Reverse Osmosis (RO) does not offer the guarantee of 100% removal of any single contaminant. In onsite water analysis, it may indeed be found that a contaminant found in the feed water is not detectable in the water produced by the RO. However, this result may vary with each successive analysis. The main takeaway from reading a membrane spec sheet is that RO rejects a percentage of contaminants and that percentage varies with different contaminants and conditions. In addition, different membranes boast different rejection characteristics. At the end of this response, I have included a specification sheet for the Dow Filmtec® membrane, This is the product by which all other membranes are judged.
To digest this voluminous information down to a manageable document. From a technical standpoint, an RO membrane will reject material that is larger than 100 Molecular Weight units. You can pick up a chemistry book to get some idea of what contaminants fall within this size range.
In general, RO membranes will remove 95%-99% of inorganic material. This refers to things like Calcium, Strontium, Copper, and compounds like Sodium Chloride. Organic material such as bacteria and viruses are removed at 99.9% at maximum effectiveness. At times, the water produced may be totally bacteria and virus free.
While the RO membrane will remove chlorine compounds, chlorine will hydrolyze and destroy it. The rate of destruction depends on how much chlorine is in the water. It is suggested to pretreat an RO membrane with Activated Carbon to remove chlorine, a water softener to remove hardness which will foul the membrane and filtration for sediment which will plug the membrane.
One Caveat regarding the removal of bacteria and viruses. An RO membrane can be operating well within its specifications but still cause contamination. Here is what happens. Should one or two fibers in the membrane rupture it won’t affect the overall rejection percentage, however, because bacteria and viruses are living, multiplying organisms, even an undetectable reduction in measurable rejection can cause a proliferation of infectious growth.