Based on my experience, I would have to agree with Don's basic ideas.  
When I was responsible for such operations, I used a conductivity 
meter with the probe in the same cell as where the water was 
introduced in the counterflow series of cells.  I set the meter so 
that it turned on at twice the conductivity of the incoming water.  
This was an arbitrary starting point; I am sure that with further 
testing, we could have lowered the setpoint.  And this was with 
softened city water.  DI and RO water have their uses, but in 
an electroless line, for example, they sound like overkill.  If 
possible, a hang time of 10 seconds over a process bath maximizes the 
drainage in a reasonable amount of time; minimizing water usage means 
minimizing the need to rinse.  As Don implied, making theoretical 
calculations is a lost cause since variable interactions are too 
costly to figure out.  An empirical approach is most cost effective: 
what works for your specific situation will be determined by trial and 
error.  Rinsing may also be most effective with warmer water vs. 
cold water and bubbling vs. no bubbling or other agitation to 
maximize mixing.  And don't forget simpler things like: no 
uncontrolled hoses, automatic shut-offs, spray rinses, restrictors, 
etc.
I once heard someplace:  What is Clean?  Better is the evil of 
good enough.

Robert Lundquist
Assistant Director -- MnTAP 
Suite 207
1313 Fifth St. SE
Minneapolis, MN  55414
612-627-4557
612-627-4769 FAX

[log in to unmask]