Speaker
Description
Cooling water is an essential part of the infrastructure of all but the smallest AMS instruments, and of particle accelerators in general. While facilities situated on larger scientific campuses often can rely on good quality water from a central supply, many instruments, including VERA at the University of Vienna, are situated on sites where a stand-alone provision is necessary. Because it is considered part of the infrastructure, little support is provided by the accelerator manufacturers.
VERA, established in 1995, operated smoothly for the first 14 years without paying much attention to water quality. Then, first blockages of magnet coils were observed. While removal by rinsing with commercial descaler liquids was successful initially, problems became more severe over time, culminating in the unrecoverable blockage of a quadrupole coil with magnetite. A major effort became necessary.
Cooling systems of particle accelerators are a combination of materials like steel, aluminum, copper, and plastics. Directly cooled magnet coils apply voltages to the water. Oxygen and CO2 diffuse through plastic tubing. This evokes all kinds of degrading chemical reactions, dissolving copper and iron, and depositing carbonates, magnetite, and elementary copper. Information on these processes and possible remedies is mainly found in literature on nuclear reactors and stator coils of power generators. Specific information on particle accelerators and literature on the best practice for cooling water is scarce (Spencer 2014). Pure water, with pH about 8, conductivity below 1 µS/cm, and dissolved oxygen below 10 ppm seems advisable.
We achieved this goal by replacing all plastic tubing with several hundred meters of diffusion tight multi-layer composite pipes, and EVOH where electrical insulation was necessary. A mixed-bed demineralization cartridge was mounted to remove any newly dissolved ions, and before addition of new water dissolved oxygen is stripped in a self-built nitrogen bubbler. As the complete removal of all accumulated magnetite in the steel tubing seemed impossible, filter frits were mounted before the individual components. Conductivity, pH, and oxygen content are regularly controlled with handheld instruments. The modifications took a lot of work, but minor financial efforts. Problems seem to have diminished since then.
The best method for clearing magnet coils seems to be rinsing with circulating EDTA solution. The solution can be introduced even into completely blocked coils with the help of a suction flask. Deposits of particulates are efficiently knocked out with packets of water pushed through at high speed with pressurized air.
We will present a summary of the relevant literature and describe the methods used at VERA.
C. M. Spencer, Improving the Reliability of Particle Accelerator Magnets: Learning From Our Failures in IEEE Transactions on Applied Superconductivity 24/3, p 1 (2014) doi: 10.1109/TASC.2013.2280923.
| Student Submission | No |
|---|
