Acceptable Large Leaks expressed as a percentage?

Normally around here when folks talk about large leaks and what is and isn't acceptable, the talks quickly goes to liters/minute. I know that's how the chart displays in SH, but both my machine's LED and Encore Basic (which it seems just mimic my machine) show Large Leaks as a percentage- which is far simpler and easier for me to wrap my head around. Plus, I figure they wouldn't report it if it weren't meant to be meaningful.

So can anyone tell me what is a maximum acceptable large leak number for any given night, expressed as a percentage? By acceptable I guess I'd have to say that percentage right before the machine "throws it's little arms up" as Pugsy says, or maybe more to the point the percentage beyond which the other reported numbers become unreliable.

Thanks all.

It's different between machines. You don't mention the type of machine that you're talking about. It's probably your Respironics. My Resmed is much different.

Since the on screen % of time in large leak doesn't tell us just how big the large leak is there is no way to tell exactly from the on screen % of time number just how bad the leaks were and it's how bad the leaks were that determine just how effective or reliable the therapy happens to be during large leak.
There's a big difference between a large leak that is 5 L/min into large leak territory and 30 L/min into large leak territory.
In other words let's say that it shows 20% of the time in large leak..and based on 7 hours that would be 420 minutes times 20% is 84 minutes in large leak...we have no way to know just how big that big leak during those 84 minutes were unless you look at the leak line graph to see just where it went.

If large leak territory starts at 90 L/min (it varies between machine and pressure used) and the worst it ever got was 95 L/min...then the machine data reported is likely still fairly accurate but if you went to say 110 L/min for those 84 minutes...the data is likely suspect.

If you are just wanting a general number in terms of % of time in large leak that probably would be "acceptable" and without having the other data to back it up to know just how bad the leak was...I would try to keep that % of time in large leak in the single digits if it were me. But that is just me.
Now Resmed seems to go by the 25% of the time in large leak before it gives a S9 user the Mr Red Frowny face for his leak.
In other words if the time spent in large leak is less than 25% the user will get a Mr Smiley face for leak.

Respironics users aren't ever told how much time in large leak is unacceptable.
So I suppose you could use the ResMed 25% if you wish...I think that is way too much to be acceptable IMHO and me personally...I would want it less than 10 % of the time.