Efficient systems for groundwater defluoridation: a pilot study for rural schools in southern Brazil

Water is essential for human life, socioeconomic development, and environmental protection. However, population growth in recent decades, climate change, and inefficient natural resource management have raised global concerns about water resource availability (Mujtaba et al. 2024). Among the 17 Sustainable Development Goals (SDGs) of the UN Agenda, SDG 6 aims explicitly to ensure safe drinking water and sanitation by 2030 (Shaji et al. 2024).

In this context, fluoride in groundwater at concentrations of 0.6–0.9 mg L⁻¹ is essential for dental and bone health (State Health Department 1999). However, at elevated levels (>1.5 mg L⁻¹), it is harmful to health (Protocol GM/MS N° 888 2021), causing problems such as skeletal and dental fluorosis, increased insulin resistance, decreased estrogen and testosterone levels, and neurological damage (Rathnayake et al. 2022; Gu et al. 2024).

Fluoride is naturally found in groundwater, and contamination occurs due to volcanic eruptions, atmospheric deposition, and the dissolution of fluoride-containing rocks (Lacson et al. 2021; Tong et al. 2023). Fluoride in water can also be attributed to human activities through industrial emissions, fertilizers, and waste disposal (El Messaoudi et al. 2024).

Thus, the World Health Organization (WHO 2011) has set the maximum acceptable fluoride limit for drinking water at 1.5 mg L⁻¹ (Lacson et al. 2021). The United States Environmental Protection Agency (USEPA) has set it at 2.0 mg L⁻¹ (Podgorski & Berg 2022), the European Union (EU) between 0.7 and 1.5 mg L⁻¹ (Deng et al. 2024), and in China, the maximum allowed fluoride concentration in drinking water is 1.0 mg L⁻¹ (Tong et al. 2023).

In Rio Grande do Sul, particularly in the central region, in the municipalities of Restinga Seca, Agudo, Paraíso do Sul, and Novo Cabrais, high fluoride concentrations have been observed in groundwater, with values of 5.6, 4.2, 5.0, and 3.6 mg L⁻¹, respectively (Luiz et al. 2018).

Due to the adverse effects caused by elevated fluoride levels, removing fluoride from groundwater and surface water has been a challenge in water resource management (Mohandas et al. 2023). Various technologies for drinking water treatment have been proposed for defluoridation, such as adsorption, biosorption, ion exchange, coagulation/precipitation, membrane processes (reverse osmosis and nanofiltration), electrodialysis, electrocoagulation (Halder et al. 2023; Shaji et al. 2024), and phytoremediation (Weerasooriyagedara et al. 2020). Adsorption for fluoride removal using activated bone charcoal offers the best cost–benefit ratio (Lacson et al. 2021) due to its simple operation and wide availability of adsorbents (Kimambo et al. 2019). It allows the development of sustainable, low-cost, and easily maintained technologies, which are also innovative for implementation in emerging countries to ensure drinking water safety (Gao et al. 2022; Medellín Castillo et al. 2023; Shimabuku et al. 2023).

Therefore, this study aimed to evaluate the defluoridation system of groundwater in rural schools of the Taquari-Antas Watershed in the municipality of Venâncio Aires, RS, Brazil, using activated bone charcoal as the adsorbent medium, applying a sustainable, low-cost technology that is easy to implement, operate, and maintain.

Data collection on the number of students, teachers, administrative staff, collection points, water consumption sources, and water supply networks was conducted in collaboration with the school administrations through the Department of Education of Venâncio Aires, RS, Brazil. Five rural schools (A–E) were pre-selected. After sample collection and analysis of all institutions' natural water supply networks, only three rural schools (A, B, and C) showed concentrations exceeding the maximum permitted in Rio Grande do Sul, Brazil (State Health Department 1999).

Defluoridation systems for groundwater were constructed and installed, using activated bone charcoal as the adsorbent (20 × 50 mesh, internal surface area of 120 m² g⁻¹, pore size range from 7.5 to 60,000 nm), acquired from Bonechar (Carvão Ativado do Brasil LTDA, Maringá, Brazil).

The Taquari-Antas is a sub-watershed located within the hydrographic region of the Guaíba watershed, RS, covering an area of 26,430 km² and with an estimated population of 1,383,442 inhabitants (2020), including 1,081,261 in urban and 302,181 in rural areas (SEMA 2024).

