- Email: [email protected]
Minimizing solid wastes in an activated sludge system treating oil refinery wastewater
Accepted Manuscript Title: Minimizing solid wastes in an activated sludge system treating oil refinery wastewater Author: V.M.F. Alexandre T.M.S. de Castro L.V. de Ara´ujo V.M.J. Santiago A.C.F.P. de Cerqueira D.M.G. Freire M.C. Cammarota PII: DOI: Reference:
S0255-2701(15)30132-X http://dx.doi.org/doi:10.1016/j.cep.2015.10.021 CEP 6702
To appear in:
Chemical Engineering and Processing
Received date: Revised date: Accepted date:
23-6-2015 17-10-2015 27-10-2015
Please cite this article as: V.M.F.Alexandre, T.M.S.de Castro, L.V.de Ara´ujo, V.M.J.Santiago, A.C.F.P.de Cerqueira, D.M.G.Freire, M.C.Cammarota, Minimizing solid wastes in an activated sludge system treating oil refinery wastewater, Chemical Engineering and Processing http://dx.doi.org/10.1016/j.cep.2015.10.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Minimizing solid wastes in an activated sludge system treating oil refinery wastewater
V. M. F. ALEXANDREa, T. M. S. de CASTROa, L. V. de ARAÚJOb, V. M. J. SANTIAGOc, A. C. F. P. de CERQUEIRAc, D. M. G. FREIREb, M. C. CAMMAROTAa* [email protected] a Laboratory of Environmental Technology, School of Chemistry, Federal University of Rio de Janeiro, Brazil b Laboratory of Microbial Biotechnology, Institute of Chemistry, Federal University of Rio de Janeiro, Brazil c
Research Center of Petrobras, Brazil
*
Corresponding author: / Tel: +55 21 3938 7568.
Highlights
Proposal of a novel sludge reduction technique.
A new application for biosurfactants in wastewater treatment plants.
52% reduction in sludge disposal maintaining high COD removal.
Area occupied by secondary clarifier can be reduced by 39 to 52%.
Abstract Biosurfactants are suitable for application in wastewater treatment systems due to their biodegradability, biocompatibility and low toxicity. In activated sludge systems, they reduce coalescence and disintegrate flakes, enabling more cells to have access to oxygen. At low concentrations, they may act as growth inhibitors. In this study, rhamnolipid was added to a bench scale sequential batch reactor operating in similar conditions as oil refinery wastewater treatment plants. Concentrations from 12 to 50 mg rhamnolipid/L were evaluated, the latter being the minimum concentration necessary to reduce sludge disposal. In this concentration, rhamnolipid reduces sludge disposal of up to 52%, maintaining COD removal of 81-97% and good sludge settling properties (SVI 120 mL/g) and could also reduce area occupied by secondary clarifier of 39 to 52%. However, biosurfactant application needs to be optimized, because its cost is even higher than the savings obtained with lower waste disposal.
Keywords: activated sludge; biosurfactant; disposal; minimization; oil refinery wastewater; solid waste.
1. Introduction Activated sludge is the system most widely used in the treatment of domestic sewage and industrial effluents, including those generated in oil refineries. However, despite its high organic matter removal efficiency, it has high sludge production, which, after suitable treatment, is often referred to final disposal in landfills [1, 2]. This practice increases the costs for the treatment plant because sludge management can reach 60% of total operation costs, even though its volume accounts for only 1-2% of the total volume of treated effluent [2]. The activated sludge system of the oil refinery in this study generates about 11 daily tones (dry basis) of sludge with a daily disposal cost of USD 770. In Brazil, a new environmental law on solid waste management was recently implemented (Law Nº 12,305, August 2, 2010), establishing a National Policy on Solid Wastes, prioritizing the prevention of waste generation, followed by reduction, reuse, treatment and final disposal. The implementation of this law in the coming years is expected to further aggravate the problem of generation and disposal of sludge in landfills because it adds an environmental aspect to sludge management costs. Some modifications may be implemented in activated sludge systems to reduce sludge generation such as extended aeration and operation with reduced food/microorganism ratio (F/M ratio) [3, 4]. Most Brazilian oil refineries have adopted these changes and are already producing less sludge. Still, the amount of sludge generated is considerably high due to the high water consumption and, consequently, high generation of effluents [5]. According to Mahmood and Elliot [6], excess sludge reduction approaches fit into two categories: (a) post treatment of the sludge generated in order to reduce the amount for disposal and (b) changes in the wastewater treatment unit, so that sludge is produced in lower quantities. The second category, i.e., preventing the generation of biomass, can be regarded as a better choice in environmental terms [7]. In terms of integration with wastewater handling units, the existing mechanisms aiming sludge reduction are cell lysis/cryptic growth (most studied technique), uncoupled metabolism, endogenous metabolism and microbial predation, and each one of them consists of a set of technologies. Several technologies have been studied and even implemented on a pilot and industrial scale, although some studies are still restricted to the laboratory scale [2, 8].
As the problem of sludge generation is a subject that is in evidence all over the world, the search for new technologies to reduce excess sludge production is still necessary. The application of biosurfactants to reduce sludge production is one of these new alternative, still little discussed in literature, and should be studied in oil refineries. Biosurfactants are molecules with chemical properties similar to surfactants, frequently obtained by microbial means. The interest in these substances has increased mainly because they are considered environmentally compatible, since they have low toxicity and are biodegradable [9, 10]. In wastewater treatment, biosurfactant can be used to reduce coalescence, disintegrate biological flakes and allow more cells to have access to oxygen in aerobic biological processes, thus improving the treatment efficiency [11]. However, depending on the concentration, biosurfactants can inhibit cell growth or act as biocides [12]. Studies have shown that rhamnolipid-type biosurfactants have biocidal and inhibitory effect on algae and may even affect cell organelles [13]. The effect of biosurfactants as microbial growth inhibitor or biocide has been very little studied and there are no reports in literature about their use in wastewater treatment systems in order to reduce sludge generation by changing microbial metabolism. The aim of this study was to evaluate the sludge disposal reduction in the treatment of oil refinery effluent by activated sludge operating in sequential batch with and without addition of biosurfactant. The study also aimed to verify whether this alternative has potential for application in reducing one of the greatest environmental and economic problems today in treatment plants of oil refineries and other types of industry. The reduction in sludge generation allows processes intensification as it offers opportunities for use of equipment already known but in reduced size (in the case of sludge sedimentation tanks) and environmental benefits (lower emissions of pollutants during transport and lower amount disposed in landfills).
2. Material and methods
2.1. Origin of effluent and sludge The bioreactor operation used a mixture of effluents from an oil refinery, collected and stored at room temperature until time of use, and a synthetic medium containing substances typically found in oil refinery wastewaters, whose composition is based on the study of Brookes [14]. The NaCl concentration in the medium was adjusted to salinity value close to that found in Brazilian oil refineries (600 mg Chloride/L).
The sludge used as inoculum in bioreactors was obtained from the activated sludge system of an oil refinery, being collected and stored at 4°C until time of use and characterized as mass of volatile solids by waste mass (74 mg VS/g wet weight).
2.2. Biosurfactant production The production of the rhamnolipid-type biosurfactant was performed according to method described by Santos et al. [15], using Pseudomonas aeruginosa PA1, a strain previously isolated from oil wells [16], preserved in ultrafreezer (-80ºC) with glycerol 10% (w/v). The pre-inoculum (1 g biomass/L) was cultivated in a rotary shaker at 30ºC and 170 rpm for 40-44 h in medium with the following composition (g/L): NaNO3 1.0; KH2PO4 3.0; K2HPO4 7.0; MgSO4.7H2O 0.2; yeast extract 5.0; peptone 5.0 and glycerol 30.0. At the end of this period, cells were recovered by centrifugation (5000 g for 15 min) and used as inoculum (1 g/L) in 1L erlenmeyer flasks containing 500 mL of medium with the following composition (g/L): NaNO3 1.4; KH2PO4 3.0; K2HPO4 7.0; MgSO4.7H2O 0.2 and glycerol 30.0. Fermentation was conducted at 30ºC for 168 h. The culture medium derived from fermentation containing rhamnolipids was centrifuged to remove cells (5000 g), autoclaved (121ºC/15 min) and maintained under refrigeration (4ºC). The cell-free broth used in this study contained 7.9 – 10.2 g/L of rhamnolipids (crude rhamnolipid solution). After purification, a 0.75% (w/v) aqueous solution was characterized according to methodologies described in Araújo et al. [17], with surface tension of 29.8 mN/m and critical micelle concentration (CMC) of 67 mg/L.
2.3. Feeding of bioreactors The feeding of bioreactors consisted of a mixture of oil refinery wastewaters and synthetic medium calculated to obtain COD around 1000 mg/L. To maintain an ideal COD: N: P ratio of 100: 5: 1, there was need for supplementation with 2 mL solution composed of (g/L) Na2HPO4.12H2O 40.42 e KH2PO4 6.58 (mixture of effluents had sufficient nitrogen concentration) per L of feed. In order to improve the sludge settleability, 5 mL of FeCl3.6H2O 6.66 and Al2(SO4)3.18H2O 4.0 solutions (g/L) were added per L of feed, as recommended by Novak et al. [18]. The feed pH was adjusted to average values of 7.2 ± 0.2.
