12 Mitigation of Excess Fluoride in Groundwater

In water-scarce and remote areas, treatment techniques to remove fluoride from groundwater may be the only mitigation option available. An enormous number of materials have been investigated for their aqueous fluoride removal capabilities. Some common groundwater fluoride mitigation approaches are listed in Table 4 at the end of this section. They have been reviewed extensively elsewhere (Bhatnagar et al., 2011; Habuda-Stanic et al., 2014; Heidweiller, 1990; Jagtap et al., 2012; Mohapatra et al., 2009; Sandoval et al., 2021; Yadav et al., 2019). Some of the longest-established methods involve low-technology coagulation/precipitation or adsorption/ion exchange. The Nalgonda coagulation technique, named after the Nalgonda District, Telangana, India, where it was developed in the 1970s, has been one of the most frequently applied (Jagtap et al., 2012; Nawlakhe and Bulusu, 1989). The method uses a combination of alum (or aluminum chloride), lime (or sodium aluminate) and bleaching powder. These materials are combined with fluoride-rich water, stirred, and the aluminum hydroxide flocs with co-precipitating fluoride are then left to settle before removal by filtration. The method has been applied at domestic scale (bucket) and community scales (fill-and-draw plant) (Nawlakhe and Bulusu, 1989). Costs are moderate and raw materials usually readily available. Use of alum results in increased concentrations of SO4 and suspended particles in the treated water and so aluminum polychloride sulphate has been used as an alternative (Lagaude et al., 1992). Major drawbacks of the Nalgonda technique, besides the high sulphate concentration in treated water, are production of sludge, presence of residual aluminum and reports that fluoride removal efficiency is only around 18 to 33 percent (Yadav et al., 2019). Other coagulation methods include addition of calcium-bearing materials such as gypsum, dolomite, calcite or calcium chloride (Nath and Dutta, 2015).

Electrocoagulation has also been developed more recently and has been reviewed by Sandoval and others (2021). Metal electrodes connected to an external power supply inserted into an electrolyte solution (groundwater) produce metallic cations by oxidation at the anode (usually aluminium, Graça et al., 2019) while reduction at the cathode produces hydrogen gas and hydroxide ions. Coagulating metal flocs produced by the electrochemical reaction remove fluoride (as e.g., aluminium fluoride hydroxide) from solution and are then removed by flotation, settling and filtration (Emamjomeh et al., 2011; Sandoval et al., 2021; Yadav et al., 2019; Zhao et al., 2011). The method shows promise (Luna et al., 2018) but to date, has not been applied at a scale large enough for fluoride removal in developing-country settings.

Numerous sorbents and ion-exchange media have been tested for the removal of fluoride from water. These include activated carbon, activated alumina (Barbier and Mazounie, 1984; Bhatnagar et al., 2011), manganese-oxide-coated alumina (Tripathy and Raichur, 2008), fluorapatite (Wei et al., 2014), chitosan (Hu et al., 2018), titanium oxides, ion-exchange resins (e.g., Defluoron 1, Defluoron 2), and several types of local materials including plant carbon (Venkata Mohan et al., 2007), clay minerals (Chaturvedi et al., 1988; Du et al., 2011; Nabbou et al., 2019a), aluminium oxides (Farrah et al., 1987), iron oxides (Tang et al., 2010b), mixed Fe-Al oxides (Sujana and Anand, 2010), zeolites, calcite (Turner et al., 2005), clay pots (Moges et al., 1996), fly ash (Chaturvedi et al., 1990), local soils (Wang and Reardon, 2001; Zevenbergen et al., 1996), rice husks, crushed bone, and bone char (Brunson and Sabatini, 2009).

The pH dependence of many fluoride sorbents is well-established. Sorption to the metal oxides (e.g. amorphous Al(OH)3, gibbsite, Al2O3, activated alumina, iron oxides) is strongly pH-controlled, with specific sorption to the aluminium oxides reported to be most effective typically around pH 4 to 8 (Farrah et al., 1987; Shimelis et al., 2006; Sujana and Anand, 2011), slightly acidic dependence of activated alumina (Mohapatra et al., 2009; Yadav et al., 2019) and neutral to mildly acidic range for ferric oxide and hydroxide (around pH 3-7, Tang et al., 2009; Tang et al., 2010b). Fluoride sorption is favored electrostatically by net positive surface charges on the variably charged oxides at acidic pH. Sorption to clays is also pH-dependent (Kau et al., 1997; Mudzielwana et al., 2016). Due to the permanent negative surface charge on clays, many approaches to fluoride removal have involved modifications of clay materials to improve the anion sorption capacity (Ma et al., 2012). Desorption from clays at high pH is commonly attributed to OH exchange on octahedral layers (Wang and Reardon, 2001).