The groundwater defluoridation system was sized for installation at the point of consumption to remove excess fluoride only from direct consumption water. Thus, the daily water consumption volume, the initial fluoride concentration in natural water, and activated bone charcoal adsorption capacity were considered. Other relevant considerations for sizing are the following:

• I. The activated bone charcoal used in the filter has the following specifications: 20 × 50 mesh, density of 0.675 g, porosity of 65%, and adsorption capacity of 1,200 mg kg⁻¹.

• II. The full concentration of fluoride ions in natural water is 4.0 mg L⁻¹.

• III. The contact time ranges from 8.5 to 250 min, with the ideal time being 60 min (F_o = 1).

Based on all considerations, calculations were performed to size the groundwater defluoridation systems for schools (A, B, and C), as shown in Table 1. To meet the needs of rural schools, the system was designed with the premise of a daily consumption of 2 L day⁻¹ per consumer and an average fluoride concentration of 0.8 mg L⁻¹. The estimated mass of activated bone charcoal ranged from 5 to 12 kg, covering 4,320 to 7,020 L of defluoridated water. Thus, the rural schools' defluoridation systems were sized with 14 kg of activated bone charcoal.

Sample collection followed the Standard methods for the examination of water and wastewater guidelines (Miner 2006). The initial water flow was discarded at a slow rate for 2 min. Then, 500 mL of water was collected in a polypropylene (PP) bottle, stored in thermal boxes with ice, and transported to the laboratory for analysis, following APHA guidelines.

Although the maximum fluoride limit in public water supply is 1.5 mg L⁻¹ according to the WHO (2011) and the GM/MS Protocol No. 888 (2021), for this study, the maximum in treated water was set at 0.9 mg L⁻¹. This value corresponds to the maximum limit stipulated by the legislation of the State of Rio Grande do Sul, which determines the range of 0.6–0.9 mg L⁻¹ as ideal for dental and bone health (State Health Department 1999). Thus, for the dimensioning of this system, the ideal value (final concentration) was determined to be 0.8 mg L⁻¹.

Identical defluoridation systems were structured, with one unit installed to perform preliminary tests on the University of Santa Cruz do Sul (UNISC) campus and the other units in rural schools (A, B, and C).

The defluoridation system was evaluated in three stages. In the first stage, the equilibrium time in the adsorption process and the adsorption capacity were determined in bench tests using a jar test agitator (Quimis, model Q-305D23, Diadema, SP, Brazil). In the second stage, a prototype of the defluoridation system was installed and monitored on the UNISC campus. In the third stage, the defluoridation systems were installed in the three rural schools and continuously monitored to assess the efficiency of the treated water under actual usage conditions.

The tests were conducted with samples of activated bone charcoal with a particle diameter of 20 × 50 mesh, which were previously dried in an oven (Quimis, model Q-317B-22, Diadema, SP, Brazil) at 100 °C for 3 h. The evaluation was done with three distinct bench tests (1, 2, and 3). In tests 1 and 2, naturally fluoridated water was used, with a fluoride concentration of 2.24 mg L⁻¹, using 2 g (test 1) and 4 g (test 2) of activated bone charcoal, respectively. For test 3, an artificially fluoridated solution was prepared using deionized water and sodium fluoride P.A. (NaF) (Nuclear brand, Diadema, SP, Brazil), with a fluoride concentration of 2.9 mg L⁻¹ and 2 g of activated bone charcoal. The three adsorption tests were monitored for 480 min in a jar test at 120 rpm.

The prototype of the defluoridation system was installed on the UNISC campus. It was monitored daily for 25 days until the fluoride concentration in the treated water reached values equal to that of the natural water (meaning the complete saturation of the adsorbent medium). Thus, the lifespan of this prototype was established when the filtered water results reached fluoride concentration values greater than 0.9 mg L⁻¹.

The defluoridation systems were structured with all the necessary components for assembly, using 14 kg of activated bone charcoal, and installed and monitored in the rural schools (A, B, and C) through physicochemical, pH, turbidity, total dissolved solids (TDS), and fluoride analysis for 60 days.