2.4. Operation of bioreactors
Two 1-L bioreactors with 500 mL working volume were used: Control (without addition of biosurfactant) and Test (with addition of biosurfactant). To ensure adequate oxygen supply, compressed air was injected through a porous diffuser located at the bottom of bioreactors. Aeration along with magnetic stirring allowed maintaining the sludge in suspension and the supply of dissolved oxygen necessary to the process. Bioreactors started with the addition of 20 g of sludge used as inoculum and feeding to complete the working volume of the reactor (500 mL) to an initial VSS (volatile suspended solids) concentration of about 3000 mg/L. Control and Test bioreactors were monitored for a total period of 266 days in sequencing batch, simulating operating conditions of oil refinery continuous reactors: 5.5 h reaction time, sludge age of 20 d, volumetric organic load (VOL) of 1.2 kg COD/m3.d and F/M of 0.4 kg consumed COD/kg VSS. Two daily medium exchanges were made. In the first exchange, aeration and agitation were shut down, sludge was left to sediment for 30 min and half the supernatant (130 mL) was replaced by a new feed with COD of 1000 mg/L, so as to simulate a recycle ratio equal to 1 (one). In the second exchange, 5.5 h after the 1st medium exchange, an aliquot of 25 mL of the mixed liquor was collected to maintain the sludge age at 20 d and again aeration and agitation were shut down. After sedimentation for 30 min, the supernatant (260 mL) was replaced by new feeding to maintain the biomass until the next day. The mixed liquor aliquot collected was analyzed for volume of settled sludge, concentration of VSS and TSS (total suspended solids), pH and centrifuged COD. Initially, the biosurfactant was added to the Test bioreactor feed to obtain the desired concentration. As this operation mode does not take into account the product biodegradation, it began to be added directly to the Test bioreactor in two daily medium exchanges, so that the entire reactor content presented the desired rhamnolipid concentration, which ranged from 12 to 50 mg/L.
2.5. Analytical methods To determine the sludge volume index (SVI), a modification of the standard method was used with sedimentation of 12.5 mL of mixed liquor in 25 mL measuring cylinder for 40 min. The other parameters (TSS, VSS, pH and centrifuged COD) were determined according to standard methods [19]. The operation of bioreactors was divided into periods based on the rhamnolipid concentration used and the results for each period were reported as mean ± standard deviation of 8 (period with 45 mg/L of rhamnolipid into the reactor), 11 (period with 40 mg/L of rhamnolipid into the reactor), 14 (period with 12 mg/L of rhamnolipid in the feed
and periods with 24 and 50 mg/L of rhamnolipid into the reactor), 16 (last period with 50 mg/L of rhamnolipid into the reactor) or 25 (period with 12 mg/L of rhamnolipid into the reactor) data obtained after 5.5h of reaction. For each period, the significance of results was evaluated by the Student's t test with 95% confidence level using the Statistica 7.0 software.
3. Results and discussion After an initial period of 15 days, in which both bioreactors received the same feed (no rhamnolipid) for adapting the sludge to the medium constituents, the Control bioreactor remained with feeding consisting of only a mixture of oil refinery effluents and the synthetic medium, while the Test bioreactor operated under different feeding conditions, based on the production batch, concentration and biosurfactant addition method. The use of different biosurfactant batches, as well as the addition method (in the feed or directly into the reactor) did not result in significant differences in the operation of the Test bioreactor. However, the rhamnolipid concentration influenced the results. The final pH of batches (after 5.5 h of reaction time) was slightly higher in the Test bioreactor in some periods of operation, but remained at values close to 7.0-7.5 in both bioreactors, a value considered suitable for aerobic biological treatment. Figure 1A shows the results obtained in terms of COD removal throughout the operation period of Control and Test bioreactors. After the adaptation period (15 days), COD removal efficiencies remained above 85%, and no significant differences were observed in any of the conditions evaluated. Values ranging from 80.6% to 97.3% and 84.7% to 97.3% were obtained for Control and Test bioreactors, respectively. Therefore, the addition of rhamnolipid did not affect the organic matter removal efficiency in any concentration evaluated, which is a primary condition for the indication of this bioproduct as a reducing agent of sludge production in treatment systems. Changing the treatment system using any technology can influence the quality of the final effluent, so any change in plant operation must be done by ensuring that there will be no harm to treatment [8]. For the addition of rhamnolipids in activated sludge systems to actually be used to reduce biomass generation, it must be ensured that other factors, in addition to COD removal, are not adversely affected. For instance, it cannot affect nitrification and cannot contribute to the final toxicity of the effluent, two major aspects regarding oil refineries wastewater treatment. In the initial stage of the study, the main focus was to evaluate the possibility
of rhamnolipids application in order to reduce sludge production. Thus, other factors will be assessed in a further stage of the study. Literature has reported some sludge production reduction technologies that increase the quality of the final effluent. For example, in a treatment system carried out in three stages, anaerobic, anoxic and aerobic, sludge heat treatment resulted in 28% sludge disposal reduction through partial solubilization of cells and increased nitrogen removal because the solubilized cell material was used as a substrate in the anoxic treatment stage [20]. Figure 1B shows the biomass concentration (VSS) in bioreactors throughout operation. The addition of 12 mg/L rhamnolipid in the feeding or directly into the reactor had no effect on the VSS concentration, which remained virtually the same in both bioreactors. Average VSS concentrations of 2068 ± 241 mg/L and 2188 ± 279 mg/L were obtained in Control and Test bioreactors, respectively, with no significant difference in the Student t test. The increase of biosurfactant concentration in the Test reactor to 24 mg/L had an effect contrary to expectations, contributing to a greater biomass growth. At this concentration, average VSS concentrations of 2166 ± 190 mg/L and 2469 ± 287 mg/L were obtained in the Control and Test bioreactors, respectively, with significant difference in the Student t test. These values resulted in an 11% increase in the sludge disposal (measured in mg TSS/day) in the Test bioreactor compared to the Control bioreactor.
Microbial growth induction observed with 24 mg/L biosurfactant may be due to the increased uptake of typical compounds from oil refinery wastewater by microorganisms such as polyaromatic hydrocarbons, which in the presence of surfactants have increased solubility and bioavailability [21]. Sponza and Gök [22] obtained 18% increase in polyaromatic hydrocarbon removal efficiency with the addition of 15 mg/L rhamnolipid in the treatment of oil refinery effluents using activated sludge system, and the models proposed to simulate the removal of polyaromatic hydrocarbons indicated that 94% of substances were removed by biodegradation. Uysal and Türkman [21] also reported an increase in the concentration of solids after addition of biosurfactant in a treatment system for removal of 2,4-dichlorophenol. The expected reduction of VSS concentration was observed only with the addition of 50 mg/L biosurfactant directly into the reactor. In this operating condition, the VSS concentration in the Control bioreactor was 2828 ± 318 mg/L, while in the Test bioreactor, 1800 ± 451 mg/L. Considering the entire operation period under this
condition (from days 149 to 181), TSS disposal was 36% lower in the Test bioreactor (53 ± 13 mg TSS/day) compared to value observed in the Control bioreactor (83 ± 8 mg TSS/day). This difference, even considering the standard deviations observed, would be significant in the Student t test. However, Figure 1B shows that the reduction of solids did not occur early in the period, but after six batches with the addition of higher biosurfactant concentration. Previous experiments conducted with other bioproducts have shown the need for an extended exposure time so that the reducing effect could be observed [7]. After this initial period, a more pronounced effect of biosurfactant was observed, with average VSS concentration in the Control bioreactor of 3061 ± 189 mg/L and the Test bioreactor of 1435 ± 196 mg/L. Thus, considering only the period from days 163 to 181, TSS disposal was 52% lower in the Test bioreactor (43 ± 5 mg TSS/day) compared to value observed in the Control bioreactor (89 ± 5 mg TSS/day), which difference is quite expressive and significant in the Student t test. It is noteworthy that this result was observed with no damage to the COD removal in the same period (Figure 1A). One hypothesis that could explain the effect observed is the fact that rhamnolipid presents different degrees of antimicrobial activity against bacteria that compose the sludge. While some species are fully inhibited and disappear, others suffer no effect, and as they have access to the substrate due to the high availability of nutrients, they can keep the process efficiency. Araújo et al. [17] observed that crude rhaminolipid was able to inhibit 2.5 and 4.6% of planktonic cell growth for two Listeria monocytogenes strains. A 100% inhibition was obtained for both strains when purified rhamnolipid was added to the medium. A rhamnolipid mixture obtained from Pseudomonas aeruginosa AT10 had surface tension of 24 mN/m, interfacial tension of 1.31 mN/m and the CMC was 120 mg/L and showed inhibitory activity against various bacteria at concentrations of 4 to 32 mg/L [23]. Vasileva-Tonkova et al. [24] evaluated the effect of rhamnolipid produced by Pseudomonas fluorescens on bacterial strains and verified that depending on the concentration, biosurfactant has a neutral or detrimental effect on the growth and protein release of a Gram (+) Bacillus subtilis strain, while the growth and protein release of a Gram (-) Pseudomonas aeruginosa strain was not influenced by the presence of biosurfactant in the medium. The effect of rhamnolipid can be associated to an interaction with the microbial membrane, opening pores that interfere in the selective permeability and destroy organelles. The cell wall of bacteria may provide protection against this effect, as suggested by Wang et al. [13]; however, the diversity of bacterial cell surface structures
