I was looking at specifications for a project earlier this year that had very tight targets for iron and manganese (combined target of >0.06 mg/L) but there was no pilot data to back up that these targets were achievable. Bidders were required to guarantee these limits would be met by the specified pressure filter system with Greensand Plus media. When there was pushback in guaranteeing the performance of the system (that the engineer had designed) without any prior testing to show this was achievable, the response from the engineer was that there were other Greensand Plus filters in the State that were achieving these levels so there should be no issues making the performance guarantee…
This was essentially saying that the performance of a treatment process on one water source should be expected on a totally different source water without considering the water quality of the two sources. Anyone who knows anything about water treatment knows while iron is relatively easy to remove, manganese is a totally different animal. While manganese may be easy to remove on one water source, it could be very difficult to remove on another.
- Oxidation of Fe and Mn: Manganese Dioxide (manganese in the oxidized form) forms a far finer floc (than ferric hydroxide), so fine at times that a granular media filter will not remove it.
- Organic Complexing of Fe and Mn: Operators experiencing difficulty in removing Fe and Mn (especially Mn) have uncovered some common factors:
o A
level of organic carbon (TOC) over 2 mg/L
o Some level of ammonia or hydrogen sulfide in the feed water
- Adsorption Removal Methods (my summary): Manganese is best removed by adsorption on a manganese dioxide media like Greensand Plus or Pyrolusite. Iron is best removed by precipitation/filtration because iron adsorption blinds the media. Therefore, when both are present, the best process used is a combination of iron oxidation/filtration and manganese adsorption.
Further to the last point, oxidation of manganese takes a much longer time than iron where you are looking at from seconds to a minute for iron (Chapter 5, Chlorination) and up to up to 30 minutes or longer for manganese. Therefore in a filter system where an oxidant is dosed in the feed piping to the filters you can have adequate time in the piping and space in the filters above the media for iron oxidation (a few minutes) while allowing adsorption removal of the manganese.
There was also a very good article in AWWA’s Opflow in December 2021 titled “Evaluate and Optimize Manganese Treatment”. This article explains that the form and levels of manganese can vary considerably between wells. All manganese removal methods described in this article are based on sorption to the filter media. Therefore conditions must be optimized for the sorption mechanism on manganese oxide coated media including ensuring there is a free oxidant residual to provide a continuously regenerated adsorptive surface. The pH also impacts the Mn reaction kinetics with pHs above 7.0 more favorable.
As mentioned above, TOC, ammonia and H2S create a chlorine demand which impacts the chlorine available to oxidize the Fe and regenerate the filter media for Mn adsorption. Because the iron oxidation reaction is a lot faster compared to TOC, you typically get iron oxidation in the presence of TOC, as long as the iron is not organically bound to the TOC. When ammonia and H2S are present you may need a higher chlorine dose to overcome the demand from these compounds and provide sufficient iron oxidation. While you are not trying to remove manganese by oxidation/filtration, you still need a free oxidant residual to keep the manganese dioxide media regenerated so that is adsorbs the manganese. Therefore, a water with a high chlorine demand can impact the ability of the media to adsorb manganese. If ammonia is present, potassium permanganate may be a good option rather than chlorine as the oxidant because it does not react with the ammonia.
If iron and/or manganese is complexed with organics, the oxidation process can be significantly impacted. At a minimum, a higher oxidation dose and longer oxidation time will be required and if this works you could still create another problem with the formation of disinfection byproducts. Coagulation may be a better option to remove organically bound manganese and possibly iron also.
So clearly, iron and particularly manganese removal chemistry is not simple, and you can’t assume if the Fe and Mn levels on one water source are similar to another water source that a particular treatment technology will work equally on both. Other constituents in the water source impact removal performance and must be taken into consideration and ideally bench and/or pilot testing should be conducted to confirm the effectiveness of a proposed treatment process. To steal a saying from an old Mobile oil commercial, Manganese Removal Ain’t Manganese Removal!
The comments and opinions in this post are my own and not those of my employer.
