Background
Arsenic contamination in drinking water presents a health problem in many areas of the world. Although the World Health Organization established a goal of less than 10 ppb for arsenic in drinking water in 1992, worldwide adoption of similar limits or treatment goals has only taken place in recent years. The 1998 European Water Quality Directive lowered the level from 50 ppb to 10 ppb, and the same reduction was set into motion by the US Environmental Protection Agency in 2001. Enforcement of the new rule in the US began in 2006, and many countries throughout the globe are now adopting similar maximum contaminant levels for arsenic. This has led to increased interest in developing more effective treatments for arsenic removal. This article from Dow Water & Process Solutions reviews the various options available and discusses their performance with respect to the water chemistry of the feed.
Technologies for removing arsenic
For large municipal applications, traditional coagulation using ferric chloride (FeCl3) followed by sand or multi-media filtration is often employed. This system generally has low operation costs and is also effective in removing iron, which has additional aesthetic benefits. Due to the need to specify the chemical dosage for each feed water, such systems require careful operation or automation. In addition, the waste water and sludge generated has to be disposed and plant size can be very large.
Several common point-of-use or point-of-entry options are available for treating arsenic, offering more simplified treatment for smaller systems. Reverse osmosis (RO) membranes are a proven technology with the advantage of improving many water quality parameters simultaneously, as a point-of-use solution. However, RO is vulnerable to the presence of oxidants and produces a waste stream of rejected water. Furthermore, it is non-selective and generally does not effectively remove the arsenic, As (III) form of arsenic. RO should therefore only be considered when there are multiple problems to be addressed (e.g. high salt content) and when only arsenate As (V) is present.
Strong base anion exchange resin can be a cost effective point-of-entry means of reducing arsenic levels, but the lack of specificity can cause the resin to require frequent regeneration if competing ions such as phosphates or sulfates are present. If the resin is not regenerated appropriately, there is a risk of sloughing arsenic if the resin column is over-run. Ion exchange resins are also not effective in removing As (III).
A selective adsorptive media is often the best approach if arsenic removal is the only consideration. A simple static bed of adsorptive media can effectively remove arsenic to non-detectable levels and may operate for several years with a minimum waste stream. In cases where the water is relatively free from particulate matter, the bed may not even need backwashing or other service during its useful life.
The most common media currently available are based on activated aluminium, iron and titanium oxides. These offer a good point-of-use or point-of-entry option for small to medium size plants. The effectiveness of a system in removing arsenic varies widely with the type of media used and the chemistry of the water to be treated. It may be difficult to predict media performance if the feed water composition changes and there could be issues related to media disposal.
Adsorptive media may be of the regenerable or non-regenerable type. Iron-impregnated ion exchange resin is a typical regenerable media which is normally regenerated off-site, requiring additional media handling and tracking, and producing a high arsenic content secondary waste stream that has to be further treated. The cost effectiveness of such a system depends highly on the efficiency of the regeneration for multiple cycles, which has yet to be proven in the field.
Activated alumina is the traditional cheap, non-regenerable media, but its low selectivity to arsenic and sensitivity to pH cause potential leaching of the disposed media. Alumina is therefore increasingly being replaced by more selective media such as iron oxide, which has higher capacity and is less pH sensitive. However, it is also affected by competing ions and arsenic leaching can occur at high pH. Other non-regenerable media based on lanthanum and zirconium oxides claim to have high capacity, but there is little field data available. As a result, accurate costing of such systems has not yet been made, although it may be high due to the high cost of such media.
Titanium oxide and hydroxide based media have higher affinity to arsenic than iron and alumina based media, as well as a superior capacity over a wide range of pH conditions and water compositions. The high affinity means lower sensitivity to competing ions such as nitrate, phosphate or sulfate and higher stability to arsenic leach both in use and when disposed. Titanium-based media has fast kinetics, which allows more flexibility in plant design, potentially less media and thereby a reduction in the size of the plant.
Factors affecting media performance
The lifetime of an adsorptive bed inversely depends on the feed arsenic concentration, so the relatively low capacity of activated alumina means that systems will become large or require frequent media change out if arsenic levels are high. It is also important to consider the speciation of arsenic, since they are adsorbed with different efficiencies. The arsenite, As (III) species is more difficult to remove, as it has no charge in neutral pH water. Some adsorbents will not remove it at all and even the most selective ones, such as titanium oxide, remove arsenite with only about half the efficiency of arsenate in typical water. If the water source is high in arsenite, a pre-oxidation step may be necessary.
The influence of feed water pH on media performance is also associated with speciation. For the arsenate, As (V) species, as the pH increases above 7, the H2AsO4- anion converts to HAsO42-, which is less effectively adsorbed from solution by selective media. Furthermore, the charge state on the surface of the adsorbent media itself is pH dependent. Titanium oxide has a positive surface charge below pH 5 and a negative charge at higher pH. For iron oxide, this transition occurs at pH ~7. These phenomena explain the sensitivity of adsorption media to pH in the typical drinking water range of 6.5 to 8.5. All adsorptive media lose efficiency at higher pH, but the effect is less dramatic for titanium oxide since its surface charge is unchanged over the common pH range found in drinking water.
Another major factor influencing media performance is the presence of other competing ions, especially silica. Although the mechanism is not fully understood, it is thought that silica acts by simply blocking the media surface, rather than fouling it. This is supported by the fact that the effect is reversible and that there is a minor accumulation of silica on the media over time. Silica levels > 20 ppm dramatically reduce the arsenic capacity of all media, although more selective adsorbents, such as titanium oxide, are affected less.
In addition to silica, a host of other interferants can impact performance, again dependent on the media in use. High concentrations of iron can complicate arsenic removal, since the arsenic present may be associated with the iron in colloidal forms. For this reason, iron removal may be required prior to arsenic removal to maximize system life. Sulfate and phosphate can interfere with non-specific adsorbents such as anion exchange resins and have varying impact relating to the degree of selectivity for more specific adsorbents (activated alumina > iron based > titanium based). Vanadium, selenium, and other oxyanions can adsorb by similar mechanisms to arsenic, so they can also impact media performance, again varying with the particular media in use.
Conclusions
Adsorptive media can provide a very cost effective means of removing arsenic to non detectable levels without significantly altering other water quality characteristics. Compared with other available technologies such as coagulation-filtration, reverse osmosis or ion exchange, selective adsorptive media offer simple operation and easily managed waste disposal, particularly for small or medium sized systems in developing as well as industrialized regions. Tests show that titanium-based media has performance advantages, especially in selectivity to arsenic, resulting in high capacity, low sensitivity to other anionic components and pH tolerance.
It is critical to have a good knowledge of the water chemistry of the water to be treated in order to select an appropriate point-of-entry or point-of use media. This will allow for effective system design, operating conditions and allow realistic performance expectations to be set. Effects such as arsenic speciation, level of silica and water pH can be even more important in determining the lifetime of a media bed than the arsenic level itself. Any model for performance provided by a media supplier should incorporate these factors as a minimum. Dow Water Solutions has developed a semi-empirical model that quantifies the impact of different parameters on media performance and allows a realistic estimate for system designs to be made. It can also be used to optimize the plant configuration and operation.