Murray Kelly
Member
Ian;
Here is an example. We were trying to determine the optimum concentration of an additive to a developer and found that 1.5 g/l gave the effect we wanted and 1.4 did not. We ran 1.6, 1.7, 1.8 and 1.9 and found that 1.5, 1.6, 1.7 and 1.8 were pretty close giving just a hair more activity but at 1.9 there was some fog.
We used 1.6 g/l. Why? Well, 1.5 was the low threshold, and 1.7 and 1.8 were just going to waste money by adding a line width of activity in the characteristic curve while 1.9 was overkill. The rule of thumb became, "look for the onset of activity and go just one step beyond". That seems to work in general production to keep values centered around usable weights. Of course, in emulsion making things are even more precise most of the time.
PE
Also is a 7% bulk manufacture error buffer

Thanks for the tale
Murray