Arsenic & Iron Removal for Public Water Systems

Technology considerations for effective co-contaminant treatment

It has been estimated in a national arsenic occurrence survey of drinking water supplies that as many as 35 to 40% of U.S. public water systems have arsenic above the maximum contaminant level (MCL) of 10 parts per billion (ppb), and have co-occurrence of iron above the secondary MCL of 0.3 mg/L. This presents an interesting challenge for small systems, which may or may not have any treatment in place for either iron or arsenic to meet their respective MCLs. Although there is a widespread distribution throughout the U.S., there is a particularly high co-occurrence in parts of the Midwest, where incidences of both iron and arsenic found above their respective MCLs are closer to 75%. These incidences have occurred in Michigan, Wisconsin, Minnesota, Ohio, Indiana, North Dakota and Illinois.

Until recently, such systems were only concerned with providing treatment to achieve the secondary (non-enforceable) aesthetic aspects of having high iron or manganese in their water—red water staining, black deposits or other accumulation of precipitants in distribution and piping problems that result in customer complaints. With the new arsenic MCL in play, this has clearly changed.

So what is the small water system to do? Which technology is most appropriate to manage both contaminants to meet the MCLs? If one removes iron, will it guarantee arsenic removal? Is there a silver bullet? How does one make practical and cost-effective technology decisions without spending a fortune? There are no simple answers to these questions. The prudent treatment provider needs to assess the current water chemistry and existing treatment (if present). The site-specific factors at each well site also must be addressed. Unfortunately, there are many design variables to consider. Many of these have been well documented or reported previously in work by the U.S. Environmental Protection Agency’s (EPA) Office of Research and Development and others. Some of these variables are mentioned below, along with some practical technology suggestions based on previous installations and experience.

Common Misconceptions

It is impossible to cover all design variables in this article, but it is safe to say that the process of iron and arsenic removal is more complicated than it seems. Without proper understanding of the behavior and interactions of iron, manganese and arsenic, unmet performance expectations and field problems are inevitable, not to mention frustrated system operators. In some cases, the wrong technology will be employed, resulting in wasted resources and non-compliance. Let’s identify a few common misconceptions.

  • If one uses a technology that removes iron and manganese in the water, it will guarantee that the arsenic problem will go away as well.
  • If the ratio (concentration) of iron to arsenic is greater than 20:1, that is all that is necessary for an iron removal process to concurrently achieve the new arsenic MCL of 10 µg/L.
  • If one adds enough iron to the water (i.e. coagulant dose), the system can achieve whatever removal efficiency of arsenic desired.
  • Adsorption media designed for arsenic removal can accommodate high con- centrations of iron without negative consequences; the adsorption vessels simply can be backwashed more fre- quently without loss of performance.

Field experience, recent EPA arsenic demonstration projects, technical papers and past case studies have brought new light to these common misconceptions in the marketplace. Let’s begin with an overview of the most common methods that deal with these contaminants individually, followed by a discussion of design considerations and some actual field applications.

Common Treatment Options

To properly evaluate alternatives for co-contaminant removal, we must first review individual contaminant technology options and then look for common denominators. Although the list is more extensive, the most common arsenic removal technologies (or technology categories) being employed today for small public water systems (less than 1,000 connections) include the following:

  • Adsorption (iron-based FeOOH media or other metal oxide media)
  • Oxidation/filtration (media types such as manganese greensand, AD26)
  • Anion exchange
  • Coagulation/filtration (often using iron salts such as FeCl3 or FeSO4)

For even smaller systems regulated under the arsenic rule, including non- transient, non-community systems such as schools, daycare centers and small subdivisions, this list becomes even shorter. These very small public water systems are selecting adsorption alternatives predominantly because of simplicity, economics, the need for less operator attention, available space and operating costs.

For iron and manganese removal, the objective is primarily to convert dissolved or reduced species to oxidized forms of these constituents to form Fe(OH)3 and MnO2 precipitants, which are filterable in subsequent unit operations. The most common methods for small systems include aeration/filtration, cation exchange (softening), manganese greensand, MnO2 media processes such as pyrolusite, AD26 and sequestering agents. Figure 1 summarizes some of these technologies with comments.

It is important to point out that what may be considered a proven technology for iron and manganese removal may not be appropriate for high efficiency arsenic removal. A case in point is manganese greensand. Glauconitic-based materials used with potassium permanganate and/or chlorine addition (continuous or intermittent) in conventional pressure filters have been used effectively for iron and manganese removal for decades. These technologies, however, have not been designed or employed historically for arsenic removal. In situations where arsenic is just above the 10 µg/L standard in the presence of high iron, this technology could be appropriate if applied properly. It is important to understand the inefficiencies of conventional filtration processes for arsenic removal to avoid misapplication.

Another recent iron/arsenic example is a 250-gal per minute (gpm) EPA demonstration site in Minnesota. The site was designed for a two-stage treatment train with a conventional iron removal plant followed by an adsorption system using granular ferric oxide. The incoming iron averages 1.1 mg/L, manganese at 0.04 mg/L and arsenic at 39 to 50 µg/L, 35% of which is As (III). The arsenic-to-iron ratio is approximately 25:1. The system utilizes a proven aeration and conventional filtration approach with 20 minutes of contact time prior to arsenic polishing. While the iron is being removed consistently in the pretreatment unit to <0.1 mg/L, the arsenic removal efficiency is only 33 to 45%, remaining still well above the 10 µg/L MCL when entering the polishing adsorption system.

Arsenic Removal and Naturally Occurring Iron—Friend or Foe?

