Brine recovery from reverse osmosis effluents using an automatic temperature control system: salt crystallization Open Access

Increasing industrial activities go hand in hand with the increasing rate of waste generation. Mining and smelting activities generate significant quantities of waste (Rybicka 1996; Rwiza et al. 2016). A global study shows that about 142 million m³/day of concentrated brine are generated from industrial and mining operations as a byproduct of reverse osmosis (RO) processes (Jones et al. 2019). Brine wastes pose major environmental problems such as impairing soil's ability to produce crops and forage as well as skin irritation for human beings (Roberts et al. 2010; Ariono et al. 2016; Chen & Yip 2018; Jones et al. 2019). The problem is more serious in developing countries where the mining sector is important for economic growth. For example, the strong brine effluent with total dissolved solids (TDS) ranging from 20,000 to 35,000 mg/L was produced from the RO process at North Mara Gold Mine (NMGM), inducing problems to surrounding communities and aquatic ecosystem health (Bitala et al. 2009; Hui et al. 2018). The trend is expected to increase with time if no proper management measures are taken (Bitala et al. 2009; Matebese et al. 2024). The primary alteration minerals present in the brines at NMGM include carbonates, sulphides, and chlorides (Almås & Manoko 2012; Kazimoto 2020).

The improper brine disposal might result in physicochemical and ecological impacts on the environment (Roberts et al. 2010; Ariono et al. 2016). The physicochemical consequence is influenced by temperature, contaminants’ type levels and salinity. Other common contaminants include metals and scaling agents (calcium and sulphate) that may change the physicochemical qualities of the receiving water (Ariono et al. 2016). This change may have a negative impact on aquatic life (Ariono et al. 2016). Ecological impacts, including pollution of subsurface aquifers because of leakage issues, and difficulties for breeding and migrating birds (National Research Council 2008; Randall 2010), can also occur.

Several brine treatment methods, such as electrodialysis (ED), membrane distillation (MD), evaporation crystallization (EC), and freeze crystallization (FC), have been reported. ED is one of the most effective methods for salt manufacturing, which utilizes brine discharge from a seawater RO facility. Conversely, ED still needs to be improved in order to meet industrial demands for reduced energy use (Tanaka 2003; Ariono et al. 2016). Another possible substitute for high-purity water harvesting and production of salt is MD; however, precipitation of inorganic (calcium) or organic constituents is a serious barrier to MD applications (Ariono et al. 2016). Despite the fact that EC can reduce brine volume, other salts are always present in the final salt output and the process is expensive due to the high-energy requirements (Randall et al. 2011). Both approaches can lower the quantity of brine, although the separation of the salt byproduct is a great challenge (Manana et al. 2015).

FC is a low-temperature brine treatment technology and a fast way of recovering salts such as chloride, sulphate and carbonate from individual and mixed brine samples at higher efficiency (Cipolletta et al. 2021). The advantages of FC include simplicity of operation, less energy consumption, and no requirement of extra chemical additives (Randall et al. 2011; du Preez et al. 2020). NMGM produces more than 27,500 mg/L (average amount) of waste from the RO process. The methods used to treat waste need more energy and are not effective in releasing some waste to the environment (Jones et al. 2014). There is a need to investigate how FC can be used to effectively treat wastes from the RO process. Therefore, this study aimed at using an automatic temperature control (ATC) system as an alternative treatment approach for the concentrated brine effluent from the RO process in mining operations.

Synthetic brine solutions were prepared using analytical-grade salts including sodium sulphate (Na₂SO₄·10H₂O), sodium carbonates (Na₂CO₃·10H₂O), and sodium chloride (NaCl·2H₂O) using tap water from the NM-AIST laboratory. Individual and mixed sample solutions were prepared using 32 g (of each in 50 and 100 mL) and 15 g (mixed from each salt in 100 mL), respectively. The prepared concentration of individual salts Na₂SO₄·10H₂O, Na₂CO₃·10H₂O, and NaCl·2H₂O ranges from 0.64 to –0.32 g/mL, while the concentration of mixed samples was 0.45 g/mL. A magnetic stirrer was used to obtain a homogeneous solution.

Analytical-grade reagents were used in this study without further purification. Batch experiments were conducted on water samples by varying parameters such as freezing temperature, brine concentration, and contact time. Parameters, such as TDS, dissolved oxygen (DO), pH, temperature, and electrical conductivity (EC), were measured using multi-parameter type Hach (HQ40d). The recovered mass was weighed using Precisa Gravimetrics AG, Series 520 BP. Cations and anions were analyzed using an atomic absorption spectrometer (AAS) and the DR 6000 spectrophotometer method 680 ⁠.

The ATC system comprises two separate refrigeration units, each optimized for different salt crystallization conditions. ATC-1, designed for Na₂CO₃ recovery, operates at temperatures ranging from −0 to −7 °C, reflecting the lower crystallization point of Na₂CO₃·10H₂O. ATC-2 operates at temperatures between −10 and −99 °C and is primarily used for NaCl·2HO, Na₂SO₄·10H₂O, and mixed brine recovery trials.