Sorption efficiency is further affected by factors such as sorbent composition, texture and aging, initial water fluoride concentration and overall chemical composition, notably presence of competing anions. Sorption to many surfaces is also reported to be endothermic (Biswas et al., 2007; Hu et al., 2018; Mejia et al., 2017; Nabbou et al., 2019a). Given high initial fluoride loadings, many of the adsorption techniques struggle to achieve fluoride concentrations below around 1 to 1.5 mg/L (Mohapatra et al., 2009), although this meets the requirements of the WHO guideline value. One of the main drawbacks of the adsorption/ion-exchange methods is the production and disposal of waste materials (Yadav et al., 2019). Activated alumina and bone materials are among the more frequently used and effective fluoride sorbents (with highest removal capacity). However, activated alumina is relatively expensive and may not be universally available, and bone products are unacceptable in some cultures.

Other removal methods include solar distillation (Antwi et al., 2011) and the membrane technologies such as electrodialysis (Gmar et al., 2015), reverse osmosis (Schneiter and Middlebrooks, 1983) and nanofiltration (Yadav et al., 2019). Electrodialysis effects passage of fluoride ions through a semi-permeable membrane via use of an electric field. Nanofiltration and reverse osmosis use a semi-permeable membrane to prevent passage of fluoride ions (and other dissolved solids) by application of pressure sufficient to reverse the natural osmotic pressure (Yadav et al., 2019). Nanofiltration involves use of slightly larger membrane pore sizes, with less resistance to the passage of solutes, lower pressure and hence lower energy needs. Much research has also gone into targeting membranes for removal of specific solutes including fluoride (Mohapatra et al., 2009). Membrane technology offers many advantages in terms of removal performance, lack of interferences and removal of solutes besides fluoride, but requires greater technical knowledge and is relatively high-cost. Membrane fouling from organic solutes, colloids, scales and biofouling accumulations can be an additional problem (Van der Bruggen et al., 2008).

Most methods designed for village-scale fluoride removal in developing-country settings have drawbacks in terms of removal efficiency, cost, local availability of materials, residual chemicals or taste in treated water, lack of monitoring of treated water and disposal of treatment chemicals. Many have not been tested beyond pilot or laboratory scale. Methods that have been tested have experienced problems with long-term sustainability. Success rates depend on factors such as fluoride removal efficiency, treatment capacity, ease of use, ease and cost of maintenance, availability of raw materials and degree of community participation and acceptance.

As examples, various pilot defluoridation schemes have been in operation in the East African Rift Valley since the 1960s. Methods have included bone char, coagulation and activated alumina (Kloos and Tekle-Haimanot, 1999). Frustrations with the efficacy and operation of the Nalgonda coagulation technique centered on inadequate removal of fluoride and production of sludge, which led to a shift towards use of bone char which has greater removal efficacy and is readily available (Dahi, 2016). In India, despite Nalgonda having been developed there, it does not appear to be in widespread use and little evidence exists for long-term use of other appropriate methods (Ganvir and Das, 2011). In Sri Lanka, reverse osmosis has been applied in some affected areas, though problems with inadequate disinfection and maintenance, scarcity of water, lack of technical capacity and brine removal have all been highlighted (Imbulana et al., 2020). In affected areas of China, applied methods have included activated alumina and electrodialysis, though piped water supplies have also been installed where feasible (Wang et al., 2012).

Given the common operational and sustainability problems of groundwater fluoride removal, potentially beneficial alternative approaches to water quality improvement include judicious borehole siting and groundwater management. Factors in borehole siting include local geology and spatial variations in groundwater fluoride concentration (e.g., with depth). Groundwater management includes consideration of optimum pumping rates, especially where there exists the possibility of mixing of groundwater with deep fluoride-rich groundwater (e.g., old groundwater or hydrothermal fluids), which could be increasingly drawn upward at high pumping rates (Carrillo-Rivera et al., 2002).