may result in different effects. The glycolipid produced by Pseudozyma fusiformata yeast enhanced nonspecific permeability of the cytoplasmic membrane in sensitive cells of many yeasts and fungi, which resulted in ATP leakage. The minimal effective glycolipid concentration varied between 130 and 1600 mg/L to different yeasts and fungi [25]. Vasileva-Tonkova et al. [24] observed that bacterial cells treated with rhamnolipid became more or less hydrophobic than untreated cells depending on individual characteristics and abilities of the strains. For all treated strains, an increase in the amount of released protein was observed with increasing the amount of biosurfactant, probably due to increased cell permeability as a result of changes in the organization of cell surface structures. As the minimum concentration required to achieve reduction in sludge disposal could be in the range from 24 to 50 mg/L, the biosurfactant concentration was reduced to 40 mg/L. In operation period with 40 mg/L biosurfactant, average VSS concentrations of 2679 ± 205 mg/L and 1989 ± 312 mg/L were found for Control and Test bioreactors, respectively. The sludge disposal in the Test bioreactor (58 ± 8 mg TSS/day) was lower than in the Control bioreactor (78 ± 5 mg TSS/day), but with a difference between bioreactors of only 26%. Although this value has been shown to be statistically significant by the Student's t test (95% confidence level), in Figure 1B it is possible to observe the increase in VSS concentration in the Test bioreactor under this condition, demonstrating that the biosurfactant has lost effect. Lower values at the beginning of the period, resulting from the previous condition (50 mg/L) have led to lower average value. Within the same period, the VSS concentration in the Control bioreactor was kept constant (2679 ± 205 mg/L) while its concentration in the Test bioreactor increased at a rate of 32 mg VSS/L.d (R2 0.7355) increasing from 1592 to 2184 mg VSS/L. The increase in biosurfactant concentration to 45 mg/L in Test bioreactor resulted in VSS concentrations of 2882 ± 187 mg/L in the Control bioreactor and 2370 ± 490 mg/L in the Test bioreactor. The sludge disposal in the Test bioreactor (67 ± 14 mg TSS/day) was lower than in the Control bioreactor (85 ± 6 mg TSS/day), but with a difference between bioreactors of only 21%. Thus, the addition of 50, 40 and 45 mg/L showed 36 (maximum 52%), 26 and 21% reduction on sludge disposal, on average. As reductions achieved have been decreasing, it follows that the concentration of 50 mg/L would be the best condition for satisfactory reduction. The increase in the biosurfactant concentration again to 50 mg/L confirmed the results obtained in the previous period using the same concentration. After the first six batches, average VSS concentrations in the Control bioreactor of 2375 ± 118 mg/L and the Test bioreactor of 1424 ± 177 mg/L were found and the TSS
disposal was 42% lower in the Test bioreactor (43 ± 6 TSS mg/day) compared to that observed in the Control bioreactor (74 ± 6 TSS mg/day). As can be seen in Figure 1C, VSS/TSS ratios in both bioreactors showed very variable results, probably due to changes in the effluent composition at each feed to the reactor, which consisted of industrial effluent with high variability [7]. However, an increase in the average VSS/TSS ratio from 0.72 (1st period) to 0.83 (last period) is observed in both bioreactors over time. This increase can be attributed to the effluent composition, which has emulsified oils and greases that adsorb microbial flocks and are quantified as VSS. In all periods assessed, the VSS/TSS ratios were statistically equal in both bioreactors. It could be therefore inferred that the addition of biosurfactant does not interfere with this relationship. Another interesting result concerns the sludge sedimentation, characterized by SVI measurements (Figure 1D). While the Control bioreactor presented several periods of poor sedimentation (24, 40-50 mg/L rhamnolipid) with SVI values above 300 mL/g, in the Test bioreactor, the addition of rhamnolipid seems to have contributed to lower the SVI values and improve the sludge sedimentation. In the Test bioreactor, the SVI values remained below 120 mL/g, threshold considered adequate for good sludge sedimentation [4]. Combining reduction of the microbial growth with improved sludge sedimentation, another hypothesis suggested for the biosurfactant effect observed would be the metabolic uncoupling. The effect of a bioproduct based on chemical surfactants and stress proteins on the sludge reduction was assessed [7], reaching up to 46% reduction in sludge disposal. Although Podella et al. [26] point out stress proteins as primarily responsible for metabolic uncoupling, the similarity between the results suggests that (bio)surfactants may also act as uncoupling agents. Since there are organic matter consumption and energy deficit, or metabolic uncoupling, the production of extracellular polymeric substances (EPS) is minimized [27], thus contributing to improved sludge sedimentation. The same study (data not shown) showed lower EPS concentration in the sludge of the Test reactor and higher concentration in the Control reactor, which also showed higher SVI value. Regardless of the technique used, the response obtained in each case depends on factors such as scale, effluent used (real or synthetic, raw or after primary sedimentation, domestic sewage or industrial wastewater) and operation conditions (F/M, SRT, etc), making it difficult to compare results obtained in literature [8]. In addition, each technology also has specific characteristics, such as product concentration and type, in the case
of addition of metabolic uncouplers and temperature, contact time and treatment frequency in the case of thermal treatment [2]. The literature is quite scarce regarding the application of (bio)surfactants to reduce sludge production and only two studies have been found: surfactant application in the post-treatment of excess sludge [11] and evaluation of the effect of chemical surfactants on biological flakes, in which the authors also observed reduction of VSS in the reactor [28]. Stark and Kalos [11] applied surfactants to promote sludge disintegration, which may be followed or not by a mechanical process. The solubilized material was then aerobically stabilized, so that microbial growth was reduced. The authors claim that it is possible to reduce up to 80% the sludge disposal in activated sludge systems. Liwarska-Bizukojc and Bizukojc [28] evaluated the addition of three anionic chemical surfactants (sodium dodecyl sulfate, sodium dodecyl benzene sulphonate and sodium alkyltrioxyethylene sulphate) in activated sludge systems at 250 mg MBAS/L and obtained increase or decrease in the VSS concentration in the reactor, depending on the dilution rate applied. In higher HRT, chemical surfactants were consumed and served as a substrate for microorganisms, increasing the VSS concentration in the reactor up to 3x. In lower HRT, however, the VSS concentration reduced by up to 6x in test bioreactors, which contributed to the loss of system efficiency. The authors also observed lower SVI values in the test bioreactor during periods when reductions in the biomass concentration were observed. These results, consistent with those obtained in this study, demonstrate the potential application of (bio)surfactants aimed at reducing sludge production, but the mechanism of action and optimal conditions must still be further studied. Figure 2 shows the sludge disposal differences (mg TSS/day) between Control and Test bioreactors with respect to the rhamnolipid concentration. Positive differences (disposal in the Test bioreactor < disposal in the Control bioreactor) are obtained only with 45 - 50 mg/L rhamnolipid, and greater differences are observed in operation periods with 50 mg/L rhamnolipid in the reactor. As the biomass concentration in the Test bioreactor varied over time, a mass relationship between rhamnolipid and VSS was evaluated in the same period, verifying that the largest differences occurred with 19.1 ± 2.7 mg rhamnolipid/g VSS. The average sludge disposal in each operation condition is summarized in Figure 3a. There is a reduction in the concentration of solids and thus sludge disposal in both bioreactors as a function of the adjustment of biomass to the feeding constituents and operation conditions. However, from the 2nd operation period (12
mg/L of rhamnolipid in the Test bioreactor), it could be inferred that the average sludge disposal in the Control bioreactor is maintained at 75 ± 9 mg TSS/day, while in the Test bioreactor, the average sludge disposal is 66 ± 6 TSS mg/day to 45 mg/L rhamnolipid. Only with 50 mg/L of rhamnolipid, significantly lower disposal in the Test bioreactor is perceived (43 ± 5 mg TSS/day). In order to demonstrate that the reduction of sludge disposal is associated with lower biomass yield coefficient, Figure 3b shows average values of this parameter in each operation condition of bioreactors. In the initial period of operation without addition of rhamnolipid, the cell yield coefficient value (Yx/s, expressed as g VSS/g COD removed) in both bioreactors is low (0.12 ± 0.02), consistent with the sludge age of 20 days [29]. This average is maintained in the Control bioreactor throughout the operation period, while in the Test bioreactor only up to 45 mg/L (0.11 ± 0.01), decreasing with 50 mg/L rhamnolipid (0.07 ± 0.02). Based on these results, it was concluded that the minimum rhamnolipid concentration required to reduce growth was 50 mg/L or 19.1 mg rhamnolipid/g VSS under the operation conditions evaluated.
Within the context of process intensification, modification of the activated sludge system by the addition of biosurfactant can contribute on two aspects: reduction of the area occupied by the treatment system and reduction in sludge disposal costs, although reagent costs should be considered. The calculation of the required surface area is the main aspect in the design of a sedimentation tank. Surface overflow has been the design parameter for many years, however, currently, the solids loading rate is considered the limiting parameter that affects the effluent concentration. With a proper hydraulic design and management of solids, the overflow rate has little or no effect on the effluent quality over a wide range (up to 82 m3/m2.d). Therefore, the area is usually determined by considering the solids loading rate, and corresponds to the quotient between the applied solids load and the surface area of the sedimentation tank. Typical surface loading rates for extended aeration activated sludge systems are 1-5 kg TSS/m2.h [3]. The surface area of the sedimentation tank is calculated by sludge production (kg TSS/h) divided by the solids loading rate, as shown in Table 1. Therefore, any reduction in the amount of sludge generated directly influences the surface area of the sedimentation tank. In this particular case, a 51.7% reduction in the sludge production obtained by adding 50 mg/L of rhamnolipid would reduce the surface area of the sedimentation tank in the same value.
Another benefit resulting from the addition of rhamnolipid is the improvement of the sludge sedimentation, verified by SVI measurements. Von Sperling [4] proposes a method for calculating the surface area of secondary sedimentation tanks based on the sludge sedimentation speed, estimated from SVI value ranges. The sedimentation velocity (v) can be calculated from Equation 1,
v v0 . e k .C
(1)
where C is the influent TSS concentration to the sedimentation tank and v0 and k are coefficients obtained according to the sedimentation classification. In order not to lose solids in the effluent, the hydraulic loading rate (HLR) should not exceed the sludge settling velocity. So, the maximum HLR values required to meet the clarification criteria and the surface area required for secondary sedimentation tanks are given by Equations 2 and 3.
