(see also https://alog.ligo-wa.caltech.edu/aLOG/index.php?callRep=27078) 

Today I inspected the "hang tags" on the N2 and Kr calibration gases at each of the currently installed aLIGO RGAs (Pfeiffer Prisma Plus, Y-end, X-end and Vertex).  It turns out that these are NOS (new old stock) calibration gases that were intended for but never used during the Beam Tube bake back in 1998.  Though I don't specifically remember, I must have decided to use these since they were the correct gases and leak rates needed for the RGAs and were unused, albeit old, and would save having to spend another $8,000.  This looks to have been a bad idea as the factory fill pressures for the N2 bottles were sub-atmospheric, ~500 torr.  This is typical for "crimped capillary" leaks for the mid 10-8 torr*L/sec leak rates and isn't an issue as long as their isolation valves never get opened at pressures > 500 torr, say like at atmosphere.  The problem here is that these sat in a drawer for 15 years along with various other cal-gas bottles and were accessible to anybody wanting to turn a knob.  

It looks very much like the N2 cal-gas used at the Vertex had been opened to atmosphere at some point during its long storage.  I am currently re-baking the Vertex RGA after having exposed it to the N2 bottle last Friday but while both cal-gas isolation valves are open.  It may be that the N2 bottle may need to isolated and tagged-out and we may be stuck with only a Kr leak at the Vertex.  I haven't reviewed the data but this might also explain the poor clean-up of the RGAs at the end stations.