Effluent conditioning of biodiesel production for biological treatment using clay as a separation coadjutant

Biodiesel is a renewable, non-toxic and sustainable biofuel, considered to be the main candidate for a fossil fuel alternative in many countries. However, its manufacturing process results in about 0.2 to 3 L of effluent per litre of biodiesel produced. In addition to an inherently high organic load, its composition includes byproducts, traces of unreacted chemicals and catalysts, which inhibit microorganism growth and prevent its direct treatment by tertiary methods. In this context, this work aims to evaluate the combination of the coagulation-flocculation technique with dissolved air flotation (DAF). Real wastewater samples were obtained by synthesising biodiesel from soybean oil in-loco and performing the required washing procedures. The highest turbidity reduction efficiency (92.03%) was obtained using 1200 mg ∙ L of clay. By using response surface methodology, it was possible to analyse the effect of the chosen experimental factors and show that the best results (81.28%, 58.95% and 89.34% for turbidity, oil and grease and chemical oxygen demand respectively) were obtained using 925 mg ∙ L of clay and 1000 mg ∙ L of coagulant. Ultimately, clay proved to be an efficient coadjutant in the removal of organic matter, oils, grease, suspended solids and soluble organic matter from the biodiesel wastewater. Moreover, its low cost over traditional flocculants makes it an attractive alternative to industrial wastewater treatment processes.


1.INTRODUCTION
The recent worldwide attention given to renewable energy sources is a result of raised concerns over long-term effects of global warming and new developments in the petroleum geopolitics [1]. In this context, owing to several other advantages regarding their sustainability, technical adaptability and environmental impact, biofuels are considered a promising alternative for mitigating greenhouse gas emissions [2].
Biodiesel is a renewable, non-toxic fuel produced through the transesterification of triglycerides [3] and considered to be the main alternative to fossil fuels in many countries [4]. The production consists of the chemical reaction of lipidsfats or oils of vegetable or animal originwith alcohol in excess, in the presence of a catalyst [5]. Subsequently, the biodiesel must be washed thoroughly, in order to remove by-products and other contaminants. This process is well-known as biodiesel washing (BW) and yields an effluent rich in unconverted oil, traces of the catalyst and significant amounts of alcohol used in production [6,7]. For each litre of biodiesel produced, approximately 0.2 to 3 L of wastewater is formed during BW [8].
According to CONAMA Resolution No. 430 of 2011 [4], inappropriate disposal of this effluent, contrary to the current environmental legislation, may result in harmful effects on the environment. Therefore, in order to comply with this resolution, wastewater should not be disposed of directly into the environment without a proper pre-treatment.
Biodiesel wastewater treatment is complex due to the high variation of its composition and exceptionally high content levels of oil, grease, chemical oxygen demand (COD), total organic carbon (TOC) and suspended solids (SS). Despite the existence of several wastewater treatment processes for BW effluent [9], the research community has focused its efforts on developing a costeffective method that can reduce successfully the content of contaminants and can be integrated into a biodiesel production plant. The current wastewater treatment solution is a combination of physico-chemical processes, which reach higher treatment efficiency than any single-step techniques [10].
Among those, the coagulation-flocculation (CF) process is considered to be the most promising pre-treatment step since it is well-known and capable of conditioning the effluent composition for a subsequent tertiary method [11]. Coagulation occurs with the addition of coagulant agents, which destabilise the colloidal suspension rapidly, and leads to sedimentationor flotationof the particle aggregates and a decrease in suspended solids content, oil and grease (OG), COD and biological oxygen demand (BOD) [12,13].
While aluminium sulphate is the most commonly used coagulant in wastewater treatment, providing excellent results for industries in several areas [14], the application of clay as an additive provides conditions for industries to treat its effluents and dispose of them with virtually no impact and at lower costs [15] compared to commercial polymeric agents.
Dissolved Air Flotation (DAF) is an effective method for removing low density particles in suspension and reducing turbidity. The air is dissolved under high pressure in a saturator vessel and then decompressed in the flotation tank under atmospheric pressure leading to formation of microbubbles. The microbubbles bond to the particles, increasing their buoyancy and allowing them to rise to the surface [16].
In this manuscript we report the application of a combined CF-DAF processusing clay as coadjutant agentin the pre-treatment of wastewater produced during biodiesel washing, a required purification step in the biofuel production. Combined, these techniques have the potential to reduce considerably the organic load in the effluent, conforming it to the constraints of tertiary treatment method.

2.1.Biodiesel Washing Wastewater (BWW)
Wastewater samples were provided by the Laboratory of Alternative Technologies (LTA) of the Federal University of Sergipe, were the biodiesel washing (BW) effluent was obtained from the biodiesel production on a bench-scale process. The process produces biodiesel through the transesterification of commercial soybean oil using methanol in a ratio of 1:6 (on a molar basis), with anhydrous potassium hydroxide as the catalyst (1% of the oil mass). After the separation of the glycerine (co-product), the biodiesel was washed successively with distilled water at 90°C until the washing water (effluent) reached neutral pH.

2.2.2.Content of Oils and Greases (OG)
OG content was determined by the solvent extraction method HORIBA S-316 and subsequent infrared spectrophotometric analysis using the HORIBA OCMA 350 equipment [17].

2.2.3.Chemical Oxygen Demand (COD)
COD was determined according to the method described by Baird et al. (2012) [17], in which the sample is digested in the presence of sulphochromic solution, silver sulphate (catalyst) and mercury sulphate (to remove chlorides). Sample digestion was performed in an ACD LABOR thermoreactor using borosilicate tubes with screw caps at 150 ± 2°C for 2 hours. The determination of COD was performed by measuring the absorbance of the sample digested at wavelength of 600 nm in a KASUAKI spectrophotometer, model IL-226.