HLR v0 . e k .C
(2)
A Q / HLR
(3)
For an influent flow of 500 m3/h, the surface area required for the secondary sedimentation tank of the Test bioreactor was 38.5% lower compared to the Control bioreactor, based on the hydraulic load applied as shown in Table 1. In the activated sludge systems with extended aeration usually applied in oil refineries, the sludge disposed from secondary sedimentation tanks are already stabilized, but have solids content below 1%, requiring a sludge dewatering step to reduce its moisture content and thus the sludge volume for transportation and disposal in landfills. This dewatering step is usually carried out in centrifuges. The reduction in sludge production would also reduce the volume and area occupied by these sludge treatment units. For the application of any technology to be successful, it is essential to know all the variables that influence the process. In the case of biological systems such as activated sludge, long-term studies should also be carried out in order to evaluate the adaptation of biomass to system requirements. Therefore, direct comparison of results obtained by different research groups becomes extremely complicated, even using the same technology. So, sludge reductions to the same technology may vary: Foladori et al. [2], for example, reported results ranging from 30 to 100%.
In order to make a comparison on the same basis, Ginestet [8] standardized the results obtained in his survey to assess technical and economic aspects of several routes that alter the wastewater treatment plant in order to reduce sludge production (ozonation, peroxide oxidation, addition of metabolic uncoupler, microbial predation and thermal, electrical, mechanical and anaerobic treatments). The responses obtained depend on the configuration (raw effluent, effluent after primary sedimentation with or without sludge digestion) and plant size. The highest sludge reduction percentages were obtained with ozonation (70% reduction), and the smallest have been assigned to predation by protozoa and addition of p-nitrophenol (10% reduction). In terms of direct operational expenditures (energetic cost, manpower costs and maintenance costs), thermal and mechanical treatment, ozonization and microbial predation present costs similar to conventional sludge treatment, while electrical treatment, oxidation with peroxide and particularly addition of p-nitrophenol were especially costly. In terms of sludge disposal costs, a typical Brazilian oil refinery with production of 4,000 tons sludge (dry basis)/year and disposal cost of USD77/ton (dry basis), can obtain a saving of USD160,160/year with a 52% reduction in sludge production (conversion of Brazilian currency into American dollar held on the basis of the dollar exchange rate of September 2015). In order not to represent an additional cost in the treatment plant, the cost of the rhamnolipid in aeration tank should not exceed the savings with the lower amount of sludge disposed in the landfill. Considering a typical Brazilian refinery (500 m3/h), the biosurfactant cost should be less than USD 0.36/kg rhamnolipid. Currently, the market cost of rhamnolipid-type biosurfactants varies from USD 16/kg of product (50% pure, Victex Chemical Industries) to USD 1,250/kg of product (90% pure, Agae Technologies), much higher than that required to make its application economically viable in this case. However, there is no need to apply products with high purity and the sludge production reduction can also reduce the energy cost in the sludge dewatering steps. These factors must make the modified process more economical. The production cost of commercial biosurfactants is still much higher than the economy they provide due to high raw material costs, high processing costs and low manufacturing output. In order to make rhamnolipid production competitive in the surfactant market, it is essential to reduce production costs and substantially increase production rates [9]. Studies have shown that the use of inexpensive substrates, such as crude or
waste materials, dramatically affects the production costs of biosurfactants [30]. More studies on improving the biotechnological production of rhamnolipid using P. aeruginosa should contribute for reduction of production costs [9]. The proposed technology is not yet comparable to existing technologies in economic terms, but the results show potential for application. It is also worth mentioning that in Brazil, the cost of the disposal of sludge in landfills is still very low. While the cost of sludge disposal in Brazil is around USD 77/ton (dry basis), in Europe, this value can reach USD 357/ton (dry basis), according to Ginestet [8]. Thus, any technology aimed at reducing sludge production will find an economic bottleneck in the country, hindering its implementation. In addition to reducing production costs of biosurfactants, other strategies need to be evaluated such as application of other types of biosurfactants or mixtures of biosurfactants with reduced concentrations of individual surfactants, and also the intermittent addition of biosurfactant to reduce application costs. It should be considered that the continuous addition of product can reduce the concentration required to obtain the same disposal reduction percentage, which would also contribute to reduce costs. Although the biosurfactant application cost is still a limiting factor, the National Policy on Solid Wastes implemented in Brazil since 2010 imposes restrictions on the final disposal of biological sludge in sanitary and industrial landfills. Thus, investments in research in this field are expected to increase considerably in the coming years, which should help to reduce costs.
4. Conclusions The addition of 50 mg/L rhamnolipid or 19.1 mg rhamnolipid/g VSS in an activated sludge system operating in sequential batch to treat oil refinery effluent allowed reduction of up to 52% of sludge disposal, without harm to the COD removal efficiency or sludge settleability. Different lots of rhamnolipid produced by P. aeruginosa added at the same concentration yielded the same result, showing a further potential application of biosurfactants in the area of effluent treatment and waste minimization. The process modification by the addition of biossurfactant can contribute to processes intensification, reducing from 38.5 to 51.7% the area occupied by the secondary sedimentation tank and other sludge treatment units and sludge disposal costs. However, the cost of biosurfactant production must be reduced to enable the modification of the proposed process.
Acknowledgements The authors would like to thank Cenpes/Petrobras for the support provided to this research project (Cooperation Agreement Nº 0050.0064561.11.9), to researchers Fernanda Ribeiro do Carmo Damasceno and Antonio Carlos de Oliveira Machado for having contributed in the biosurfactant production and technician João Paulo Garuzi Luz Machado for having assisted in the conduction of experiments.
References [1] MA, H.; ZHANG, S.; LU, X.; XI, B.; GUO, X.; WANG, H.; DUAN, J. Excess sludge reduction using pilot-scale
lysis-cryptic
growth
system
integrated
ultrasonic/alkaline
disintegration
and
hydrolysis/acidogenesis pretreatment. Bioresource Technology, v. 116, p. 441-447, 2012. [2] FOLADORI, P.; ANDREOTTOLA, G.; ZIGLIO, G. Sludge reduction technologies in wastewater treatment plants. IWA Publishing, London, 2010. [3] METCALF & EDDY. Wastewater engineering: treatment and reuse. 4ª Ed., McGraw-Hill, 2003. [4] VON SPERLING, M. Activated sludge and aerobic biofilm reactors. 1st Ed., IWA Publishing, London, 2007. [5] POMBO, F. R.; MAGRINI, A.; SZKLO, A. Technology roadmap for wastewater reuse in petroleum refineries in Brazil. In: Elzbieta Broniewicz. (Org.). Environmental Management in Practice. InTech, v. 1, p. 425-448, 2011. [6] MAHMOOD, T.; ELLIOTT, A. A review of secondary sludge reduction technologies for the pulp and paper industry. Water Research, v. 40, p. 2093-112, 2006. [7] ALEXANDRE, V. M. F.; DE CERQUEIRA, A. C. F. P.; SANTIAGO, V. M. J.; CAMMAROTA, M. C. C. Bioproducts for sludge reduction in activated sludge systems treating oil refinery wastewater. Oil & Gas Science and Technology – Rev. IFP Energies Nouvelles, 2015. [8] GINESTET, P. Comparative evaluation of sludge reduction routes. IWA Publishing, 2007. [9] REIS, R. S.; PEREIRA, A. G.; NEVES, B. C.; FREIRE, D. M. G. Gene regulation of rhamnolipid production in Pseudomonas aeruginosa – A review. Bioresource Technology, v. 102, p. 6377–6384, 2011. [10] MAIER, R. M.; SOBERÓN-CHÁVEZ, G. Pseudomonas aeruginosa rhamnolipids: biosynthesis and potential applications. Applied Microbiology and Biotechnology, v. 54, p. 625-633, 2000. [11] STARK, O.; KALOS, G. Reduction of excess sludge by application of surface-active substances biochemical desintegration. Retrieved May, 2015, from: www.stark-consult.com. [12] ARAÚJO, L. V.; FREIRE, D. M. G.; NITSCHKE, M. Perspectives on using biosurfactants in food industry. In: NaimKosaric and Fazilet Vardar Sukan (Ed.). Biosurfactants: Production and Utilization— Processes, Technologies, and Economics. CRC Press, p. 295-312, 2014.
