This paper analyzes and compares optimization results for simulated di
abatic columns Co theoretically derived optimization results from the
isoforce method for the same system. The separation of ethanol and wat
er is used to illustrate the two methods and bring out their common an
d special features. The results, which are presented in McCabe-Thiele
diagrams, show partial compatibility between the methods. The isoforce
method is found to require total reflux ratios close to the azeotrope
, and a reformulation of the method is suggested. It is also shown tha
t the separation work obtained in an adiabatic stripping column can no
t be achieved in a diabatic column without adding a (distributed) cond
enser in the upper section of the column, or more plates. (C) 1997 Els
evier Science Ltd.