Arsenic in its oxidized form, As (V), has a natural affinity for iron. In normal pH ranges, arsenate binds with ferric iron, Fe(OH)3, in natural waters. This is of course why iron-based media, such as granular ferric oxide (GFO), works so well for arsenic adsorption. The media carries a net positive charge below its zero point of charge (ZPC), while As (V) behaves as a negatively charged anion (H2AsO4-). The capacity of these synthetic, high surface area, high iron content media (over a wide range of pH, from 5.5 to 8.5) is well demonstrated.

When the naturally occurring iron or manganese concentrations are significantly higher than secondary MCL levels, however, a careful selection of technology is needed. Adsorption media processes, anion exchange and reverse osmosis are best employed in water profiles where iron and manganese are present in concentrations at or below 0.3 and 0.05 mg/L, respectively. While some iron may actually help arsenic removal in adsorption processes (in the short run), the more common consequences are as follows:

  • Binding to the media, resulting in fouling of the media surface and/or blocking the pore structure (where the majority of adsorption occurs);
  • Accumulation of precipitants in the adsorption beds, resulting in high pressure drop across the system; and
  • Reduced capacity for arsenic adsorption (which is the primary intent rather than iron or manganese removal), resulting in more frequent media change-outs and higher operating costs.

Technology Selection

Therefore, in scenarios where adsorption media solutions are being considered for arsenic removal in waters containing high iron or manganese levels, pretreatment to remove these interferents may be essential for good performance. Pretreatment would likely include one of the technologies mentioned in Figure 1. Although an effective approach, the downside is that this means a two-stage treatment process would be used. Could both contaminants be removed effectively in a single treatment step? The answer is yes—sometimes. The benefits can be attractive, with capital savings of potentially 30 to 50%.

Given the right combination of water quality conditions and effective use of the naturally occurring iron in the water, one can remove both contaminants with a single technology. To do this, however, one would not employ an engineered adsorption media product as a stand-alone technology. Figure 2 is a technology matrix for co-occurrence of iron and arsenic and relationships between the two for technology selection.

What is the right combination of water conditions and which technologies are suggested? Based on experience, the most important variables for effective co-occurring iron and arsenic removal include:

  • Conversion of As (III), if present, to 100% As (V)
  • Use of the right oxidizer to accelerate/effect removal
  • Location of the oxidant in the treatment train
  • Presence of high stoichiometric ratios of oxidized iron to arsenic, at least a 30:1 ratio
  • Sufficient contact time and intimate association of oxidized iron and arsenic to form a precipitant
  • pH in the normal ranges of 6.5 to 8.5
  • Surface-active and effective filtration media and backwashing to remove the precipitants

Figure 2 plots iron and arsenic concentrations and shows a partitioning of various technology options. Notwithstanding other potential interferents, low iron waters may be ideally suited for adsorption or other options without any pretreatment. Although adding iron salts to the water is often less desirable for small systems from a residuals management and operations standpoint, low iron water with arsenic may be a suitable candidate for coagulation-enhanced filtration processes.

The green shaded area in Figure 2 depicts scenarios with high iron and moderate to low arsenic (below 25 µg/L). Generally speaking, the higher the iron-to-arsenic ratio, the better chance for success under the right treatment conditions. As pointed out earlier, the efficiencies of conventional iron removal processes (aeration/filtration, greensand, etc.) differ from site to site. While many can achieve excellent iron and manganese removal (>85%), efficiencies for arsenic removal may be far less—35 to 50%. Therefore, caution is advised to avoid expecting too much from a technology.

AD26 Applications for Co-Contaminant Removal of Iron, Manganese and Arsenic

A technology that has been employed in multiple applications for co-contaminant removal is AD26, a MnO2 media-based process. It is utilized in conventional pressure filtration vessels using chlorine to oxidize As (III) to As (V), enhance oxidation of naturally occurring iron (Fe+2 to Fe+3) and provide optimal conditions for ferric arsenate complexes to form on the surface of the catalytically active media. Similar to other iron removal processes, the filters are backwashed periodically to remove the accumulated iron and manganese precipitants. The solids consist of oxidized insoluble, non-hazardous particles.

AD26 is in the same family of technology as pyrolusite, a high-content MnO2 media with a successful history of use for iron and manganese removal, particularly in the western U.S. In contrast to pyrolusite, the AD26 is attrition resistant, does not leach manganese and has NSF 61 certification. The oxidant injection for AD26 occurs just before entering the filtration vessels, requiring only a very short contact time of 15 to 30 seconds, compared to minutes with other alternatives. Flow rates through the media bed(s) are three times that of conventional filtration with high efficiency removal, requiring smaller vessels and less space. Where adequate iron-to-arsenic ratios, oxidizing conditions, pH and other design variables are favorable, the technology has been very successful for co-contaminant removal. Figure 3 illustrates three full-scale co-contaminant arsenic and iron/manganese sites where the technology has been employed for achieving both primary and secondary MCLs.


Conventional iron and manganese removal processes alone, while designed and effective for those target contaminants, may or may not be capable of providing the treatment horsepower for effective arsenic removal to meet the 10 µg/L standard. Could an existing iron removal plant be optimized or modified to achieve arsenic compliance? Sure, but only under the right conditions. What are the tradeoffs? How efficient is the process currently? A careful understanding of the water chemistry and process limitations of each technology must guide the decision-making process. It is the author’s experience that more often than not, overly simplistic conclusions such as, “If I just remove the iron, my arsenic problem will disappear” are being proposed without the proper due diligence and data. The good news is that there is hope for co-contaminant sites, and finding a cost-effective solution may be closer than believed.



Hello. This is a great article. I wondered where the figures 1 and 2 can be viewed? Are they on the links or somewhere else.

many thanks, [email protected]

Post new comment

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.

More information about formatting options


About the author

Gregory C. Gilles is vice president and principal for AdEdge Technologies, Inc. He can be reached at 678.835.0052, or by e-mail at [email protected].