[13] WANG, X.; GONG, L.; LIANG, S.; HAN, X.; ZHU, C.; LI, Y. Algicidal activity of rhamnolipid biosurfactants produced by Pseudomonas aeruginosa. Harmful Algae, v. 4, p. 433-443, 2005. [14] BROOKES, A. Immersed membrane bioreactors for produced water treatment. Thesis (Ph.D.), Cranfield University, 2005. [15] SANTOS, A. S.; SAMPAIO, A. P. W.; VASQUEZ, P. S.; SANTA ANNA, L. M.; PEREIRA Jr, N.; FREIRE, D. M. G. Evaluation of different carbon and nitrogen sources in production of rhamnolipids by a strain of Pseudomonas aeruginosa. Applied Biochemistry and Biotechnology, v. 98-100, p. 1025-1035, 2002. [16] SANTA ANNA, L. M. M.; FREIRE, D. M. G.; SEBASTIAN, G. V.; MENEZES, E. P.; ROSADO, A. S.; ALVES, T. L. M.; PEREIRA JR, N. Production of biosurfactants from Pseudomonas aeruginosa PA1 isolated in oil environments. Brazilian Journal of Chemical Engineering, v. 19, p. 159-166, 2002. [17] ARAÚJO, L. V.; ABREU, F.; LINS, U.; SANTA ANNA, L. M. M.; NITSCHKE, M.; FREIRE, D. M. G. Rhamnolipid and surfactin inhibit Listeria monocytogenes adhesion. Food Research International, v. 44, p. 481-488, 2011. [18] NOVAK, J. T.; CHON, D. H.; CURTIS, B. A.; DOYLE, M. Biological solids reduction using the Cannibal process. Water Environment Research, v. 79, p. 2380-2386, 2007. [19] APHA, AWWA, WEF. Standard Methods for the Examination of Water & Wastewater. 21ª Ed., Washington, 2005. [20] RAJ, S. E.; BANU, J. R.; KALIAPPAN, S.; YEOM, I.; KUMAR, S. A. Effects of side-stream, low temperature phosphorus recovery on the performance of anaerobic/anoxic/oxic systems integrated with sludge pretreatment. Bioresource Technology, v. 140, p. 376-384, 2013. [21] UYSAL, A.; TÜRKMAN, A. Effect of biosurfactant on 2,4-dichlorophenol biodegradation in an activated sludge bioreactor. Process Biochemistry, v. 40, p. 2745-2749, 2005. [22] SPONZA, D. T.; GÖK, O. Effect of rhamnolipid on the aerobic removal of polyaromatic hydrocarbons (PAHs) and COD components from petrochemical wastewater. Bioresource Technology, v. 101, p. 914-924, 2010. [23] BENINCASA, M.; ABALOS, A.; OLIVEIRA, I.; MANRESA, A. Chemical structure, surface properties and biological activities of the biosurfactant produced by Pseudomonas aeruginosa LBI from soapstock. Antonie Van Leeuwenhoek,v. 85, p. 1-8, 2004.
[24] VASILEVA-TONKOVA, E.; SOTIROVA, A.; GALABOVA, D. The effect of rhamnolipid biosurfactant produced by Pseudomonas fluorescens on model bacterial strains and isolates from industrial wastewater. Current Microbiology, v. 62, p. 427-433, 2011. [25] KULAKOVSKAYA, T. V.; KULAKOVSKAYA, E. V.; GOLUBEV, W. I.ATP leakage from yeast cells treated by extracellular glycolipids of Pseudozyma fusiformata. FEMS Yeast Research, v. 3, p. 401-404, 2003. [26] PODELLA C.W., HOOSHNAM N., KRASSNER S.M., GOLDFELD M.G. Yeast protein-surfactant complexes uncouple microbial electron transfer and increase transmembrane leak of protons. Journal of Applied Microbiology, 106, 1, 140–148, 2008. [27] JIANG, B.; LIU, Y. Roles of ATP-dependent N-acylhomoserine lactones (AHLs) and extracellular polymeric substances (EPSs) in aerobic granulation. Chemosphere, v. 88, p. 1058–1064, 2012. [28] LIWARSKA-BIZUKOJC, E.; BIZUKOJC, M. Effect of selected anionic surfactants on activated sludge flocs. Enzyme and Microbial Technology, v. 39, p. 660–668, 2006. [29] GRADY, C. P. L.; DAIGGER, G. T.; LOVE, N. G.; FILIPE, C. D. M. Biological wastewater treatment. 3rd ed, CRC Press, USA, 2011. [30] HENKEL, M.; MÜLLER, M. M; KÜGLER, J. H.; LOVAGLIO, R. B.; CONTIERO, J.; SYLDATK, C. et al. Rhamnolipids as biosurfactants from renewable resources: Concepts for next-generation rhamnolipid production. Process Biochemistry, v. 47, p. 1207-1219, 2012.
Figure Captions
Figure 1. COD removal efficiency (A), biomass concentration (VSS) (B), VSS/TSS ratio in mixed liquor (C) and sludge volumetric index (SVI) (D) in Control and Test bioreactors under different rhamnolipid addition conditions.
Figure 2. Sludge disposal difference between bioreactors (Control - Test) and rhamnolipid concentration throughout the operation time.
Figure 3. Sludge Disposal (A) and yield coefficient (B) in Control and Test bioreactors under different rhamnolipid concentrations.
Table Table 1 – Example of sizing of the secondary sedimentation tank surface area for activated sludge systems with and without addition of biosurfactant. Parameters
Control
Test
Average influent flow (m3/h)a
500
500
Average return sludge flow (m3/h)b
500
500
3,558 ± 183
1,717 ± 203
TSS in the reactor (g/m3)c
Design based on solids loading rate Influent TSS load to the secondary sedimentation tank (kg/h)
3,558
1,717
3.0
3.0
1,186
572.3
Average solids loading rate (kg TSS/m2.h)d Required surface area (m2) Reduction of occupied area (%)
51.7 Design based on hydraulic loading rate
Range of sludge volume index values (mL/g)c
300-400
50-100
Very poor
Good
v0 (m/h)e
5.6
9.0
k (m3/kg)e
0.73
0.35
v (m/h)e
1.6
2.6
Hydraulic loading rate (m3/m2.h)e
1.6
2.6
312.5
192.3
Settleability
Required surfacearea (m2) Reduction of occupied area (%) a
38.5
Typical value of a large oil refinery, b recirculation ratio - r = 1, cvalues obtained in this study in period with
addition of 50 mg/L biosurfactant, d [3],e [4].
S0255-2701(15)30132-X http://dx.doi.org/doi:10.1016/j.cep.2015.10.021 CEP 6702
To appear in:
Chemical Engineering and Processing
Received date: Revised date: Accepted date:
23-6-2015 17-10-2015 27-10-2015
Please cite this article as: V.M.F.Alexandre, T.M.S.de Castro, L.V.de Ara´ujo, V.M.J.Santiago, A.C.F.P.de Cerqueira, D.M.G.Freire, M.C.Cammarota, Minimizing solid wastes in an activated sludge system treating oil refinery wastewater, Chemical Engineering and Processing http://dx.doi.org/10.1016/j.cep.2015.10.021 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Minimizing solid wastes in an activated sludge system treating oil refinery wastewater
V. M. F. ALEXANDREa, T. M. S. de CASTROa, L. V. de ARAÚJOb, V. M. J. SANTIAGOc, A. C. F. P. de CERQUEIRAc, D. M. G. FREIREb, M. C. CAMMAROTAa* [email protected] a Laboratory of Environmental Technology, School of Chemistry, Federal University of Rio de Janeiro, Brazil b Laboratory of Microbial Biotechnology, Institute of Chemistry, Federal University of Rio de Janeiro, Brazil c
Research Center of Petrobras, Brazil
*
Corresponding author: / Tel: +55 21 3938 7568.
Highlights
Proposal of a novel sludge reduction technique.
A new application for biosurfactants in wastewater treatment plants.
52% reduction in sludge disposal maintaining high COD removal.
Area occupied by secondary clarifier can be reduced by 39 to 52%.
Abstract Biosurfactants are suitable for application in wastewater treatment systems due to their biodegradability, biocompatibility and low toxicity. In activated sludge systems, they reduce coalescence and disintegrate flakes, enabling more cells to have access to oxygen. At low concentrations, they may act as growth inhibitors. In this study, rhamnolipid was added to a bench scale sequential batch reactor operating in similar conditions as oil refinery wastewater treatment plants. Concentrations from 12 to 50 mg rhamnolipid/L were evaluated, the latter being the minimum concentration necessary to reduce sludge disposal. In this concentration, rhamnolipid reduces sludge disposal of up to 52%, maintaining COD removal of 81-97% and good sludge settling properties (SVI 120 mL/g) and could also reduce area occupied by secondary clarifier of 39 to 52%. However, biosurfactant application needs to be optimized, because its cost is even higher than the savings obtained with lower waste disposal.
Keywords: activated sludge; biosurfactant; disposal; minimization; oil refinery wastewater; solid waste.
1. Introduction Activated sludge is the system most widely used in the treatment of domestic sewage and industrial effluents, including those generated in oil refineries. However, despite its high organic matter removal efficiency, it has high sludge production, which, after suitable treatment, is often referred to final disposal in landfills [1, 2]. This practice increases the costs for the treatment plant because sludge management can reach 60% of total operation costs, even though its volume accounts for only 1-2% of the total volume of treated effluent [2]. The activated sludge system of the oil refinery in this study generates about 11 daily tones (dry basis) of sludge with a daily disposal cost of USD 770. In Brazil, a new environmental law on solid waste management was recently implemented (Law Nº 12,305, August 2, 2010), establishing a National Policy on Solid Wastes, prioritizing the prevention of waste generation, followed by reduction, reuse, treatment and final disposal. The implementation of this law in the coming years is expected to further aggravate the problem of generation and disposal of sludge in landfills because it adds an environmental aspect to sludge management costs. Some modifications may be implemented in activated sludge systems to reduce sludge generation such as extended aeration and operation with reduced food/microorganism ratio (F/M ratio) [3, 4]. Most Brazilian oil refineries have adopted these changes and are already producing less sludge. Still, the amount of sludge generated is considerably high due to the high water consumption and, consequently, high generation of effluents [5]. According to Mahmood and Elliot [6], excess sludge reduction approaches fit into two categories: (a) post treatment of the sludge generated in order to reduce the amount for disposal and (b) changes in the wastewater treatment unit, so that sludge is produced in lower quantities. The second category, i.e., preventing the generation of biomass, can be regarded as a better choice in environmental terms [7]. In terms of integration with wastewater handling units, the existing mechanisms aiming sludge reduction are cell lysis/cryptic growth (most studied technique), uncoupled metabolism, endogenous metabolism and microbial predation, and each one of them consists of a set of technologies. Several technologies have been studied and even implemented on a pilot and industrial scale, although some studies are still restricted to the laboratory scale [2, 8].
