[ilhami]

Hi all,
in my drinking water plant we have a reverse osmozis membran unit and then a tank to blend it with raw water, then this mixture is exposed to ozone approximitely at 0,20 ppm. Raw water is 500 μS/cm, purified water is 20 μS/cm, and the final mixture is 50 μS/cm.
Last week i had a lot of complaints about yellow coloured water in bottles, exactly in the same minute. One minute later the bottles are fine. They were filled just in the starting time of the line.
First i thought they may be green algae but after the microbiological analysis, i found out they weren't.
I measured conductivity and pH of both yellow and clear ones, the result was interesting;
Clear water:..   7,32 pH - 35,1 μS/cm (27,6°C) (blending unit was not operating at that moment)
Yellow water:...10,07 pH - 46,5 μS/cm (28,0°C)
Filtered yellow water: 9,76 pH - 49,2 μS/cm (26,1°C)
When i filtered it through a 0,45 μm membran filter i found a yellowish green residue with fine grain size.
And in the bottle they were distrubited in water in a colloidal state - no sedimentation.
I have added every chemical material that is used in the plant in our water to reach this pH in this conductivity value, but conductivity was always very high at 10 pH.

I am totally confused, what compound it may be to raise the pH to 10 from 7,3 while giving the conductivity a slight increase

And how can it be so sudden while we have a 10 ton buffer tank, but out ozone tank is really small.

Does anyone has any ideas, or got through such an incident

[Borek]

No idea about exact numbers, but they should be not difficult to estimate - increasing pH means going from about 10^-7 M OH^- to 10^-4 M OH^- - by how much this ion alone will increase conductivity? (there must be a counterion as well, but it will have much lower mobility and its effect will be much lower as well)

[Arkcon]

This is a pretty interesting analytical problem.  I'm having a little bit of trouble understanding the engineering of your water bottling procedures.  That's out of my depth in that regard.

Its great that you have a filtered sample of the colloidal solid.  You may want to submit that for elemental analysis.  Find out what metals are there, to understand what went wrong.  While you're waiting for those results, you can review the construction of your equipment with process engineers -- is maintenance up to date?  Are there any jury-rigged connections i.e. welded seams that haven't been sanitized afterward.

Do you have a good understanding of your input water?  Has that ever been submitted for elemental analysis?  Levels will of course be much lower, but you'll want to have that handy, maybe something is at a very high level.

What are your cleaning validation protocols like?  Speaking in the most general terms -- what are the results of elemental analysis from a sample of "dirty" equipment?  What about samples when you've begun to clean?  What samples result from how you define "clean" equipment.

I'm going to suspect that you're going to tell us that you don't have this information, and that you can't take the line out for service for these sorts of procedures.  That's OK.  But do try to tells us what cleaning procedures are required for your municipality, and how you comply with those, and what the results are.  If you follow those procedures diligently, problems such as your are easier to detect, prevent and fix, and that saves a company more money.

Now, this part:

Clear water:..   7,32 pH - 35,1 μS/cm (27,6°C) (blending unit was not operating at that moment)
Yellow water:...10,07 pH - 46,5 μS/cm (28,0°C)
Filtered yellow water: 9,76 pH - 49,2 μS/cm (26,1°C)

OK, this makes me feel a little ... itchy.  You filtered it, and the pH changed, and quite significantly.  But pH only measures activity of a solution -- filterable solids, colloidal suspensions, those don't matter.  Now, if you vacuum filtered, you may have removed a volatile component -- maybe even dissolved CO₂.  Or some oxygen may have changed something.  But you should take multiple samples of each, and apply statistics to them.  Speaking broadly, if you take multiple clear water samples, and the pH varies from say 6.9 to 7.4, then your variation from 9.76 to 10.07 for two water samples doesn't mean much.