Chlorine will irreversibly damage a polyamide RO membrane, and as such, a chlorine residual cannot be allowed to contact the membrane. This means that once you dechlorinate, surviving bacteria will once again flourish, and dead bacteria will be available as assimilable carbon sources for the survivors.   Many components of existing TOC will also be broken down to AOC (Assimilable organic carbon). These will also become a source of carbon for increases in biological populations and a resulting biomass. This is especially a problem with water that has very high organics content.   Research has shown that fouling by dead bacteria can result in flux decline almost as severe as what would be observed with dead bacteria, due to a phenomenon know as “cake enhanced osmotic pressure”.

The use of dechlorination chemicals such as sodium bisulfite (SBS) is also tricky because it encourages growth of sulfate oxidizing bacteria that produce a very dense slime.  This type of slime is difficult to penetrate even with aggressive high pH CIP chemicals. More importantly, bisulfite requires extended contact time to react with chlorine but most plants dose immediately upstream of the cartridge filters; this results in halogenation of the RO membrane, resulting in an irreversible loss of salt rejection.  Almost all autopsies performed on membranes from plants using chlorination followed by SBS dechlorination find a combination of lost salt rejection and dense biofouling.

Carbon filters are not much better, as the media in the top layers removes the chlorine while the media in the lower layers with adsorbed organics provide an AOC supply to the surviving bacteria. Studies have shown higher bacterial counts post carbon filter than pre-chlorination.

Many wastewater reuse RO systems inject chlorine into their feed water which already contains high levels of ammonia. This produces chloramines which are safe for direct contact with the membrane up to 3 ppm as combined chlorine. This gives the advantage of maintaining a biostatic environment throughout the membrane system. However, if you do not already have ammonia in the water, this would be risky – any failure leading to a loss of ammonium sulfate dosage would result in irreversible damage to your membranes.  Additionally, recent experience at water reuse facilities applying chloramination has found that certain polyamide membranes irreversibly lose permeability when feedwater pH is decreased below 7. Further information about this phenomenon can be found in our paper titled ‘The perils of using chloramines for pretreatment of water reuse RO‘.

In most cases, it is best to avoid chlorination as it is more likely to increase biofouling issues. Slug dosing non-oxidizing biocides such as DBNPA is generally a better solution for disinfecting the membranes on a periodic basis so as to extend times between cleanings. Permeate flushing before every shut down, and/or every 12 hours can also help reduce frequency of cleaning.  There are also new membrane-compatible oxidizing biocides that can be used in feed water that doesn’t contain oxidizable components such as ferrous iron or H2S.