Reducing the North Area Power
Consumption
Adrian Fabich
in collaboration with EPC, CO, OP
EATM, 25. October 2016
Power saving scheme 2016
Reducing unnecessary electrical powering
- Relying on existing infrastructure
Details in EDMS 1610139
- no hardware modifications
(working draft)
- minimum software modifications
• During unavailability of primary beam
• During access to a beam line/zone
• During shot-down periods (MDs …)
Overall saving like 3-4 GWh per NA year
25/10/2016
A. Fabich
2015
SPSFT
DYNECO
FULLECO
2
FullEco fully operational
• Handled by SIS
25/10/2016
– Operational in H2/H4
– Available also for H8 to M2
A. Fabich
3
FullEco examples
• Access on H4
25/10/2016
• Injector stop today
• SPS FullEco
A. Fabich
4
DynEco mode
Implementation
• Implementation based on local timing
• In case of SPS-DynEco, FT Start-Cycle is suppressed to the
power converters
Test results:
• Without Start-Cycle, the power converters drop into
VoltageLimit (error mode)
• Some time-out logic in the internal electronics of the power
converters
Investigations are ongoing …
25/10/2016
A. Fabich
5
Summary
• FullEco mode of a beam line operational
– Fully operational
– Working without difficulties
– Transparent to users
• Further investigating the DynEco mode
• Collaboration with Yves Gaillard (EPC), Ioan
Kozsar (CO), Michal Dudek (CO), Mathieu Gabriel
(CO)
25/10/2016
A. Fabich
6
North Area – consumption
jenergy
~300 magnet circuits
With laminated magnets only and reliable power supplies, the saving
potential is about 20 GWh/y (based on 2015 operation) from magnet power
circuits
• Details presented to the Energy Saving Panel
• The necessary hardware changes are proposed to be included in the
consolidation project study.
… in the meanwhile …
25/10/2016
A. Fabich
8
Technical Implementation
Short-term (single/few cycles) absence of beam (=DYNECO)
– Applied to all NA beam lines
– Acting only on PULSED magnet circuits
– Early BO will cause a much shorter flat top
•
SSC
(DynEco)
•
NA magnet cycle
Long-term absence of beam (longer a few minutes)
– If a beam line is in access mode (defined by XTAX dump position)
– If no SFT% cycle in the SPS-SC (e.g. MD)
– If FullEco of SPS is activated
– SIS requests magnet circuits of (one or all) beamlines into STAND-BY
– Acting only on PULSED/DC magnet circuits (spectrometers not included for safety/operational
aspects)
– In the test phase applied to H2/H4 only
25/10/2016
A. Fabich
9
Risk assessment
Risks induced by this power saving concept:
Failure of re-starting after STAND-BY, physics time lost
Degradation of DC magnet circuits due to additional STAND-BY periods
•
Applying DynEco, the only difference (compared with 2015) is, that the flat top is
finished earlier. Pulsing frequency remains the same.
 No additional risk for failure or degradation
•
Long-term STAND-BY (initiated by SIS)
•
•
•
Impact only for DC magnet circuits
From a few 10 down times in 2015 to a few 100 down times in 2016
Testing the reliability of re-starting after STAND-BY, initially applying to H2/H4 only
•
In case of excessive need for FirstLine interventions, the concept cannot be applied to existing
hardware.
Re-start after an MD period should not exceed the FirstLine effort in 2015 (a few power
converters)
•
The implemented procedures can be revoked within an hour.
25/10/2016
A. Fabich
10
Further needs
• EPC: power supply studies during MDs, exempt particular, single power
supplies from power saving scheme
• An easy suspend mode e.g. for OP and EA for re-starting after an MD
• Studying the impact of this approach on p0/K12 operation; special
ingredients like high-intensity beams, p0-survey (only on long-term
scheme)
• CALS logging of events related to this power saving
• Display power saving mode in CESAR control GUI
• The ongoing investigations revealed some improvements, which could be
included in any future upgrade of NA power converters controls, e.g.
swapping to FESA3
25/10/2016
A. Fabich
11
North Area – consumption
jenergy
25/10/2016
~300 magnet circuits
A. Fabich
12