Due to the effects on human health caused by high fluoride concentrations, removal from groundwater has been a challenge in water resource management (Mohandas et al. 2023). Thus, traditional and effective techniques for water defluoridation, such as adsorption, biosorption, ion exchange, precipitation, membrane processes (reverse osmosis and nanofiltration), electrodialysis, and electrocoagulation are managed. Some techniques have disadvantages; for example, the ion exchange process has been tested in India, and it has been shown that its efficiency is affected by other ions such as carbonate, sulfate, alkalinity, and phosphate. In addition, it is not economically viable due to the high cost of synthetic resins. Nanofiltration is highly efficient at removing excess fluoride from water, while membrane separation technology removes high levels of ions, some of which are vital for human health (Halder et al. 2023). Compared with existing studies, among the technologies mentioned, in the global context, adsorption on activated bone carbon presents the best cost–benefit ratio (Lacson et al. 2021), allowing for the development of low-cost, sustainable technologies that are easy to maintain and innovative enough to be implemented in emerging countries (Medellín Castillo et al. 2023).

The results indicate that the equilibrium time for the adsorption process occurs after 360 min of contact. It is observed that in the artificial sample (test 3), the adsorption process extends until the end of the contact time of the tests.

Considering the principle of adsorption (Costa et al. 2013; El Messaoudi et al. 2024), which involves adsorbate ions on solid surfaces, reducing the concentration in water, the results in this study showed maximum adsorption capacity values of 0.446, 0.331, and 0.759 mg g⁻¹ for tests 1, 2, and 3, respectively (Table 2).

Table 3 presents the operational results of the defluoridation system.

The results obtained from the filtered water show an immediate reduction in fluoride concentration, maintaining below 0.1 mg L⁻¹ until reaching 4,300 L of treated water. Fluoride concentration increased in the filtered water, reaching 0.9 mg L⁻¹ with 7,100 L of treated water.

The filtration process continued to be monitored until the fluoride concentration in the water reached values equivalent to those of the natural water, indicating the complete saturation of the adsorbent medium. However, this prototype's lifespan was determined when the fluoride concentration in the filtered water exceeded 0.9 mg L⁻¹, indicating the saturation of the adsorbent medium. At this point, the ratio C_i/C₀ corresponded to approximately 0.39, where C₀ is the initial average fluoride concentration (2.41 ± 0.20 mg L⁻¹, n = 19) and C_i is the fluoride concentration at time i (0.9 mg L⁻¹). Thus, the lifespan of this system is reached after the filtration of 7,100 L of natural water (C_i of 2.41 ± 0.20 mg L⁻¹; n = 19). The breakthrough point was reached after the filtration of 4,300 L. At that point, C_i/C₀ = 0.1, i.e., when the fluoride concentration in the treated water reached 10% of the initial concentration (C₀).

The performance of the campus defluoridation system aligns with the design for the sizing, intended for the consumption of a school unit with up to 39 people consuming 2 L day⁻¹ for a minimum period of 90 days.

The evaluation of the monitoring of natural water indicates alkaline properties (pH = 9.0 ± 0.1, n = 18), absence of turbidity (<0.05, n = 18), and an average TDS concentration of 340 ± 27 mg L⁻¹, n = 18. In the filtered water, an initial increase in alkalinity was observed due to the alkaline properties of the adsorbent medium, activated bone charcoal. This pH increase is due to the natural water having high values, as evidenced by the results at the beginning of the filtration process. At the beginning of the filtration process, the TDS concentration is above the maximum allowable limit (Protocol GM/MS N° 888 2021). This occurs due to the leaching of dissolved salts in the activated bone charcoal. As for the turbidity results, only in the first sampling were the values above the maximum permissible value (MPV), corresponding to when the defluoridation system was installed (State Health Department 1999).

Upon reaching a volume of 1,218 L of filtered water from the campus prototype defluoridation system, natural and filtered water samples were collected and analyzed as stipulated in Protocol GM/MS N° 888 (2021). The results are presented in Supplementary material 2.

The results showed that pH (9.35) and ammonia concentration (1.7 mg L⁻¹) did not comply with Protocol GM/MS N° 888 (2021). As previously discussed, the alkaline properties of the natural water and the adsorbent medium resulted in filtered water with values close to the permissible limit for public water supply. This interference in pH at high levels in the natural water is minimized using the defluoridation system.

The ammonia results (1.7 mg L⁻¹) in the filtered water exceeded the maximum permissible limit set by Protocol GM/MS N° 888 (2021) of 1.2 mg L⁻¹, while (WHO 2011) considers that groundwater can have concentrations of up to 3 mg L⁻¹, as ammonia in water is due to bacterial contamination, effluents, or animal excrement (Seben et al. 2021).

