The pdf file shows a log/log plot of the pumpdown of initial LIGO BSCs and the advanced LIGO BSC 8 in the LVEA as well
as the current one at the yend. The BSC 8 pumpdown is below the others most likely because there were several days of pumping
not accounted for before the broken feedthrough was replaced. The BSC 8 pumpdown is besides also anomolously fast. The pumpdown curve for the current advanced LIGO at the yend shows a somewhat smaller exponent n<1 P(t) ~ 1/t^n where t is the time. The areas for the yend pumpdown are:
BSC chamber = 5.2 e 5 cm^2
ISI         = 1.2 e 6
Quad        = 2.3 e 5
Transmon    = 1.6 e 5

there are two BSC chambers being pumped together one having the advanced LIGO components and the other the initial LIGO
components. The total area is 2.7 e 6 cm^2. Using the empirical Dayton relation for water outgassing, the outgassing rate
varies as J(300K) = 1.7e-7/t(hours) torr liters/sec/cm^2/hour. The predicted water vapor flow is then 4.6 e-1/t(hrs) torr liters/sec/hour.
Assuming 1500 liters/sec for the turbo pump speed, the pressure of water after 10 days is estimated as 1.3 e -6 torr. Not a hugh
amount different than an extrapolation of the data would indicate.

In summary, the slow pumpdown is unfortunately to be expected had we only put the known numbers into a calculation before hand.
The question now becomes how to best deal with this.