I am assuming your question is specific to testing at each stage of the water treatment plant
Water sampling for pre-treatment should be carried out during phase one validation testing.
The intent of the test from a quality and validation point of view is two fold:
- To determine the functionality of the step is correct
- To verify that the step does not introduce new contaminants to the stage.
Point of use testing is to be carried out as per the last comment but I think you are asking specifically about the pretreatment steps, and there are requirements for this too.
Looking at your treatment plant…
For purified water: TVC, TOC and absense of objectionable organisms are the main criteria. Conductivity and TOC will vary at the different treatment points especially around water softening steps as the water contents will vary. For WFI you should add endotoxin testing. see examples below…
For example: sample before and after a storage tank to test for TVC to verify there is no significant increase. do not put figures on the acceptance critiera. vent filters can be a common source of contamination if not designed or maintained properly
Another example: Water softening tends to replace Calcium ions in the water with sodium ions in the water and will change your conductivity (it actually raises it if io remember correctly) the purpose of RO downstream from this is to then remove this extra salt content. so I hope you can see the dangers of putting specific criteria on the individual steps. RO unit manufacturers will claim that the water is reduced to about 3% conductivity of the incoming water… but again this is not enough to put an acceptance critiera on.
Remember you need to verify that the treatment steps you carry out on the water do not ‘dirty’ the water further, so you should test that each step does not ‘dirty’ the water any more (TVC endotoxin). also focus on the last few steps in your treatment process along with any tanks rather than the initial steps.
I hope this helps