Scope Review (draft): Purity Monitors for DUNE Jianming Bian (UC - - PowerPoint PPT Presentation
Scope Review (draft): Purity Monitors for DUNE Jianming Bian (UC Irvine) Andrew Renshaw (U of Houston) Stephen Pordes (Fermilab) 1 Motivation of Purity Monitors Cryostat Operation (cryostat+inline): monitor argon filling during
1
2
3
Start to Monitoring purity right after bottom PrM submerged LAr, found filter saturation during the filling à report to cryo people to regenerate cartridges
times a day, alerted the experiment solely to serious problems several times
level gauges. prevented situations which otherwise would have continued unnoticed for some time, with severe consequences to the ability to take any data. Neither the gas analyzers nor the TPC caught these problems in time.
benchmark LAr purities for recirculation studies and TPC calibration, measure e-lifetime for TPC
recirculation study
4
5
As soon as the lowest purity monitor was immersed: ~40 us -> 7.5 ppb O2eq On Thursday 30st of August purity was compatible with ~60 us Cathode signal Anode signal ProtoDUNE-SP: On Friday 31st of August, 2018 the purity of the bulk liquid argon dropped from 40 us purification cartridges needed to be regenerated. Regeneration took till the 3rd of September. Filling restarted immediately after. Filippo Resnati - DUNE Collaboration Meeting - CERN - 28th January 2019
Monitor purity during LAr filling, find saturation during the filling
6
During data taking, ProtoDUNE-SP, PrMs have alerted the experiment solely to serious problems several times (dips): Recirculation pump stoppages, false alarms, and problems from the cryostat-level gauges. These alerts are critical to the ProtoDUNE-SP project's success, as they prevented situations which otherwise would have continued unnoticed for some time, with severe consequences to the ability to take any data. Neither the gas analyzers nor the TPC caught these problems in time.
7
Mar 12, 2019 8
reduced pump speed All boil-off filtered pump off All boil-off vented pump off All boil-off GAr condensed and returned Pump restart P1 P2 P3 P4 P5
Ilsoo Seong
PrMs are sensitive to purity change Typically no regular TPC data when testing recirculation
9
10
Bottom PrM Mean = 4.85ms RMS = 0.19 ms Mid PrM Mean = 6.20ms RMS = 0.40 ms Top PrM Mean = 6.99ms RMS = 0.36 ms TPC lifetime 2 Mean = 6.1ms RMS = 2.0 ms PrM : 200 flashes/measurement TPC lifetime 1 Mean = 4.6, 6.1ms RMS = 1.0, 1.4 ms
PrM e-lifetime stat. error is much smaller than TPC because it measures localized purity with large statistics
11
Choose TPC cosmic runs under same PrM purities for e-lifetime calibration Important for DUNE because cosmic rate is low
12
Electron lifetime: tau = 1/k_A*ns , where k_A is the electron attachment rate and ns is the concentration of a certain type of impurity. Attachment rates k_A at different E-fields are different, make electron lifetime and Qa/Qc different in TPC and PrM different due to their different operation HVs Develop TPC-PrM combined lifetime measurement: (Qa/Qc)_TPC= f*(Qa/Qc)_PrM f obtained from fit to (1/tau0-1/tau)_TPC vs. (1/tau0-1/tau)_PrM in data, space charge effects largely cancelled Result consistent with prediction from SCE corrected e-lifetime time prediction by Flavio and Xiao)
Craig Thorn, LBNE-Doc-4482-v1, Xin Qianhttps://indico.fnal.gov/event/14296/contributi
13
14
τ lifetime = − tdrift log QA QC f RC
A
f RC
C
ftrans !
<latexit sha1_base64="(nul)">(nul)</latexit><latexit sha1_base64="(nul)">(nul)</latexit><latexit sha1_base64="(nul)">(nul)</latexit><latexit sha1_base64="(nul)">(nul)</latexit>Relative uncertainties, affect sensitivity to catch purity change
0.55% in eq. Qa/Qc
15 Absolute uncertainty 5-9% in lifetime at ~7ms, dominated by transparency correction and anode/cathode gain correction Gain uncertainty can be calibrated in vacuum à Will do so when we pull PrM Assembly out to prepare for protoDUNE-SP run2 Transparency correction uncertainty can be prevented if we have longer purity monitors
Hints for e-lifetime stratification
16
Bottom PrM Mean = 4.85ms RMS = 0.19 ms Mid PrM Mean = 6.20ms RMS = 0.40 ms Top PrM Mean = 6.99ms RMS = 0.36 ms TPC lifetime 2 Mean = 6.1ms RMS = 2.0 ms PrM : 200 flashes/measurement TPC lifetime 1 Mean = 4.6, 6.1ms RMS = 1.0, 1.4 ms
PrM e-lifetime stat. error is much smaller than TPC because it measures localized purity with large statistics
What are the limitations of the purity monitors in terms of measuring high purity levels and what are the benefits that would be obtained by implementing proposed improvements to these devices? For Detector Operation (relative measurement):
for LArTPC operation
equivalent to 42 ms lifetime for regular PrM drift time 2.2ms For absolute lifetime measurement:
17
18
Uncertainty of lifetime measurement relative to each purity monitor is small Very sensitive to catch purity change caused by recirculation problems
19 Absolute uncertainty 5-9% in lifetime at ~7ms, dominated by transparency correction and anode/cathode gain correction Gain uncertainty can be calibrated in vacuum à Will do so when we pull PrM Assembly out to prepare for protoDUNE-SP run2 Transparency correction uncertainty can be prevented if we have longer purity monitors
Hints for e-lifetime stratification
Are the proposed number of purity monitors per far detector module (10 total, 6 inside cryostat, 4 inline within cryogenic infrastructure) necessary to meet critical system requirements?
