Analysis of High -Intensity Proton Accelerator Availability over the … · 2018-11-21 · Injector...
27
WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN Angelina Parfenova :: Large Research Facilities :: Paul Scherrer Institut Analysis of High-Intensity Proton Accelerator Availability over the Last 15 Years of Operation @ PSI ARW 2017 Versailles 20.10.2017
Transcript of Analysis of High -Intensity Proton Accelerator Availability over the … · 2018-11-21 · Injector...
WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN
Angelina Parfenova :: Large Research Facilities :: Paul Scherrer Institut
Analysis of High-Intensity Proton Accelerator
Availability over the Last 15 Years of Operation
@ PSI
ARW 2017 Versailles 20.10.2017
Layout of the high intensity proton facility at PSI
Page 2
Paul Scherrer Institute - PSI
PSI Control room How do we produce a 590 MeV CW beam of more than 1 Megawatt
PSI aerial view
Page 3
Experimental Hall
PSI research covers the following areas: - Particle physics and astrophysics - Solid-state research and material science - Energy and environmental research - Biology and medicine
PSI develops and operates complex research installations: - The 590 MeV proton accelerator facility (HIPA) with spallation Neutron Source (SINQ) - The Synchrotron Light Source (SLS) - Proton Therapy facility (Proscan) - Free-electron X-ray laser (SwissFEL)
Proscan
Injector
cyclotron
72 MeV
Ring
cyclotron
590 MeV
Targets
SINQ
SLS
SwissFEL
Control room at PSI
Page 4
HIPA = high intensity proton accelerator
SLS = Swiss light source
Proscan = 250 MeV proton accelerator for biomedical applications (cancer treatment)
Infrastructure & personal safety
SWISSFEL Project
HIPA development
o Beam current increased from 100 μA (1974) to 2400 μA @ 590 MeV
(stable test operations 2015)
o In 2015 the average beam current to meson production targets was about 2100 μA
with the beam loss under 400 nA
Figure: courtesy of M. Seidel (PSI)
Page 5
2400 μA
o Mesons (possible with beam on SINQ or beam dump <1.7mA)
o Neutrons SINQ with 60% or 70% of the main beam depending on the mesons target (4 or 6 cm)
UCN full beam on UCN target is kicked (typically 8s every 5 min.)
o Isotopes production IP2 (splitted or direct 72 MeV-beam under 100 mA): 40 MeV (Degrader), 72 MeV
High Intensity Proton Accelerator
Page 6
Cockcroft-Walton
870 keV, 10 mA
Injector 2: 72 MeV, 2.7 mA, 4 cavities, 82 turns
beam current can be adjusted by a moving collimator
Isotopes production 2
40 or 72 MeV, 100 mA
Injection beam line 2
72 MeV, 2 mA Ring Cyclotron: 590 MeV, 2.4 mA, 4 cavities (1
MV ), 1 flattop cavity, 184 turns => low losses at
extraction (efficiency 99.98%).
Layout and installation: courtesy of C. Sattler
SINQ UCN
Targets, experimental areas
Beam dump
Multiple modes of operation:
HIPA user annual operations schedule
Shutdown from 24. December till 8. May (128
days)
Service and machine development shifts
Begin of the next shutdown
200 days of user-operation
Page 7
Run weekly operational status meetings
Which metrics do we use to analyze operational data
Page 8
Metrics, Availability and Analysis
Availability over the last decades What do we see from our analysis
We measure operational efficiency by monitoring the
following Key Performance Indicators:
Operational metrics
Page 9
o Beam availability o Current integrals at meson targets, SINQ, UCN, IP2 o Trip rate (number of trips)
Find “Happy-User-Index”! A. Lüdeke (PSI)
Availability 2001 – 2017 (I> 1 mA)
Electrostatic elements
RF-discharges in the
Ring cyclotron and
deformed trim coils
Flattop
Cavity
problems
Availability averaged 89%
Electrostatic
septum
discharges,
beam dump,
SINQ problems
Page 10
Availability record
New intensity record I = 2.4 mA
Availability is defined as: beam time with current I > 1 mA extracted from
the Ring cyclotron over the scheduled beam time
Interlock system
malfunctioning
Current integrals
beam trips
Trips due to discharges at the septum’s reach from 300 trips/week to 2000 trips/week (after) breaking the vacuum, or
while damaged by the beam). Beam trips have an impact on the availability. To power up the machine again takes
about 60s. Only 50 beam trips per day means 2% lost availability! Page 11
Electrostatic elements Flattop
Cavity
problems
Electrostatic septum
discharges, poor vacuum
Interlock system
malfunctioning
Poor vacuum in
the Ring cyclotron
Ch
arge
Current integral records
Operational events in detail, that have impacted availability over the last decade 1 MW facility, what challenges do we have What problems did and do we have 2016 operations What do we do to successfully operate HIPA facility
Page 12
Problems with electrostatic elements (2005-2006)
o Discharges with switched on RF
o Ionized gas in the surrounding of the
electrostatic elements
o Defects on insulators
Extraction
Injection
Page 13
Shielding of the electrostatic elements against:
o Decoupled RF-Power
o Ionized gas (plasma) Photos: R. Kan, D. Götz (PSI)
o Deformation because of a failure of the cooling circuit
o Repair would take up to several week, continued operation (under 2.2 mA!)