The monitoring results of the defluoridation systems indicated average fluoride values in the natural water supplying schools A, B, and C of 2.6 ± 0.29 mg L⁻¹, n = 13, 2.4 ± 0.20 mg L⁻¹, n = 10, and 3.3 ± 0.14 mg L⁻¹, n = 14, respectively. After filtration, the average values were less than 0.1 mg L⁻¹. The operational data of the system can be viewed in Supplementary material 3.

The results of fluoride concentration in the natural water of the rural schools show high values, with rural school C presenting the highest values (3.4 ± 0.1 mg L⁻¹; n = 14), corresponding to 3.8 times the MPV (0.9 mg L⁻¹) (State Health Department 1999).

In Protocol GM/MS N° 888 (2021), the MPV for fluoride is 1.5 mg L⁻¹, 2.3 times the maximum limit. The natural occurrence is due to the interaction and weathering between water and rocks. Seben et al. (2022) reported a variation of 0.8–39.5 mg L⁻¹ in fluoride concentration in groundwater in Malawi, Africa. Similar results have been reported in other studies in which other techniques were used for defluoridation (Hadoudi et al. 2023). Thus, the chemical processes involved in fluoride removal are complex and have been widely studied in research (Zaki et al. 2023).

The evaluation of fluoride concentration in the filtered water showed values below 0.1 mg L⁻¹ throughout the monitoring period of the defluoridation systems. The volumes of filtered water in rural schools A, B, and C were 5,660, 2,612, and 4,805 L, respectively. This variation is due to the number of consumers, 31, 24, and 39, respectively, over 60 days. The defluoridation efficiency results were >97.4, >98.3, and >97.6%, respectively. The average results were above 97%, guaranteeing the supply of water with fluoride levels suitable for human consumption. More recent innovations in water treatment technology have shown excellent results, with 90–100% removal of fluoride from groundwater (Fraiha et al. 2024).

Although the State Health Department (1999) defines acceptable fluoride concentration levels in the water supply to be between 0.6 and 0.9 mg L⁻¹, the near-complete removal of fluoride by defluoridation systems does not compromise the water's potability standards. The sustainability of these systems is of great importance, especially in rural areas, as highlighted in recent research (Hadoudi et al. 2021).

Implementing large-scale defluoridation systems is challenging, but they can be scaled up regionally and nationally. However, installation and operating costs must be considered, as well as the replacement of activated bovine bone charcoal and parts and equipment used to size the system.

The long-term viability and maintenance requirements of defluoridation systems in areas with limited resources involve various factors, such as operating costs, financial resource availability, infrastructure, maintenance, and adsorbent material supply. Simpler technologies, such as adsorption with cheap and easily accessible materials, tend to be more viable in resource-limited locations, although they still require ongoing attention to ensure safe and efficient operation.

The study on the defluoridation system, using activated bone charcoal as the adsorbent medium, provided promising information for implementing effective and sustainable methods in groundwater treatment.

The performance results of the campus defluoridation system aligned with the sizing structure. Thus, this system can meet the consumption needs of a school with up to 39 people consuming 2 L day⁻¹, with a minimum flow rate of 3.3 L h⁻¹, for a minimum period of 90 days.

This study demonstrates the removal of excess fluoride from natural water using activated bone charcoal as the adsorbent and its adsorption efficiency in rural schools A (>97.4%), B (>98.3%), and C (>97.6%) throughout the entire operation period, meeting potability standards.

The adsorption process for removing fluoride ions is the most sustainable method due to its main characteristics: practicality, wide availability of adsorbents, efficiency in removing fluoride ions, operational cost feasibility, and sustainability. Indeed, the system developed by adsorption with activated bone charcoal presented an investment cost of US$115.

The adsorption system for groundwater defluoridation implemented in this study features an easy-to-apply technology. It is low cost to implement, operate, and maintain, contributing to improving sanitation conditions in rural schools.

The authors thank the National Council for Scientific and Technological Development (CNPQ) – protocols 306216/2022-1 and 302816/2023-2, Coordination for the Improvement of Higher Education Personnel (CAPES) – code 001, 88881.710390/2022-1, Senai Institute of Leather and Environmental Technology, and Philip Morris Brasil.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.