As the problem of sludge generation is a subject that is in evidence all over the world, the search for new technologies to reduce excess sludge production is still necessary. The application of biosurfactants to reduce sludge production is one of these new alternative, still little discussed in literature, and should be studied in oil refineries. Biosurfactants are molecules with chemical properties similar to surfactants, frequently obtained by microbial means. The interest in these substances has increased mainly because they are considered environmentally compatible, since they have low toxicity and are biodegradable [9, 10]. In wastewater treatment, biosurfactant can be used to reduce coalescence, disintegrate biological flakes and allow more cells to have access to oxygen in aerobic biological processes, thus improving the treatment efficiency [11]. However, depending on the concentration, biosurfactants can inhibit cell growth or act as biocides [12]. Studies have shown that rhamnolipid-type biosurfactants have biocidal and inhibitory effect on algae and may even affect cell organelles [13]. The effect of biosurfactants as microbial growth inhibitor or biocide has been very little studied and there are no reports in literature about their use in wastewater treatment systems in order to reduce sludge generation by changing microbial metabolism. The aim of this study was to evaluate the sludge disposal reduction in the treatment of oil refinery effluent by activated sludge operating in sequential batch with and without addition of biosurfactant. The study also aimed to verify whether this alternative has potential for application in reducing one of the greatest environmental and economic problems today in treatment plants of oil refineries and other types of industry. The reduction in sludge generation allows processes intensification as it offers opportunities for use of equipment already known but in reduced size (in the case of sludge sedimentation tanks) and environmental benefits (lower emissions of pollutants during transport and lower amount disposed in landfills).
2. Material and methods
2.1. Origin of effluent and sludge The bioreactor operation used a mixture of effluents from an oil refinery, collected and stored at room temperature until time of use, and a synthetic medium containing substances typically found in oil refinery wastewaters, whose composition is based on the study of Brookes [14]. The NaCl concentration in the medium was adjusted to salinity value close to that found in Brazilian oil refineries (600 mg Chloride/L).
The sludge used as inoculum in bioreactors was obtained from the activated sludge system of an oil refinery, being collected and stored at 4°C until time of use and characterized as mass of volatile solids by waste mass (74 mg VS/g wet weight).
2.2. Biosurfactant production The production of the rhamnolipid-type biosurfactant was performed according to method described by Santos et al. [15], using Pseudomonas aeruginosa PA1, a strain previously isolated from oil wells [16], preserved in ultrafreezer (-80ºC) with glycerol 10% (w/v). The pre-inoculum (1 g biomass/L) was cultivated in a rotary shaker at 30ºC and 170 rpm for 40-44 h in medium with the following composition (g/L): NaNO3 1.0; KH2PO4 3.0; K2HPO4 7.0; MgSO4.7H2O 0.2; yeast extract 5.0; peptone 5.0 and glycerol 30.0. At the end of this period, cells were recovered by centrifugation (5000 g for 15 min) and used as inoculum (1 g/L) in 1L erlenmeyer flasks containing 500 mL of medium with the following composition (g/L): NaNO3 1.4; KH2PO4 3.0; K2HPO4 7.0; MgSO4.7H2O 0.2 and glycerol 30.0. Fermentation was conducted at 30ºC for 168 h. The culture medium derived from fermentation containing rhamnolipids was centrifuged to remove cells (5000 g), autoclaved (121ºC/15 min) and maintained under refrigeration (4ºC). The cell-free broth used in this study contained 7.9 – 10.2 g/L of rhamnolipids (crude rhamnolipid solution). After purification, a 0.75% (w/v) aqueous solution was characterized according to methodologies described in Araújo et al. [17], with surface tension of 29.8 mN/m and critical micelle concentration (CMC) of 67 mg/L.
2.3. Feeding of bioreactors The feeding of bioreactors consisted of a mixture of oil refinery wastewaters and synthetic medium calculated to obtain COD around 1000 mg/L. To maintain an ideal COD: N: P ratio of 100: 5: 1, there was need for supplementation with 2 mL solution composed of (g/L) Na2HPO4.12H2O 40.42 e KH2PO4 6.58 (mixture of effluents had sufficient nitrogen concentration) per L of feed. In order to improve the sludge settleability, 5 mL of FeCl3.6H2O 6.66 and Al2(SO4)3.18H2O 4.0 solutions (g/L) were added per L of feed, as recommended by Novak et al. [18]. The feed pH was adjusted to average values of 7.2 ± 0.2.
2.4. Operation of bioreactors
Two 1-L bioreactors with 500 mL working volume were used: Control (without addition of biosurfactant) and Test (with addition of biosurfactant). To ensure adequate oxygen supply, compressed air was injected through a porous diffuser located at the bottom of bioreactors. Aeration along with magnetic stirring allowed maintaining the sludge in suspension and the supply of dissolved oxygen necessary to the process. Bioreactors started with the addition of 20 g of sludge used as inoculum and feeding to complete the working volume of the reactor (500 mL) to an initial VSS (volatile suspended solids) concentration of about 3000 mg/L. Control and Test bioreactors were monitored for a total period of 266 days in sequencing batch, simulating operating conditions of oil refinery continuous reactors: 5.5 h reaction time, sludge age of 20 d, volumetric organic load (VOL) of 1.2 kg COD/m3.d and F/M of 0.4 kg consumed COD/kg VSS. Two daily medium exchanges were made. In the first exchange, aeration and agitation were shut down, sludge was left to sediment for 30 min and half the supernatant (130 mL) was replaced by a new feed with COD of 1000 mg/L, so as to simulate a recycle ratio equal to 1 (one). In the second exchange, 5.5 h after the 1st medium exchange, an aliquot of 25 mL of the mixed liquor was collected to maintain the sludge age at 20 d and again aeration and agitation were shut down. After sedimentation for 30 min, the supernatant (260 mL) was replaced by new feeding to maintain the biomass until the next day. The mixed liquor aliquot collected was analyzed for volume of settled sludge, concentration of VSS and TSS (total suspended solids), pH and centrifuged COD. Initially, the biosurfactant was added to the Test bioreactor feed to obtain the desired concentration. As this operation mode does not take into account the product biodegradation, it began to be added directly to the Test bioreactor in two daily medium exchanges, so that the entire reactor content presented the desired rhamnolipid concentration, which ranged from 12 to 50 mg/L.
2.5. Analytical methods To determine the sludge volume index (SVI), a modification of the standard method was used with sedimentation of 12.5 mL of mixed liquor in 25 mL measuring cylinder for 40 min. The other parameters (TSS, VSS, pH and centrifuged COD) were determined according to standard methods [19]. The operation of bioreactors was divided into periods based on the rhamnolipid concentration used and the results for each period were reported as mean ± standard deviation of 8 (period with 45 mg/L of rhamnolipid into the reactor), 11 (period with 40 mg/L of rhamnolipid into the reactor), 14 (period with 12 mg/L of rhamnolipid in the feed
and periods with 24 and 50 mg/L of rhamnolipid into the reactor), 16 (last period with 50 mg/L of rhamnolipid into the reactor) or 25 (period with 12 mg/L of rhamnolipid into the reactor) data obtained after 5.5h of reaction. For each period, the significance of results was evaluated by the Student's t test with 95% confidence level using the Statistica 7.0 software.
3. Results and discussion After an initial period of 15 days, in which both bioreactors received the same feed (no rhamnolipid) for adapting the sludge to the medium constituents, the Control bioreactor remained with feeding consisting of only a mixture of oil refinery effluents and the synthetic medium, while the Test bioreactor operated under different feeding conditions, based on the production batch, concentration and biosurfactant addition method. The use of different biosurfactant batches, as well as the addition method (in the feed or directly into the reactor) did not result in significant differences in the operation of the Test bioreactor. However, the rhamnolipid concentration influenced the results. The final pH of batches (after 5.5 h of reaction time) was slightly higher in the Test bioreactor in some periods of operation, but remained at values close to 7.0-7.5 in both bioreactors, a value considered suitable for aerobic biological treatment. Figure 1A shows the results obtained in terms of COD removal throughout the operation period of Control and Test bioreactors. After the adaptation period (15 days), COD removal efficiencies remained above 85%, and no significant differences were observed in any of the conditions evaluated. Values ranging from 80.6% to 97.3% and 84.7% to 97.3% were obtained for Control and Test bioreactors, respectively. Therefore, the addition of rhamnolipid did not affect the organic matter removal efficiency in any concentration evaluated, which is a primary condition for the indication of this bioproduct as a reducing agent of sludge production in treatment systems. Changing the treatment system using any technology can influence the quality of the final effluent, so any change in plant operation must be done by ensuring that there will be no harm to treatment [8]. For the addition of rhamnolipids in activated sludge systems to actually be used to reduce biomass generation, it must be ensured that other factors, in addition to COD removal, are not adversely affected. For instance, it cannot affect nitrification and cannot contribute to the final toxicity of the effluent, two major aspects regarding oil refineries wastewater treatment. In the initial stage of the study, the main focus was to evaluate the possibility
of rhamnolipids application in order to reduce sludge production. Thus, other factors will be assessed in a further stage of the study. Literature has reported some sludge production reduction technologies that increase the quality of the final effluent. For example, in a treatment system carried out in three stages, anaerobic, anoxic and aerobic, sludge heat treatment resulted in 28% sludge disposal reduction through partial solubilization of cells and increased nitrogen removal because the solubilized cell material was used as a substrate in the anoxic treatment stage [20]. Figure 1B shows the biomass concentration (VSS) in bioreactors throughout operation. The addition of 12 mg/L rhamnolipid in the feeding or directly into the reactor had no effect on the VSS concentration, which remained virtually the same in both bioreactors. Average VSS concentrations of 2068 ± 241 mg/L and 2188 ± 279 mg/L were obtained in Control and Test bioreactors, respectively, with no significant difference in the Student t test. The increase of biosurfactant concentration in the Test reactor to 24 mg/L had an effect contrary to expectations, contributing to a greater biomass growth. At this concentration, average VSS concentrations of 2166 ± 190 mg/L and 2469 ± 287 mg/L were obtained in the Control and Test bioreactors, respectively, with significant difference in the Student t test. These values resulted in an 11% increase in the sludge disposal (measured in mg TSS/day) in the Test bioreactor compared to the Control bioreactor.