The selected freezing temperatures and times were based on the phase diagrams of NaCl·2H₂O, Na₂SO₄·10H₂O, and Na₂CO₃·10H₂O. For example, the crystallization temperature for NaCl·2H₂O was chosen at −26 °C because preliminary trials showed that lower temperatures led to excessive ice formation, reducing salt yield. Similarly, Na₂SO₄·10H₂O was recovered at −10 °C over 96 h to ensure optimal separation from co-precipitating salts, while Na₂CO₃·10H₂O required shorter freezing times due to its smaller crystal size and faster nucleation rate.

The ATC system maintains precise temperature control through a programmable thermostat, which adjusts freezing rates to prevent thermal shock. The gradual cooling (at 0.1 °C/min) ensures uniform nucleation and crystal growth.

Experiments began with the addition of about 32 g of brine into 50 and 100 mL of tap water using a glass reactor kept at room temperature. A 0.1 °C/min decrease in the bulk solution temperature was adopted to prevent thermal shock and consequent breakage of the glass reactor. The components of the reactor were collected and placed in a 1-L beaker that was maintained in a freezer. After allowing the separation of salt from ice in the separate beaker for ±10 min, the ice and its entrained brine were kept separately in separate beakers. A normal funnel with Whatmann filter paper (pore size of 0.45 μm) coupled to a 0.5-L flask, was applied in salt filtration for the experimentations that generate salt.

Experimental data were collected according to the effect of the tested parameters.

Data values for tap water presented the initial conditions of the water before the freezing crystallization process (Table 1). The tap water of the Nelson Mandela–African Institution of Science and Technology (NM-AIST) laboratory has a pH of 6.8 which is nearly neutral, its level of salts is within the recommended level (e.g. 17.11 mg/L of chloride and 11.00 mg/L of sulphate which are below the acceptable limit set by the World Health Organization (WHO) and Tanzania bureau of standards (TBS) of 250 and 50 mg/L, respectively. The tap water's conditions were suitable for this study because its salt levels have no significant effect on the final salt levels of the synthetic water.

• (a) Chloride

The mass of recovered salt (Na₂SO₄·10H₂O) crystals was found to be influenced by solution concentration. Figure 4 shows that for about 96 h of the freezing time, the maximum recovered mass of salt crystals, up to 29.9 g (93.3%) and 22.7 g (71.1%) for sulphate solution in 50 and 100 mL, respectively, with an initial mass of 32 g at −10 °C fixed freezing temperature, has been achieved. The mass of salt crystals recovered was directly proportional to the solution concentration. Moreover, the recovery efficiency for a mass of the salt crystals from the brine solution was reduced with a further increase in freezing time (96–144 h) from 93.4 to 82.75% in 50 mL. The visual investigations revealed that more freezing time enhanced the growth and stabilization of the crystals. The findings of this study were supported by Randall & Nathoo (2015) who reported that the mean crystal size was enhanced by the longer residence times.

• (c) Carbonate

Furthermore, the brine recovery based on different solution concentrations in 50 and 100 mL, the summary for the final mass recovery, freezing temperature, freezing time, and efficiency of both salts are present in Table 2.

However, at a constant mixed brine solution of 100 mL and a freezing time of 3 h, the effectiveness of mass recovered brine was reduced from 26.86 to 0.67 g with an increasing freezing temperature ranging from −10, −15, −20, −25, and −30 °C, as shown in Figure 4. This verifies that ­10 °C was the optimum temperature required to recover the mass of salt crystals from the mixed salts samples with the following composition Na⁺ (34.1%), Cl⁻ (1.6%), ⁠, ⁠, and impurities (20.57%). Moreover, this study suggests that at lower freezing temperatures (e.g., −30 °C), the mobility of ions decreases, reducing the rate of crystal formation. Therefore, at very low temperatures, the system may show incomplete crystallization within the fixed time.

For analysing the individual constituents of the recovered brine mass, the ATC system used in this study was feasible. However, the recovered mixed brine mass revealed that its capacity to crystallize and separate the mixture to their constituents to obtain a high-purity product was low. The high impurity content (e.g., 20.57%) in mixed samples suggested to interfere with the nucleation and crystal growth processes, further lowering the recovery efficiency of the salts.

The low selectivity observed in mixed-salt experiments was consistent with thermodynamic solubility predictions. The phase diagram analysis shows overlapping crystallization regions for NaCl ·2H₂O, Na₂SO₄·10H₂O, and Na₂CO₃·10H₂O at similar temperature ranges, which explains the difficulty in isolating pure salts. Moreover, the additional experiments with stepwise freezing temperature adjustments, which improved selectivity still resulted in mixed-salt formation.

In mixed-brine samples, the competitive crystallization process significantly affected salt recovery. NaCl·2H₂O has the lowest solubility at lower temperatures, crystallized first, and followed by Na₂SO₄·10H₂O. However, Na₂CO₃·10H₂O exhibits a lower recovery efficiency (32%) due to its strong interactions with which inhibits crystal growth. These interactions reduced the effectiveness of phase separation, explaining the lower purity in recovered salts.

