hydrate software

I have calculated several hydrate prediction curves in Excel with different amounts of Methanol with the rigorous model included in Prode Properties,
as far as I can observe the model predicts values in reasonable agreement with experimental data (from Sloan's book) for fractions of methanol below 20-30% ,
in my tests Hammerschmidt's equation gives equivalent (or higher) errors,
my question : which is (in engineer's good practice) considered the maximum amount of methanol to add into a stream ?
My client says they add up to 30% weight fraction, do you know of higher amounts ?

Pat,
asanter has replied in this thread

'http://www.egpet.net/vb/threads/27092-Excel-library-for-process-calc-s-including-distillation/page19'

hoping this helps,
Carlo