Deformed trim coils of sector magnets (2010)
reduced gap (ca. 1.8 cm)
Page 14
Beam- stopper
flattop cavity
EEC
plasma region
RF-discharges in the Ring cyclotron (2010-2011)
RF discharge ignites at 470 kV below the
operational flattop voltage of 555 kV
Page 15 Photos. R. Kan (PSI)
X-rays emission at the flattop cavity (2014-2016)
Loss monitor
Page 16
Painting with AQUADAG (yearly)
o Suppression of multipactoring
o Reduction of X-rays emission
o Reduction of discharges at the
electrostatic elements
o Less decoupled HR power
1.5 1.7 1.9 1.11
0
100
200
300
400
500
600
700
MR
I11
(n
A)
date (2016)
Ion
izat
ion
ch
am
be
r
Problems with the flattop cavity (2013-2014)
Page 17
o Shutdown 2015
o Dismount of the flattop cavity
o Service at the tuning-system (hydraulics)
o Vacuum (new seals)
o Tightening up of the electrodes
09.06.2016: 2.4 mA with BAG-License
1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12
0
500
1000
1500
2000
2500
Targets
SINQ
curr
ent (m
A)
date (2016)
0
20
40
60
80
100
Operations 2016 12. - 20.6.: AHA and Target E exchange
25. 6.: SINQ-shutdown 9.10.: vacuum pressure increase in the area of the beam dump
Beam dump-operation Imax < 1.7 mA 31.10: recommissioning SINQ
o Limited to 5 Ah total charge
o Current density < 30 mA/cm2
o max. 29000 thermal cycles
Electrostatic septum discharges every 20 minutes
Page 18
Number of outages 2012 – 2017
Page 19
2012: Change to
EPICS Control
System
Flattop cavity
Aquadaq flattop cavity
Integral outage representation averaged over 2004 – 2016
Cooling 16%
HF 14%
Electrostatic elements 13%
Targets 10%
Vacuum 8%
Ion source 5%
Controls 5%
Power supplies 5%
PSYS 3%
Miscellaneous 3%
Site power 3%
Magnets 3% Others
Page 20
Operational tools
Page 21
o Beam loss monitors important to protect against misbehavior of the beam and
also gives appropriate feedback to the operator for minimizing the losses
o Phase measurement devices beam dynamic and tuning tools
o Beam position monitors automated beam
steering using bpm’s, steering and bending magnets
o Feedback tools for stabilizing phases, ion-source, …
o Beam dynamics tools
Ref. A. Parfenova, C. Baumgarten, M. Humbel, A. Mezger, A. Petrenko, “Measurements of the Beam Phase Response to correcting Magnetic Fields in PSI Cyclotrons“, TUPMR019, IPAC 2016.
J. Snuverink (PSI)
New ECR-ion source since 2009
o Longer intervals between services ( > 8 weeks)
o Higher beam stability
o More proton efficiency (80% instead of 33%)
o Better beam quality (smaller emittance)
o Up to 2.7 mA from Injector 2
Page 22
Steps to achieve higher availability and currents
Page 23
Improvements and Operations
Safe operations and operational risks
Conclusion
Steps to achieve higher availability
Page 24
o Enough time for the setup after the shutdown (at least 1 week)
o Longer beam time periods (4 weeks between maintenances)
o Lessons learned: Dismount of the electrostatic elements at the beginning of
the shutdown. Storage of the electrostatic elements in a vacuum and under
the high voltage. After RF-conditioning switching on the electrostatic
elements and then slowly ramp up
o Injector2 upgrade with new resonators (spare parts are not provided
anymore, better beam quality)
o Continuous updating and modernizing the operational tools (GUI, panels,
programs)
o Operation 2030 +
o Higher acceleration voltages (1.2 MV possible → less turns, less beam loss)
o For this a new flattop-system will be needed (limited now by Umax = 550 kV)
o New trim coils (secondary electron emission, ionized gas)
o Target, collimators upgrades, SINQ-Target (Imax > 2.4 mA), beam dump
o Injector 2 resonator exchange → better quality beam injected into the Ring
cyclotron
o Machine developments (simulations, diagnostics, mutual work
beam dynamics and operation)
o Energy efficiency of the proton facility
Today:
2.4 mA, UKav= 850 kV, 186 turns
ηacc ≈ 0.18
High current -Upgrade:
3.0 mA, UKav= 1 MV, 160 turns
ηacc ≈ 0.2
03
2
/
/
PqEIIc
qEI
P
P
kin
kin
grid
beamacc
3
max kavUI
Page 25
Steps to achieve higher currents
Ref. A. Kovach, J. Grillenberger, A. Parfenova, and M. Seidel, “Energy Efficiency and Saving Potential Analysis of the High Intensity Proton Accelerator HIPA at PSI”, TUPVA129, IPAC 2017.
Overview of HIPA Operations
PRESENT STATUS:
o HIPA delivers up to 1.4 MW beam power o Average availability is 89% (maximum availability reached: 95% in 2015) o Maximum power increase was reached through the increase of cavity
voltages o Maximum beam current is presently 2.4 mA (licensed for routine operation) o Modular design allows fast repairs in case of failures o Good infrastructure (radiation protection, shielding, disposal)
OPERATIONAL RISKS:
o Careful planning of maintenances (time for repair, availability and personal doses)
o Spare parts problems, because the facility is aged (magnets, flattop cavity, power supplies, cooling )
o Man power and retirements (loss of Know-How)
Page 26
Page 27
Creating Knowledge Today for Tomorrow
My thanks go to
• All my colleagues that
contributed to this
presentation
• PSI Operations team
In particular to:
• Andreas Lüdeke
• Joachim Grillenberger
• Richard Kan
• Anton Mezger
• Mike Seidel
• Andras Kovach
Thank you for your attention
