John and I started up HU-2 yesterday at Y End to add some humidity to the VEA. In the process of starting this unit up we removed the lid on the reservoir and found that a large piece (~9" square) of bubble wrap had been left in the water discharge area of the reservoir. We removed the bubble wrap and started the humidifier. It is set at ~50% (5V on a 0-10V scale). The 5 day plot shows an increase in the humidity. Note:The 14 KW heater is powered from a electrical panel located in the VEA and was cycling several times a minute.
Tagging DetChar, OpsInfo, and ISC. There is suspicion from DetChar and Schofield that humidity, or the lack there of, at EY is causing electronics to glitch which propagate through the interferometer. So this *may* start to improve the IFO glitch rate. Maybe. Also note the "cycling several times a minute." The humidifier's water heater is power from the YVEA, and it's controlled with a bang-bang servo, so look for changes in electronic / electromagnetic signals that show some characteristic ON / OFF signatures. Again -- keep your eyes peeled!
We have been working on a low latency blip glitch hunter. It uses PyCBC Live but with a smaller template bank focused on finding possibly only blip glitches. We also generate summary pages, updated every 5 minutes:
https://ldas-jobs.ligo.caltech.edu/~miriam.cabero/blips_live/H1/day/20170209/detchar/pycbc_live/
Unfortunately it has only been running stably since Feb 3, so we cannot really compare directly if the humidifier added on Feb 1 made a significant change in the rate of blips. However, I started using this code on Jan 30, so there are some results for previous days but the code was still a little bit unstable. It can be looked at anway:
https://ldas-jobs.ligo-wa.caltech.edu/~miriam.cabero/blips_live/H1/day/20170131/detchar/pycbc_live/
By glancing at this plot from Jan 31
and from e.g. yesterday
it doesn't look like there has been an improvement...