We extracted from ENV channels some indications about the nature of the structures in  the frequency noise which is coupling to DARM.

Figure 1 indicates the peaks in DARM (MICH) we have investigated.

Figure 2 shows the ENV coherence projection of RFC channel. Indicated are the structures which are common to MICH and DARM noise, as reported in 37537.

• 20Hz, large cohence with seismometers on the External Inj Bench, it looks like a bench resonace, Figure 3
• 24.6Hz, looks like a fan or scroll pump. It is present more or less in all accelerometers in CEB plus the microphone in DET lab. Antonio says there are only two active Scroll pumps at present: one is in the MC tunnel close to MC end (but the absence of coherence with MC environemnet would exclude it) and the other one is the CEB remote scroll. Figure 4.
• 29.8Hz, this is NOT present in RFC signal and frequency noise. It seems it is not a SDB2 resonance (as I erroneosly said in 37512) since it did not reduce significantly after SDB LC improvements, and moreover  it is observed to move in frequency: see spectro in Figure 5
• 30.9Hz, this seems stable in frequency and it is a good candidate for being a mechanical resonance of something suspended. There is some coherence with accelerometers placed around the IB tower. In particular the peak at 30.5Hz in the accelerometers on IB cryoTrap seems to match the left "shoulder" of the 30.9Hz peak. Figure 6.
• 1Hz comb between 40 and 70Hz: it is coherent with CEB magnetometers and UPS, good candidates are the CT heaters. Indeed the NI CT heater power modulation seems to matche the comb on/off: Figure 7.
• 62.5 Hz, this peak is stable, another candidate for being a resonance of some suspended bench, no coherence at all with ENV channels.
• 101.3Hz, this peak is stable. As for the 30.9Hz, there is coherence with accelerometers around IB, the most coherent is the accelerometer on the IB flange facing SIB2, yet there is no correspondent peak in accelerometers. Figure 8.
• 107 Hz, no coherent ENV channels. Yet the peak frequency moves. We notice one long and one short periodicity,The short periodicity  is of about 1 minute: Figure 9.  It resambles something driven by temperature.

SUMMARY

Some switch off can help sources identification (CT heaters for the 1Hz comb, remote scoll for the 24.6Hz line and possibly coherent seism noise in the 30Hz region). It could be worth a deeper search of the IB area (SIB2, SIB1, cryo-trap,...) for example with shakings and tappings.