdown work significantly improved kicker and cryo performance. Wai Run-2/3 shutdown work tingmodest in scope to ensure we can begin to b e qu eper ued day. taking data on Oct-1 at 4% BNL Qu e job depend on outcome of Run-1/2 analyses. �� - vs �� + Run-4 runninguedwill s 36 05/09/19 Mark Lancaster | TDR Objectives and Run-1, 2 Experience