Microbial growth induction observed with 24 mg/L biosurfactant may be due to the increased uptake of typical compounds from oil refinery wastewater by microorganisms such as polyaromatic hydrocarbons, which in the presence of surfactants have increased solubility and bioavailability [21]. Sponza and Gök [22] obtained 18% increase in polyaromatic hydrocarbon removal efficiency with the addition of 15 mg/L rhamnolipid in the treatment of oil refinery effluents using activated sludge system, and the models proposed to simulate the removal of polyaromatic hydrocarbons indicated that 94% of substances were removed by biodegradation. Uysal and Türkman [21] also reported an increase in the concentration of solids after addition of biosurfactant in a treatment system for removal of 2,4-dichlorophenol. The expected reduction of VSS concentration was observed only with the addition of 50 mg/L biosurfactant directly into the reactor. In this operating condition, the VSS concentration in the Control bioreactor was 2828 ± 318 mg/L, while in the Test bioreactor, 1800 ± 451 mg/L. Considering the entire operation period under this
condition (from days 149 to 181), TSS disposal was 36% lower in the Test bioreactor (53 ± 13 mg TSS/day) compared to value observed in the Control bioreactor (83 ± 8 mg TSS/day). This difference, even considering the standard deviations observed, would be significant in the Student t test. However, Figure 1B shows that the reduction of solids did not occur early in the period, but after six batches with the addition of higher biosurfactant concentration. Previous experiments conducted with other bioproducts have shown the need for an extended exposure time so that the reducing effect could be observed [7]. After this initial period, a more pronounced effect of biosurfactant was observed, with average VSS concentration in the Control bioreactor of 3061 ± 189 mg/L and the Test bioreactor of 1435 ± 196 mg/L. Thus, considering only the period from days 163 to 181, TSS disposal was 52% lower in the Test bioreactor (43 ± 5 mg TSS/day) compared to value observed in the Control bioreactor (89 ± 5 mg TSS/day), which difference is quite expressive and significant in the Student t test. It is noteworthy that this result was observed with no damage to the COD removal in the same period (Figure 1A). One hypothesis that could explain the effect observed is the fact that rhamnolipid presents different degrees of antimicrobial activity against bacteria that compose the sludge. While some species are fully inhibited and disappear, others suffer no effect, and as they have access to the substrate due to the high availability of nutrients, they can keep the process efficiency. Araújo et al. [17] observed that crude rhaminolipid was able to inhibit 2.5 and 4.6% of planktonic cell growth for two Listeria monocytogenes strains. A 100% inhibition was obtained for both strains when purified rhamnolipid was added to the medium. A rhamnolipid mixture obtained from Pseudomonas aeruginosa AT10 had surface tension of 24 mN/m, interfacial tension of 1.31 mN/m and the CMC was 120 mg/L and showed inhibitory activity against various bacteria at concentrations of 4 to 32 mg/L [23]. Vasileva-Tonkova et al. [24] evaluated the effect of rhamnolipid produced by Pseudomonas fluorescens on bacterial strains and verified that depending on the concentration, biosurfactant has a neutral or detrimental effect on the growth and protein release of a Gram (+) Bacillus subtilis strain, while the growth and protein release of a Gram (-) Pseudomonas aeruginosa strain was not influenced by the presence of biosurfactant in the medium. The effect of rhamnolipid can be associated to an interaction with the microbial membrane, opening pores that interfere in the selective permeability and destroy organelles. The cell wall of bacteria may provide protection against this effect, as suggested by Wang et al. [13]; however, the diversity of bacterial cell surface structures
may result in different effects. The glycolipid produced by Pseudozyma fusiformata yeast enhanced nonspecific permeability of the cytoplasmic membrane in sensitive cells of many yeasts and fungi, which resulted in ATP leakage. The minimal effective glycolipid concentration varied between 130 and 1600 mg/L to different yeasts and fungi [25]. Vasileva-Tonkova et al. [24] observed that bacterial cells treated with rhamnolipid became more or less hydrophobic than untreated cells depending on individual characteristics and abilities of the strains. For all treated strains, an increase in the amount of released protein was observed with increasing the amount of biosurfactant, probably due to increased cell permeability as a result of changes in the organization of cell surface structures. As the minimum concentration required to achieve reduction in sludge disposal could be in the range from 24 to 50 mg/L, the biosurfactant concentration was reduced to 40 mg/L. In operation period with 40 mg/L biosurfactant, average VSS concentrations of 2679 ± 205 mg/L and 1989 ± 312 mg/L were found for Control and Test bioreactors, respectively. The sludge disposal in the Test bioreactor (58 ± 8 mg TSS/day) was lower than in the Control bioreactor (78 ± 5 mg TSS/day), but with a difference between bioreactors of only 26%. Although this value has been shown to be statistically significant by the Student's t test (95% confidence level), in Figure 1B it is possible to observe the increase in VSS concentration in the Test bioreactor under this condition, demonstrating that the biosurfactant has lost effect. Lower values at the beginning of the period, resulting from the previous condition (50 mg/L) have led to lower average value. Within the same period, the VSS concentration in the Control bioreactor was kept constant (2679 ± 205 mg/L) while its concentration in the Test bioreactor increased at a rate of 32 mg VSS/L.d (R2 0.7355) increasing from 1592 to 2184 mg VSS/L. The increase in biosurfactant concentration to 45 mg/L in Test bioreactor resulted in VSS concentrations of 2882 ± 187 mg/L in the Control bioreactor and 2370 ± 490 mg/L in the Test bioreactor. The sludge disposal in the Test bioreactor (67 ± 14 mg TSS/day) was lower than in the Control bioreactor (85 ± 6 mg TSS/day), but with a difference between bioreactors of only 21%. Thus, the addition of 50, 40 and 45 mg/L showed 36 (maximum 52%), 26 and 21% reduction on sludge disposal, on average. As reductions achieved have been decreasing, it follows that the concentration of 50 mg/L would be the best condition for satisfactory reduction. The increase in the biosurfactant concentration again to 50 mg/L confirmed the results obtained in the previous period using the same concentration. After the first six batches, average VSS concentrations in the Control bioreactor of 2375 ± 118 mg/L and the Test bioreactor of 1424 ± 177 mg/L were found and the TSS
disposal was 42% lower in the Test bioreactor (43 ± 6 TSS mg/day) compared to that observed in the Control bioreactor (74 ± 6 TSS mg/day). As can be seen in Figure 1C, VSS/TSS ratios in both bioreactors showed very variable results, probably due to changes in the effluent composition at each feed to the reactor, which consisted of industrial effluent with high variability [7]. However, an increase in the average VSS/TSS ratio from 0.72 (1st period) to 0.83 (last period) is observed in both bioreactors over time. This increase can be attributed to the effluent composition, which has emulsified oils and greases that adsorb microbial flocks and are quantified as VSS. In all periods assessed, the VSS/TSS ratios were statistically equal in both bioreactors. It could be therefore inferred that the addition of biosurfactant does not interfere with this relationship. Another interesting result concerns the sludge sedimentation, characterized by SVI measurements (Figure 1D). While the Control bioreactor presented several periods of poor sedimentation (24, 40-50 mg/L rhamnolipid) with SVI values above 300 mL/g, in the Test bioreactor, the addition of rhamnolipid seems to have contributed to lower the SVI values and improve the sludge sedimentation. In the Test bioreactor, the SVI values remained below 120 mL/g, threshold considered adequate for good sludge sedimentation [4]. Combining reduction of the microbial growth with improved sludge sedimentation, another hypothesis suggested for the biosurfactant effect observed would be the metabolic uncoupling. The effect of a bioproduct based on chemical surfactants and stress proteins on the sludge reduction was assessed [7], reaching up to 46% reduction in sludge disposal. Although Podella et al. [26] point out stress proteins as primarily responsible for metabolic uncoupling, the similarity between the results suggests that (bio)surfactants may also act as uncoupling agents. Since there are organic matter consumption and energy deficit, or metabolic uncoupling, the production of extracellular polymeric substances (EPS) is minimized [27], thus contributing to improved sludge sedimentation. The same study (data not shown) showed lower EPS concentration in the sludge of the Test reactor and higher concentration in the Control reactor, which also showed higher SVI value. Regardless of the technique used, the response obtained in each case depends on factors such as scale, effluent used (real or synthetic, raw or after primary sedimentation, domestic sewage or industrial wastewater) and operation conditions (F/M, SRT, etc), making it difficult to compare results obtained in literature [8]. In addition, each technology also has specific characteristics, such as product concentration and type, in the case
