1. Performances of the French synchrotron light source SOLEIL Actions to enhance the reliability of the equipment
J.F. Lamarre*, E. Patry
X. Delétoille, A. Bence, R. Cuoq, Y. Denis, T Marion, D. Pereira, S. Petit, G. Roux, C. Tournier
Synchrotron SOLEIL, Gif-sur-Yvette, France
Performances of the French synchrotron light source SOLEIL has been delivering photon beams to the users since January In constant progress, the number of beam operation hours in 2012 has reached hours including a total of hours dedicated to the 26 beamlines. Five modes of operation are available for the users and all are in Top-Up injection mode. The maximum beam current provided in Hybrid mode (3 quarters + 1 single high intensity bunch) was raised to 430 mA, while a total current of 500 mA is possible in the uniform mode. In parallel to the constant evolutions of the accelerators, different actions were led in order to enhance the reliability of the equipment (especially power supplies) and to reduce accidental beam losses. Thus 2011 had a record high efficiency of 98.4 % and a MTBF of 47 hours. The very good reliability of the RF solid state amplifiers contributes to these good results, several other facilities migrate to this technology pioneered by SOLEIL.
In 2012 was hit by several long term beam interruptions due to electrical power outages and to failures of the cooling systems reducing the beam availability down to 96.1%. Strategy and remedies for improvement have been identified and applied or are in progress.
All the filling modes are in Top-Up. In 2012, the multibunch mode has been 100% performed with a hybrid filling pattern, currently fixed to 430 mA. Intensity should be raised up to 450mA during 2013.
An average of 24 beamlines a day out of the 26 SOLEIL beamlines uses the beam. The beam is delivered in Top-Up mode, with a reinjection every 4 minutes, the current is maintained better than 0,5%. To stabilize the electron beam to better than 1µm in position and 1 µrad angle, to maintain and control the beam size and the lifetime, irrespectively of the configuration of the 25 insertion devices, 5 feedback systems are running (SOFB, FOFB, FB-Tune, TFB, FB-Coupling).
The installation continues to evolve. With the construction of two long beamlines (150 and 200 meters) with two in-vacuum undulators installed on a canted long straight section. Also in progress, the installation of Femto-Slicing which will produce shorter bunches, and whose the commissioning is scheduled in 2013.
RF System
After 6 years of operation, the innovative design of Solid State Amplifier (SSA), which was developed in house, has proved itself and demonstrated that it is an attractive alternative to the vacuum tube amplifiers, featuring an outstanding reliability and a MTBF > 1 year.
A transfer of technology agreement was concluded with an industrial company. ESRF has already adopted this technology at 352 MHz. Amplifiers at 500 MHz are being built for ThomX (50 kW) and SESAME (80 kW). Prototypes of amplifier modules at 176 MHz and 200 MHz are also developed.
Power Supplies: first cause of failures in 2007 and Actions have been taken to improve their reliability such as changing components and purchasing additional spares.
A new dipole power-supply, has been in-house developed and is currently undergoing tests. Switching between power-supplies takes less than 5 minutes.
Another spare power supply is under development to be used in case of failure of one storage ring sextupole power supply (10) or dipole power supply of the booster to storage ring transfer line.
After suffering from quite a few IGBT explosions on 3Hz Booster dipole power supplies, a campaign to upgrade IGBT class (400A to 600A) as the first step. But was not sufficient, the number of operating cycles is larger than expected. A redesign of these units is progressing, especially by oversizing the IGBT modules and by improving the cooling system.
The solid state amplifier consists in combination of a large number of 330 W elementary modules.
4 x 724 modules -> 4 x 180 kW ( Storage Ring )
1 x 147 modules -> 1 x 35 kW ( Booster )
Built in-house Spare dipole power supply: 580A / 610V +/- 10ppm
Topology based on 2 interleaved step-down converters 10kHz
In the context of preventive maintenance, some solders were redone and some capacitors replaced in the amplifier modules; the progressive replacement of the actual transistors by more robust ones is scheduled in a second stage.
Spare compressed air
An interruption, for few minutes, of the superconducting RF cavities liquefier generates their unavailability during several hours. Equipment have been redundant like a spare compressor and the installation of a compressed-air system based on simple nitrogen bottles.
Power outage
Partial power outage and power cuts are a leading cause of breakdowns. Despite increasingly strict enforcement of storage ring RF system and power-supplies, beam losses due to power outages are continually growing. Even more concerning is the impact of power cuts, 3 in 2012, which causes an average beam interruption of 10 hours. To alleviate this issues some technical solutions are explored, and using the experience of other facilities.
Machine protection system
The introduction of a delay on infrared mirrors protection system has reduce by 5 the beam losses due to erratic beam swellings. Since then a delay has been extended to other interlock system equipment such as BPM, flow meters, thermocouples and ion pumps.
Spike BPM
Cooling system
By the very end of 2012, SOLEIL suffered from its longest failure, nearly 60 hours without beam. A tap located on the beamlines cooling system broke. Because of PVC tubing we had to wait until the glue used to repair dried. In 2012, there were more than 100 hours of outage due to cooling system and air compressed problems. Several correctives actions have been identified and launched. An audit of the facility will be performed soon.
Number of beam instability losses between 2010 and 2012
47 beam losses in 2010, 19 beam losses in 2011 and 9 beam losses in 2012
Introduction of a delay
on triggers of IR mirrors TC
Balance-Sheet in 2012
In 2012, after utilities and power supplies, RF system was the third cause of failures. Most of these failures were concentrated during 2 consecutive weeks, caused by bad contacts on cavity diagnostic cables, which was difficult to identify.
MTBF is constantly increasing. However, after very good results in 2011, efficiency and MTTR outright declined during If we remove breakdowns due to the electric grid and utilities, 2012 results should have been: Efficiency = 98.9%, MTBF= 75h, MTTR=00h50. *: