Characterization of effluents from FSEs
Four pertinent wastewater quality parameters were analysed, viz combined fats, oil and grease (FOG); total suspended solids (TSS); pH; and chemical oxygen demand (COD). It should be noted that a composite sample (C) was also analysed to represent the realistic situation of what any grease hauler would bring to a wastewater treatment plant for treatment. This sample “C” is the mixture of FSE effluent samples “K” (baking-based), “M” (frying-based) and “W” (mixed-based)
Table 1 summarizes the characteristics of FOG-bearing effluent (emulsified layer) from three FSEs (via composite sample “C”) based on sampling and analyses discussed in the forgoing section.
Table 1: Characteristics of FSE Effluent
It may be noted that the parameter with the highest concentration is COD (average of 2229 mg/l), followed by FOG (average, 511 mg/l) and TSS (average, 446 mg/l). The above results reveal that the combination of samples would be ideal and realistic to be envisaged for its physico – chemical treatment by coagulation-flocculation-sedimentation.
The physico-chemical treatment selected in this study is conventional coagulation, flocculation and settling via Jar Tests. Poly-aluminium Chloride was used to perform coagulation. The effects of additions of poly-electrolytes (supplied by General Electric) were also examined. Poly-aluminium Chloride ([Al(OH)aClb]n, where (a + b) = 3, with a >1.05) (Water New Zealand 2013)  is a class of soluble aluminium products where the aluminium chloride has been partially reacted with a base. It contains some of the highly cationic oligomers of aluminium, which makes it an excellent choice for coagulation. Polyelectrolytes are used for floc build up as they bridge between particles and charged aluminium flocs (Eckenfelder Jr. 1966, 280) . They are also absorbable meaning small dosages are effective to encourage floc build up. Colloids of oil and grease are hydrophobic which often retard flocculation and frequently require special treatment to achieve effective coagulation (Eckenfelder Jr. 1966, 280) . Additionally, the vast majority of colloids are negatively charged (with a range of the zeta potential between -12 to -40 mv). Consequently, coagulation is induced by the addition of high valence cations. Optimum coagulation occurs when the zeta potential is zero (isoelectric point – which occurs at a specific pH for the respective chemical), with effective coagulation at +/- (-5) mv (Eckenfelder Jr. 1966, 283).
Figure 1 shows removal rates of FOG at varying dosages (0 mg/l – 650 mg/l) of a 5% Stock Solution of PACl subsequent to coagulation and flocculation of the emulsified sample. Better removal rates were seen for dosages ranging between 150mg/l to 300 mg/l, with a drastic drop in removal for dosages higher than 300 mg/l. 250 mg/l dosage shows the optimum removal (89.90%), indicative of the dose which makes the emulsion most unstable. The 86.7% removal rate for 0 mg/L is indicative of a larger percent of grease being floatables after coagulation and flocculation. pH remained on the higher end of the characterization range i.e. 6, which is expected as PAC is reacted on the base end of the pH scale even though carrying hydrogen ions (H+) which contributes to acidity. The removals of TSS ranged between 11%-65% and that of COD in the range of 52% - 91%.
Figure 2 shows that oil and grease removal rates not only increased with altered pH, but removal rates generally remained above 90 percent over a pH range of 6-10. Optimum removal of 98.98% was achieved for a pH of 8. TSS removal rates (93 % to 99 %) were also in alignment with FOG rates which may be attributed to the increase in destabilisation of colloids. pH range varied between 6 and 8 which is more towards the base end, as expected with the addition of NaOH. On the other hand, the upper range of removal COD dropped to 61%, which is expected with the addition of a chemical (NaOH).
Figure 3 illustrates the effects of the optimum PAC dosage (250 mg/L) and pH adjustment of samples to 8, with addition of low anionic polyelectrolyte dosages to assess its efficacy in the treatment. Figure 3 shows there was no significant change in FOG removals with the addition of the low anionic polyelectrolyte, rather, with increased dosages past 1mg/l, there was a decreased trend in the removal of oil and grease. This is an indication that the probable net charge left after coagulation is negative. TSS removal remained in the 90%, with the exception of the polyelectrolyte dosage of 6mg/l, which gave a removal of 41%. COD removals generally dropped in a range of 40% - 60% with two high removals of 76% and 85% for 0 mg/l and 7 mg/l of polyelectrolyte added respectively. pH went unaffected by the addition of the polyelectrolyte.
Based on the insignificant change by the addition of the anionic polyelectrolyte, a low cationic polyelectrolyte was tried. Figure 4 shows the results.
It is seen that optimum performance regarding the removal of emulsified oil and grease was achieved with a poly-aluminium chloride (PACl) dosage of 250 mg/l for a 5% Stock with the influent sample adjusted to a pH of 8 (seen to be isoelectric point for PACl) and with a low cationic polyelectrolyte dosage of 4 mg/l. This resulted in a removal rate of 99.9% with an FOG residual of 0.17 mg/l. TSS removals remained in a high range (90 percentile range) as anticipated. However, COD removals significantly dropped because of the polymeric organic matters therein; noting the removal rate at 4mg/l was 53 % which is consistent with other results.
