It's already there - it doesn't break down completely in the digestive tract and some of it gets flushed out with the urine, which, well, sooner or later ends up in the water supply.
There is an obvious stopping point provided by the concentrations that naturally occur in water drunk by large populations. If chemical X occurs naturally at varying concentrations one can interpret poor health in areas of low concentration as suggesting that it is an essential mineral and that a little should be added where the naturally occuring concentration is low.
Conversely, if high concentrations are linked with poor health that suggests it is toxic and water treatment in areas with high concentrations should seek to remove some of it.
If however the concentration in question does not already occur naturally in water drunk by large populations, you lack the vital data on the consequences of low-level, life time exposure. You are basically stuck and should leave well alone.