20
21
One of DN250 instrumentation ports on each side, if not available then use part of manhole on each side
Need straight detector port Two strings of purity monitor assemblies on TCO and back sides, each string mounts 3 purity monitors on a supporting tube, in total 6 purity monitors in cryostat Similar system runs successfully in ProtoDUNE-SP Locations for Inline Purity monitors under discussion
22
23
4 purity monitors outside of cryostat but within both in front of and behind the filtration system.
24
25
26
27
Successfully run at ProtoDUNE-PS, find saturation, alert pump stoppage, pump recirculation, measure e-lifetime combined with TPC
Top PrM Middle PrM Bottom PrM Individual PrMon:
electron generation
screening at readout
PrMon cage and flange
Qanode/Qcathode = e-tdrift/t
28
Clean LAr Distribution Dirty LAr pump
29
making electron drift time range from 150 us to 3 ms.
Signals on anode and cathode are 6 times larger.
precision.
no saturation ProtoDUNE PrM signals at e-lifetime = 6 ms
30
Uncertainty of lifetime measurement relative to each purity monitor is small Very sensitive to catch purity change caused by recirculation problems
04/26/17 ProtoDUNE-SP Cryogenics Instrumentation Review 31
Fermilab Particle Physics Division Site Support Department S i z e FSCH No DWG No R e v S c a l e S h e e t Issued Originated: Last Revision: Drawn by: T i t l e B Originator: P r o j e c t Gerard Visser Walter Jaskierny 30 Aug 2004Purity Monitor Electronics Type 2, Two Channel
FLARE 1 of 2 GND 750 pf 15 KV 10 M Ω 10 M Ω 10 M Ω 50 M Ω 750 pf 15 KV 7 3 9, 5 1 2,4 6,8 DZero Preamp 5 pf 20 Meg 16 Aug 2007 Amp 1 Anode SHV Negative Cathode Supply 2kV Max. 100 M Ω 499 Ω *See Notes LM317 7906 249 Ω 1N458 1.33 KΩ 22 µf 15 WVDC 0.1 µf 0.1 µf 0.1 µf 0.1 µf 1N4002 1N4002 0.1 µf 0.1 µf +12 V 4X 2-3/4 CF Flanges (3-HV and signal) (1-Fiber optics) 2X 1/5” VCR (For gas fill and relief) 1X 1.33” CF Flange (Vacuum pumping) Custom support tube adapter, holes for HV cables to be inserted Side view
3XOptical feedthroughs HV feedthroughs
HV Feedthroughs: ~ 10kV Optical Feedthroughs:
fabrication of the parts will being immediately, flange will be first
so no worry about breakage from installation
Flange will have support tube adapter Support tube will also contain 0.5” tube to contain fibers Hole in adapter for HV cables to enter support tube
power supply for the light source.
35
36
PrM electronics Signal: 2 channels – cathode, anode < 5V PrM HV Cathod -150V Anode 2500V DAQ PC 110V Digitizers NIM Bin 110 V
Need to Develop slow control
37
38
To TPC/PDS DAQ
39
"Why measure with purity monitors and not in TPC data”
conditions were to persist, it could cause irreversible contamination to the LAr and terminate useful data
mitigate this risk is unique because the cosmic-ray rate in DUNE's deep-underground FD is too low for TPC to measure e-lifetime frequently with its data.
fluctuation in the online e-lifetime from TPC data makes it hard to catch the purity change caused by liquid argon recirculation issue in time. On the other hand, after lower the purity monitor anode high voltage, electron drift time in purity monitors reaches 2.2 ms, the same as the TPC, but the purity monitor can take large statistics in a short period of time at the same location, so the statical error and effects from location dependent uncertainties are smaller in purity monitors compared with TPC. Each purity monitor e-lifetime measurement is based on purity monitor Qa/Qc from 200 UV flashes within 40 seconds at the same location, and the statistic error on the purity monitor e-lifetime is less than 0.03 ms when the purity is 6 ms. With this high precision, purity monitors can catch the purity drop immediately and make the alert to the experiment.
problems several times. The first time was for filter saturation during LAr filling, and the rest were recirculation pump stoppages, false alarms, and problems from the cryostat-level gauges. These alerts are crucial to the ProtoDUNE-SP project's success, as they prevented situations which otherwise would have continued unnoticed for some time, with severe consequences to the ability to take any data. Neither the gas analyzers nor the TPC caught these problems in time.
40
SS tube Stephen, Filippo
Crane Port Use multimeter to test connectivity to feedthrough/topflange for cathode, anode, anode grid and ground with faraday cage Insert slowly, stop when checking connections If fibers are not closely attached to the cathode surface, open the side window of the Farady cage and tune Testing cables x 9 UCI sling <150 lbs