Ideal countercurrent recycle cascades are characterized by two propert
ies: 1) The compositions of heads and tails streams forming the feed s
tream to individual stages are the same, and 2) the heads separation f
actor for each stage is constant. When these two criteria are met, the
heads and tails separation factors are constant and equal to the squa
re root of the stage separation factor. The mixing of streams with dif
ferent compositons within a separation cascade obviously constitutes a
n inefficiency since it is precisely the reverse of this process that
is desired, hence Condition 1, which is often referred to as the no-mi
x criterion. Separation cascades for which both criteria are valid are
termed ideal. However, the ramifications of Condition 2 are not obvio
us and it is possible to design no-mix cascades which do not meet the
second condition. Mathematical relationships between ideal and no-mix
separation cascades are derived to quantify these differences. It is s
hown that Condition 2 minimizes the total interstage flow required to
make a given separation for any no-mix cascade design, i.e., it is an
ideal cascade by definition. Finally, the required number of ideal sta
ges necessary to perform a specific separation with no-mix cascades ar
e evaluated and compared to those of the ideal cascade.