Rick, Peter, Jason, Jeff B.

   A series of apparent glitches and dropouts of the PSL chiller water flow caused the PSL to shut down several times. We swapped out the online chiller (Gromit) with the backup chiller (Wallace) from 05/07/15 until 05/20/15. During this time, we replaced Gromit’s Central Chiller Controllers (for both the Crystal and Diode chillers). The older turbine flow sensors, which were flagged as the source of the PSL shut downs, were also replaced with new vortex flow sensors.

   After the upgrades, we swapped Gromit back in as the online chiller and put Wallace into backup. Gromit has been running for the past 8 days without any glitches or dropouts. The turbine flow sensors in Wallace will be replaced by new vortex flow sensors. 

   The Di-Water cartridges in both chillers have not been replaced in sometime, and there was concern for their proper functioning. The attached plots show the 180 and 30-day trends for the flow rates and the conductivity of the Di-cartridge’s for both chillers. 

   The Diode chiller is functioning within operational parameters. The conductance ranges between 4 and 7 µs/cm (the low/high set points) during a 14 to 15 day period. The flow rate is very steady at 21 l/min. The fluctuations at the right of the plots are when Gromit and Wallace were swapped on 05/07 and then swapped back on 05/20. The conductance appears to be normal, although we need a longer periods of data to reconfirm the previous trends. The flow rate appears to have climbed from 21 l/min to around 30 l/min. The flows through the diodes are all normal. This appears to be a calibration issue with the new vortex flow sensors. 

   The data for the Crystal chiller appears to have the same patterns as the Diode chiller, with two notable exceptions. (1) The conductance rise and regeneration period is three to four months (vs 14 to 15 days for the Diode chiller) and (2) the flow rate for the Crystal chiller has dropped from 18 l/min to 10 l/min. Again, this is likely to be a calibration issue as the flows through the PSL components are all in normal ranges.
 
   We will discuss this calibration question with the laser group in Germany.