Groundwater management options potentially include application of managed aquifer recharge (MAR) schemes. MAR has long been suggested to improve groundwater quality, as well as to augment groundwater resources. MAR schemes have been particularly popular in India and have included constructed check dams (Bhagavan and Raghu, 2005; Brindha and Elango, 2011; Brindha et al., 2016), dug recharge wells (Brindha et al., 2016), percolation ponds/tanks and infiltration galleries. Some positive benefits in terms of fluoride reduction have been observed, although documentation on MAR implementation has appeared to suggest mixed outcomes for fluoride mitigation (Brindha et al., 2016) as well as for water budgets (Boisson et al., 2015). Some supply wells have shown limited changes or even increased fluoride concentrations (Bhagavan and Raghu, 2005). Raising the groundwater level can bring previously unsaturated aquifer horizons into the zone of water-level fluctuation and can result in mobilization of solutes (e.g., Hallett et al., 2015). It could also increase concentrations of fluoride and dissolved salts through evaporation. Thepotential exists for MAR schemes in fluoride mitigation, but the methodology requires careful monitoring and is likely to be site-specific. Direct rainwater harvesting through installation of surface or subsurface containers also offers prospects for collection of low-fluoride water supplies, at least seasonally during and shortly after periods of active rainfall.

Table 4 Commonly applied methods for removal of fluoride from drinking water (after Heidweiller, 1990; Jagtap et al., 2012; Mohapatra et al., 2009; Sandoval et al., 2021; Van der Bruggen et al., 2008). [View a full-width version of Table 4 on a separate web page.]

Treatment method Capacity/dose Working pH Interferences Advantages Disadvantages Relative cost
Alum (aluminium sulphate) 150 mg/mg F Non-specific Established process Sludge produced, treated water is acidic, residual Al present, may have adverse taste Medium-high
Lime 30 mg/mg F Non-specific Established process Sludge produced, treated water is alkaline, may have adverse taste Medium-high
Alum + lime (Nalgonda) 150 mg alum + 7mg lime/mg F Non-specific, optimum 6.5 Low-tech, established Sludge produced, high chemical dose, residual Al present, may have adverse taste Medium-high
Gypsum + fluorite 5 mg gypsum + < 2 mg SO4 /mg F Non-specific Simple Needs trained operators, low efficiency, high residual Ca, SO4 Low-medium
Calcium chloride 3 mg CaCl2/mg F 6.5-8.0 Simple Needs additional flocculent (e.g., FeCl3) Medium-high
Electrocoagulation High 6.0-8.0 Sulphate, phosphate, bicarbonate Few chemicals Needs electrode replacements, power; passivated film formation, potential residual Al in treated water Medium-high
Adsorption/ion exchange
Activated carbon Variable < 3 Many Large pH changes before and after treatment High
Plant carbon 300 mg F/kg 7 Locally available Requires soaking in potassium hydroxide Low-medium
Zeolites 100 mg F/kg Non-specific Poor capacity High
Defluoron 2 360 g F/m3 Non-specific Alkalinity Disposal of chemicals used in resin generation, Cl in treated water Medium
Clay pots 80 mg F/kg Non-specific Locally available Low capacity, slow Medium
Activated alumina 1200 g F/m3 5.5 Alkalinity Effective, well-established Needs trained operators, chemicals not always available Medium
Bone 900 g F/m3 > 7 Arsenic Locally available May give taste, degenerates, not universally accepted Low
Bone char 1000 g F/m3 > 7 Arsenic Locally available, high capacity Not universally accepted, may give adverse color, taste Low
Membrane techniques
Electrodialysis High Non-specific Can remove other ions, used for high salinity, no chemicals Skilled operators, high cost, membrane fouling Very high
Reverse osmosis High Non-specific Can remove other ions, used for high salinity, no chemicals Skilled operators, high cost, membrane fouling, can remove beneficial solutes, residual saline wastewater Very high
Nanofiltration High Non-specific Can remove other ions, no chemicals Skilled operators, high cost, membrane fouling, can remove beneficial solutes, residual saline wastewater Very high


Fluoride in Groundwater Copyright © 2022 by Pauline L. Smedley. All Rights Reserved.