2.2.4.Coagulation and Flocculation Assay
The coagulation-flocculation tests were performed using a PoliControl jar test, model Floc Control II; aluminium sulphate was chosen as coagulating agent. The coagulation process was conducted in stages, namely: rapid mixing at 120 rpm for 2 minutes; flocculation at 20 rpm for 15 minutes and sedimentation for 15 minutes [15]. Each mixture resulting from the coagulationflocculation test was characterized and subsequently sent to the DAF assay. The operating pH range for the coagulant is 5.5 to 8 [18]. Thus, the pH of the effluent was adjusted to this range using a HCl solution 0.1 M.

2.2.5.Experimental Design
To evaluate the best operating conditions in the coagulation / flocculation processes, a factorial design 5 2 was used, in which the concentration of the coagulating agent and the additive (clay), were the independent variables. The experimental design levels and their respective real values are shown in Table 1. Statistica 7.0 was used for the data treatment. The process efficiencies in the removal of COD, OG and turbidity were used as response variables to obtain optimal process conditions. The results were evaluated by response surface methodology.

2.2.6.Dissolved Air Flotation (DAF) Assay
The experimental apparatus was composed of air compressor, saturation vessel and flotation tank (Figure 1). In this system, the water was initially saturated with air under pressure of 4 bar for 30 minutes at 25°C [19]. Subsequently, the saturated water was added to the effluent in the proportion of 15% by volume, causing the drag of particle aggregates to the surface. Samples from the clarified effluent were collected from an outlet valve located at the lower half of the treatment vessel and stored at room temperature before characterization.

Removal Efficiency
The results obtained in the characterization of the raw and treated effluents were used in the calculation of the efficiency of removal of the composition parameters, according to Equation 1.
Where is the removal efficiency (%) for the parameter P, 0 and are the initial and final concentrations for the parameter P, respectively. Table 2 presents the values of the turbidity, COD and OG removal efficiencies obtained for each test shown in the factorial design. The highest turbidity removal efficiency, 92.03%, was obtained when the maximum concentration of clay (1200 • −1 ) was used. Similar behaviour was observed for the removal of OG, with maximum efficiency of 74.43%, but using 200 • −1 of coagulant. For COD, the highest efficiency was 92.58%, using 100 • −1 of clay and 800

3.RESULTS AND DISCUSSION
• −1 of coagulant. Figure 2 shows the response surface and respective level curves for turbidity removal.
Higher efficiencies for turbidity removal were observed in two distinct regions: (1) high clay concentrations and low concentrations of coagulant; (2) high concentrations of clay and high concentrations of coagulant. Considering economic factors, as clay is the material of lower cost, the best combination of factors corresponds to the situation (1), since it is the one that minimizes the use of the coagulating agent. On the other hand, a single optimum peak scenario was observed for OG removal, as it is shown in Figure 3 -which presents the response surface for OG removal.   At such scenario, DAF treatment method is capable of decreasing both levels for turbidity and OG at favourable operational conditionslow levels of coagulant agent and high levels of clayand efficiencies comparable to more complex methods [9]. The response surface shown in Figure  4 presents COD removal efficiency as a function of the clay and coagulant concentrations. Considering the results found, it is evident that the combinations of factors that confer the highest reduction efficiencies of the three composition parameters studied (turbidity, OG and COD) were not consistent. The behaviour of the OG was not in agreement with the conclusions drawn for the COD, since the region of maximum removal of this parameter corresponds to the region of minimum removal of OG. Therefore, a test was performed without the use of clay in order to eliminate doubts as to the efficacy of this additive in the treatment process. Removal efficiencies of 24.00% for OG and 82.82% for COD were obtained. The low removal efficiency of OG in the absence of clay proves that it has a positive effect on the proposed treatment. In the experiments in which clay was used, efficiencies for the removal of OG and COD were as high as 74.43% and 92.58%, respectively.
The OG contributes to the COD of the effluent, whose parameter can be subdivided into two contributions: soluble COD, related to organic compounds soluble or emulsified in aqueous medium, and insoluble COD, which is associated with the content of oils and greases or the presence of organic solids in suspension such as flocculated material. In this context, the effect of the coagulating agent is most evident in the removal of material that can be segregated in a separate step, such as oils and greases and the colloidal suspended solids that form flakes under certain conditions. On the other hand, soluble COD is less affected by the addition of the coagulating agent and more dependent on the presence of suspended solids with adsorption capacity to remove part of the organic molecules. From this point of view, high clay contents have greater potential to remove soluble COD by the mechanism discussed here.

4.CONCLUSION
The results obtained in this study show the influence of the concentration of coagulant and clay as additives for the conditioning of the effluent of biodiesel production studied in the present work in terms of the removal of turbidity, OG and COD.
According to the analysis, the best results were obtained when using 1000 • −1 of coagulant (aluminium sulphate) and 925 • −1 of clay. This conditions yielded 81.28, 58.95 and 89.34% for turbidity, OG and COD respectively.
Thus, clay proved to be efficient as a separation additive for the removal of organic matter, mainly oils and greases, suspended solids and soluble organic matter from the biodiesel washing effluent. While those parameters are still above the limits established by the CONAMA 430 resolution, the treatment demonstrated in this manuscript successfully lowered them significatively, conditioning the effluent for the secondarybiologicaltreatment. In addition, the low cost attributed of using clay over traditional flocculants makes it an attractive alternative to industrial wastewater treatment processes.