Initially we were unable to do some needed leak testing on GV6 a few weeks ago as the helium background signal in the site vacuum system was too high (~2 x 10-6 torr*L/sec).  We have reduced this via valving-in the YBM turbo during the day then valving it out at the end of the day over the past few weeks (must be attended when open to the BT).  As of the end of today the helium background is about 2.5 x 10-7 torr*L/sec, still too high for leak testing.  

Based upon the physical parameters, the rate of removal via the YBM turbo pump suggests a reservoir of helium with some independent conductance to the vacuum system as opposed to an initial quantity of helium fully present in the vacuum system.  As such we have vented then pumped and or purged adjacent volumes which had been exposed to helium spraying in the recent past (GV1 annulus, GV3,4 gate annulus, HAM4 annulus, HAM4-5 OMC volume, HAM1-2 annulus, HAM1 interior.  This had no significant effect, i.e. the reservoir+conductance could be withing the vacuum envelope(?)

Today we were able to "take the gloves off" (HIFO Y ended) and demonstrate the the large, 2500L/sec, ion pumps are the source of helium.  We conclude this by noting that soft-closing GV5 resulted in a slight increase in the helium background at the Vertex (IP9 and IP11 not yet saturated-still pump helium a little bit) and that valving-out IP1, IP2, IP5 and IP6 resulted in the plummeting of the helium signal, i.e. t0=1100hours=3.1x10-7 torr*L/sec, 1110hours=1.7x10-7, 1120hours=9x10-8, 1130hours=5.0x10-8, 1140hours=2.7x10-8, 1150hours=1.5x10-8, 1305hours=1.2x10-9, 1430hours=1.1x10-9.  

Conclusion: We believe that this is the first instance that we have ever had a leak detector valved-in while one or more 2500L/s ion pump(s) was simultaneously valved-in.  Nominally, leak testing is performed on an isolatable volume pumped only by a turbo which is backed by a leak detector.  When initially attempting to leak test GV6, the YBM turbo, IP1, IP2, IP9 and IP11 were all valved-in (GV6 was open at the time).  Therefore we don't know if the helium concentrated/dissolved in the ion pumps that we see now is the result of years of low level residual exposure or, conversely, one single large recent exposure "event".  So we don't know if this is a new issue or an old issue(?)