In summary, the aforementioned optimum result was obtained following several experiments, where initially the addition of no chemicals were explored, followed by the addition of coagulants only. Upon deriving at the best coagulant dosage, the variation of pH ranges were explored for optimum results. Then under the optimal coagulant dosage and pH range, experiments were conducted to determine the effect on removal levels at various dosages of cationic and anionic polyelectrolytes.
Pilot Scale Study
The pilot scale study was conducted to not only confirm the results at the bench level study but to provide additional data that would aid in scaling up. It should be noted that pilot scale limit is designed and built so that the process can be better understood. It provides boundary conditions that allow scale up from the prototype to the commercial plant (Whalley 2016) . A pilot plant is meant to be flexible and adaptable so that the researcher can do modifications to test configurations.
The engineering parameters critical for the run of each sample were: wastewater volume, temporal mean velocity gradient (G), tip speed and the resulting rpm for the mechanical agitators. Average volume of samples at the pilot scale was 40 litres, as compared to bench scale level volume of 2 litres. Two agitators were used, namely a Cole Palmer 50000-40 and a Cole Palmer 50000-60, each of propeller type, of diameter 0.0635 m, which were used both in the coagulation and flocculation tanks. In the absence of values for the “Kt” constant normally provided by the suppliers of the mechanical agitators, use was made of the literature. As reported for Propeller type agitators with a pitch of two and 3 blades Kt for turbulent and laminar flows are 1 and 43.5, respectively (Tchobanglous, Burton, and Metcalf and Eddy 1991) . Tip speed (TS) and Reynold’s Number were monitored to ensure all parameters remained in correct ranges (as listed in the following paragraph); however, velocity gradient was the parameter controlled. Equations for the aforementioned are as follows:
a) G = √(P/(µ*V))
b) TS = 2Rὠπ
c) Power (P) = Kt/g *ν*N3*D5
d) Re =( ρND2)/µ
The power of the stirrers are 74.5 W with an efficiency of 80% as stated by Cole Palmer. Where ideal parameter ranges are: Coagulation; G 300-1000s-1 ,Tip Speed x>3 ms-1 and RPM < 1500. Flocculation; G 20-60 s-1 , Tip Speed 0.15 ms-1<x<0.6ms-1 , RPM x< 100 (Tchobanglous, Burton, and Metcalf and Eddy 1991) .
All the chemical parameters considered at the bench level study remained at the optimum dosages and strengths (PACl 250 mg/l dosage and 6N Sodium Hydroxide for adjustment of pH at 8) except the polyelectrolytes. Polyelectrolyte dosages and strengths coupled with varying settling times were manipulated in order to attain optimum results based on limitations faced mainly by equipment type and dimensions i.e. agitator type used in flocculation tank. Subsequently, the Kt value and subsequent calculations for mixing were used to determine the necessary rpm to allow optimum bridging of particles, in the flocculation tank. Detention times were also varied in the first tank (Pre-sedimentation tank) to allow a proper representation of emulsified oil and grease to be treated. Time varied between thirty (30) minutes and one hundred and twenty (120) minutes with an average of one hundred and seven (107) minutes. In addition, varying strengths and dosages of the polyelectrolytes were manipulated so as to obtain the optimum results, this aids as increased strength and dosages (strength – increased functional groups /unit grams and increased dosages – number of polymer groups per unit volume of liquid) would have served by increasing the charge density for agglomeration where the mechanical aspect was insufficient for the same.
Figure 5 illustrates the results at the pilot plant study. With all pre-determined bench level treatment parameters remaining at optimum dosages and strengths, with variation only in the polyelectrolyte dosages and strengths, and the settling times at pre sedimentation and final tanks; optimum results were attained for a dosage of 5mg/l Medium Cationic (0.1% Stock) and a final settling time of six (6) hours respectively. Results yielded 97.4%, FOG removal and resultant residual of 16.8 mg/l. This result differs from the dosing strength for the polyelectrolyte that attained optimum results at the bench level study of 4mg/l Low Cationic (0.1% Stock) for a removal rate of 99.9% (0.17 mg/l residual). It is concluded that the residual concentration in the bench level study is significantly below that generated at the pilot plant may be attributed to the limitations of equipment used in the pilot plant study, as discussed above.
The secondary parameters after treatment were as follows; pH consistently remained at 7 after treatment, TSS had over 90 percent removals and residuals ranged between 30 mg/l and 14 mg/l. COD removal rates on the other hand were not significant ranging between - 17% to 45%, as predetermined with the use of the increased strengths and dosages of cationic polyelectrolytes. The effluent had negligible amount of aluminium from operations, recorded at 0.16 mg/l, indicative of most of the aluminium in the process settled out. Formula determined concentration of aluminium in sludge to be 6999mg/l.