If you look at the way a flare connection is assembled, it's design
allows for a vacuum to
really suck the tubing flare down on the cone.
That makes for a very tight seal.
While testing with my test line-set, during my repair project,
http://ecorenovator.org/forum/geothe...html#post16037
I vacuumed out the line-set many times (with both Sanyo valves closed) and removed the test line-set later.
A few times, when I loosened the 5/8" flare nut, there was no hiss of air rushing in.
I had to
pull the 5/8" tube off the cone, before the air rushed in..
That tells me that
long duration vacuum testing on a line set that can get down under 200 microns is not very useful.
IMHO, If you can get down to <200 and maintain that with the pump running, 10 or 20 minutes (during warm weather)
should remove all the moisture from the line set.
So, looking at the flare connection design with 400 PSI of nitrogen
inside the line set,
then you need
tight flare nuts! That nut becomes very important.
The lack of any small defect on the mating face of the flare or cone become very important.
The nitrogen pressure test is more important! It simulates real world conditions. Pressure
inside the line-set.
My only small worry is putting too much pressure in the line set, which might possibly
force nitrogen into one of the closed valves and into the refrigerant.