I agree with esshup about testing a sample.

You can even try a preliminary test with just a glass bottle. Put a sample of murky water in a dark closet for three days. If it is still murky, then you definitely have a chemistry problem! That problem requires a true sample test so you can have data to apply the correct response.