The non-target ions (e.g., Mg²⁺, Ca²⁺) and organic matter led to impurity incorporation into crystal lattices, reducing separation efficiency. It has been found that impurities were particularly high in mixed samples (20.57%), which lowered crystallization rates and affected crystal morphology. These findings align with previous studies (Randall & Nathoo 2015) that reported impurity entrapment as a great challenge in mixed-salt FC.

Moreover, this method is more efficient based on treatment time, temperature, and energy than freeze-drying technologies, such as microwave drying, vaporization, and surface heating. Jones et al. (2014) reported that their inefficiency, mostly due to several times high heating rates, high heat of vaporization, and energy of surface heating was lost as radiant energy, respectively.

Energy consumption plays a critical role in the recovery of valuable chemicals such as NaCl, Na₂SO₄, and Na₂CO₃ from brine. Efficient energy consumption is essential to minimize operational costs. In this study, the energy consumption of the ATC system was found consistently to increase with freezing time. For instance, at 24 h, the energy consumption reached 9.6 kWh/m³, while at 144 h, escalated to 57.6 kWh/m³. The linear increase in energy consumption reflects the stable performance of the ATC system.

By comparing the ATC system with other desalination and recovery technologies, the energy consumption of ion exchange processes typically ranges from 0.2 to 2 kWh/m³, depending on the specific application, short operating time (h), and system design (Youssef et al. 2014). The energy consumption of nanofiltration systems ranges from 0.5 to 4 kWh/m³, depending on the feedwater quality, pressure requirements, system configuration, and short operating time (h) (Schäfer et al. 2018; Njau et al. 2023). Sustainable distillation using solar energy has an average energy requirement of 30 to 150 kWh/m³ for large-scale systems; however, the energy comes from the sun, the operational costs are nearly zero and long operating time (days). Table 3 compares the energy consumption and the cost of operation between ATC and the methods mentioned.

Moreover, the cost of electricity directly affects the economic feasibility of the ATC system. Tanzania's competitive electricity tariff of 0.087 USD/kWh significantly reduces the overall operating costs, making the ATC system a financially viable option for salt recovery. Therefore, the longer the system operates, the more energy it consumes. This was supported by Figure 8, where energy consumption increases proportionally with freezing time.

Additionally, the present study primarily focused on the feasibility of ATC for brine recovery through freezing crystallization, however, established technologies, such as ED, MD, and EC, have been extensively studied for brine treatment. ED requires significant energy input and pre-treatment to mitigate scaling, MD faces limitations due to membrane fouling and high operational costs, and EC requires high-energy consumption. ATC offers a low-energy, chemical-free alternative for selective salt crystallization, particularly in resource-constrained settings. Table 4 summarizes the advantages and disadvantages of mentioned technologies in terms of energy consumption, recovery efficiency, and operational complexity.

The laboratory trials for the recovery of salt crystals in synthetic brine solutions of individual and mixed samples using a locally available ATC system have been presented in this study. The maximum recovered efficiencies, 85.3% (NaCl·2H₂O), 93.3% (Na₂SO₄·10H₂O), and 32.0% (Na₂CO₃·10H₂O), were achieved from 32 g initial dissolved mass in 50 mL under fixed operating conditions of freezing temperatures −26, −10, and −2 °C, respectively, at the varied freezing times (1 to 144 h). For the mixed brine samples at −10 °C and 3 h in 50 mL, 29.37 g of brine (65.27%) was recovered with main constituents such as Na⁺ (34.1%), Cl⁻ (1.6%), , ⁠, and impurities (20.57%). This shows that the system was not feasible in separating mixed brine from its salt. The efficiency of ATC method is efficient in the recovery of salt crystals for NaCl·2H₂O (85.3%), Na₂SO₄·10H₂O (93.3%) and inefficient to Na₂CO₃·10H₂O (32.0%). The findings reported by this study through batch experiment suggested that ATC could be used as an alternative technology for the recovery of the concentrated brine effluent from the industry and mining operations sites. The ATC system demonstrates a significant advantage in terms of energy efficiency and cost-effectiveness, especially in regions with low electricity tariffs. It maintains low and predictable energy consumption due to its constant power demand of 0.4 kW.

Testing the recovery of the concentrated brine effluent from the industry and mining operations sites using an ATC system is highly recommended for future research. The analysis of the percentage composition of the recovered mixed sample to its entire constituents at varied freezing temperatures is also recommended in future studies. An advanced technology, such as a crystallizer, is needed to ensure simultaneous crystallization and separation (ice–salt–brine) of the salt crystals from mixed brine samples to improve the recovery efficiency of the useful products. Further research could explore process optimizations to reduce energy consumption or improve system efficiency. To improve purity, the integration of a secondary washing stage using deionized water or sequential crystallization with controlled temperature gradients and the use of controlled seeding techniques to promote selective crystallization were also recommended.

That this work was financially supported by The Tanzania Gemmological Centre 's highly appreciated. The North Mara Gold Mine in Tarime, Tanzania is also acknowledged for supporting laboratory work on synthetic brine recovery using an automatic temperature control system.

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

The authors declare there is no conflict.