of addition of metabolic uncouplers and temperature, contact time and treatment frequency in the case of thermal treatment [2]. The literature is quite scarce regarding the application of (bio)surfactants to reduce sludge production and only two studies have been found: surfactant application in the post-treatment of excess sludge [11] and evaluation of the effect of chemical surfactants on biological flakes, in which the authors also observed reduction of VSS in the reactor [28]. Stark and Kalos [11] applied surfactants to promote sludge disintegration, which may be followed or not by a mechanical process. The solubilized material was then aerobically stabilized, so that microbial growth was reduced. The authors claim that it is possible to reduce up to 80% the sludge disposal in activated sludge systems. Liwarska-Bizukojc and Bizukojc [28] evaluated the addition of three anionic chemical surfactants (sodium dodecyl sulfate, sodium dodecyl benzene sulphonate and sodium alkyltrioxyethylene sulphate) in activated sludge systems at 250 mg MBAS/L and obtained increase or decrease in the VSS concentration in the reactor, depending on the dilution rate applied. In higher HRT, chemical surfactants were consumed and served as a substrate for microorganisms, increasing the VSS concentration in the reactor up to 3x. In lower HRT, however, the VSS concentration reduced by up to 6x in test bioreactors, which contributed to the loss of system efficiency. The authors also observed lower SVI values in the test bioreactor during periods when reductions in the biomass concentration were observed. These results, consistent with those obtained in this study, demonstrate the potential application of (bio)surfactants aimed at reducing sludge production, but the mechanism of action and optimal conditions must still be further studied. Figure 2 shows the sludge disposal differences (mg TSS/day) between Control and Test bioreactors with respect to the rhamnolipid concentration. Positive differences (disposal in the Test bioreactor < disposal in the Control bioreactor) are obtained only with 45 - 50 mg/L rhamnolipid, and greater differences are observed in operation periods with 50 mg/L rhamnolipid in the reactor. As the biomass concentration in the Test bioreactor varied over time, a mass relationship between rhamnolipid and VSS was evaluated in the same period, verifying that the largest differences occurred with 19.1 ± 2.7 mg rhamnolipid/g VSS. The average sludge disposal in each operation condition is summarized in Figure 3a. There is a reduction in the concentration of solids and thus sludge disposal in both bioreactors as a function of the adjustment of biomass to the feeding constituents and operation conditions. However, from the 2nd operation period (12
mg/L of rhamnolipid in the Test bioreactor), it could be inferred that the average sludge disposal in the Control bioreactor is maintained at 75 ± 9 mg TSS/day, while in the Test bioreactor, the average sludge disposal is 66 ± 6 TSS mg/day to 45 mg/L rhamnolipid. Only with 50 mg/L of rhamnolipid, significantly lower disposal in the Test bioreactor is perceived (43 ± 5 mg TSS/day). In order to demonstrate that the reduction of sludge disposal is associated with lower biomass yield coefficient, Figure 3b shows average values of this parameter in each operation condition of bioreactors. In the initial period of operation without addition of rhamnolipid, the cell yield coefficient value (Yx/s, expressed as g VSS/g COD removed) in both bioreactors is low (0.12 ± 0.02), consistent with the sludge age of 20 days [29]. This average is maintained in the Control bioreactor throughout the operation period, while in the Test bioreactor only up to 45 mg/L (0.11 ± 0.01), decreasing with 50 mg/L rhamnolipid (0.07 ± 0.02). Based on these results, it was concluded that the minimum rhamnolipid concentration required to reduce growth was 50 mg/L or 19.1 mg rhamnolipid/g VSS under the operation conditions evaluated.
Within the context of process intensification, modification of the activated sludge system by the addition of biosurfactant can contribute on two aspects: reduction of the area occupied by the treatment system and reduction in sludge disposal costs, although reagent costs should be considered. The calculation of the required surface area is the main aspect in the design of a sedimentation tank. Surface overflow has been the design parameter for many years, however, currently, the solids loading rate is considered the limiting parameter that affects the effluent concentration. With a proper hydraulic design and management of solids, the overflow rate has little or no effect on the effluent quality over a wide range (up to 82 m3/m2.d). Therefore, the area is usually determined by considering the solids loading rate, and corresponds to the quotient between the applied solids load and the surface area of the sedimentation tank. Typical surface loading rates for extended aeration activated sludge systems are 1-5 kg TSS/m2.h [3]. The surface area of the sedimentation tank is calculated by sludge production (kg TSS/h) divided by the solids loading rate, as shown in Table 1. Therefore, any reduction in the amount of sludge generated directly influences the surface area of the sedimentation tank. In this particular case, a 51.7% reduction in the sludge production obtained by adding 50 mg/L of rhamnolipid would reduce the surface area of the sedimentation tank in the same value.
Another benefit resulting from the addition of rhamnolipid is the improvement of the sludge sedimentation, verified by SVI measurements. Von Sperling [4] proposes a method for calculating the surface area of secondary sedimentation tanks based on the sludge sedimentation speed, estimated from SVI value ranges. The sedimentation velocity (v) can be calculated from Equation 1,
v v0 . e k .C
(1)
where C is the influent TSS concentration to the sedimentation tank and v0 and k are coefficients obtained according to the sedimentation classification. In order not to lose solids in the effluent, the hydraulic loading rate (HLR) should not exceed the sludge settling velocity. So, the maximum HLR values required to meet the clarification criteria and the surface area required for secondary sedimentation tanks are given by Equations 2 and 3.
HLR v0 . e k .C
(2)
A Q / HLR
(3)
For an influent flow of 500 m3/h, the surface area required for the secondary sedimentation tank of the Test bioreactor was 38.5% lower compared to the Control bioreactor, based on the hydraulic load applied as shown in Table 1. In the activated sludge systems with extended aeration usually applied in oil refineries, the sludge disposed from secondary sedimentation tanks are already stabilized, but have solids content below 1%, requiring a sludge dewatering step to reduce its moisture content and thus the sludge volume for transportation and disposal in landfills. This dewatering step is usually carried out in centrifuges. The reduction in sludge production would also reduce the volume and area occupied by these sludge treatment units. For the application of any technology to be successful, it is essential to know all the variables that influence the process. In the case of biological systems such as activated sludge, long-term studies should also be carried out in order to evaluate the adaptation of biomass to system requirements. Therefore, direct comparison of results obtained by different research groups becomes extremely complicated, even using the same technology. So, sludge reductions to the same technology may vary: Foladori et al. [2], for example, reported results ranging from 30 to 100%.
In order to make a comparison on the same basis, Ginestet [8] standardized the results obtained in his survey to assess technical and economic aspects of several routes that alter the wastewater treatment plant in order to reduce sludge production (ozonation, peroxide oxidation, addition of metabolic uncoupler, microbial predation and thermal, electrical, mechanical and anaerobic treatments). The responses obtained depend on the configuration (raw effluent, effluent after primary sedimentation with or without sludge digestion) and plant size. The highest sludge reduction percentages were obtained with ozonation (70% reduction), and the smallest have been assigned to predation by protozoa and addition of p-nitrophenol (10% reduction). In terms of direct operational expenditures (energetic cost, manpower costs and maintenance costs), thermal and mechanical treatment, ozonization and microbial predation present costs similar to conventional sludge treatment, while electrical treatment, oxidation with peroxide and particularly addition of p-nitrophenol were especially costly. In terms of sludge disposal costs, a typical Brazilian oil refinery with production of 4,000 tons sludge (dry basis)/year and disposal cost of USD77/ton (dry basis), can obtain a saving of USD160,160/year with a 52% reduction in sludge production (conversion of Brazilian currency into American dollar held on the basis of the dollar exchange rate of September 2015). In order not to represent an additional cost in the treatment plant, the cost of the rhamnolipid in aeration tank should not exceed the savings with the lower amount of sludge disposed in the landfill. Considering a typical Brazilian refinery (500 m3/h), the biosurfactant cost should be less than USD 0.36/kg rhamnolipid. Currently, the market cost of rhamnolipid-type biosurfactants varies from USD 16/kg of product (50% pure, Victex Chemical Industries) to USD 1,250/kg of product (90% pure, Agae Technologies), much higher than that required to make its application economically viable in this case. However, there is no need to apply products with high purity and the sludge production reduction can also reduce the energy cost in the sludge dewatering steps. These factors must make the modified process more economical. The production cost of commercial biosurfactants is still much higher than the economy they provide due to high raw material costs, high processing costs and low manufacturing output. In order to make rhamnolipid production competitive in the surfactant market, it is essential to reduce production costs and substantially increase production rates [9]. Studies have shown that the use of inexpensive substrates, such as crude or
waste materials, dramatically affects the production costs of biosurfactants [30]. More studies on improving the biotechnological production of rhamnolipid using P. aeruginosa should contribute for reduction of production costs [9]. The proposed technology is not yet comparable to existing technologies in economic terms, but the results show potential for application. It is also worth mentioning that in Brazil, the cost of the disposal of sludge in landfills is still very low. While the cost of sludge disposal in Brazil is around USD 77/ton (dry basis), in Europe, this value can reach USD 357/ton (dry basis), according to Ginestet [8]. Thus, any technology aimed at reducing sludge production will find an economic bottleneck in the country, hindering its implementation. In addition to reducing production costs of biosurfactants, other strategies need to be evaluated such as application of other types of biosurfactants or mixtures of biosurfactants with reduced concentrations of individual surfactants, and also the intermittent addition of biosurfactant to reduce application costs. It should be considered that the continuous addition of product can reduce the concentration required to obtain the same disposal reduction percentage, which would also contribute to reduce costs. Although the biosurfactant application cost is still a limiting factor, the National Policy on Solid Wastes implemented in Brazil since 2010 imposes restrictions on the final disposal of biological sludge in sanitary and industrial landfills. Thus, investments in research in this field are expected to increase considerably in the coming years, which should help to reduce costs.
4. Conclusions The addition of 50 mg/L rhamnolipid or 19.1 mg rhamnolipid/g VSS in an activated sludge system operating in sequential batch to treat oil refinery effluent allowed reduction of up to 52% of sludge disposal, without harm to the COD removal efficiency or sludge settleability. Different lots of rhamnolipid produced by P. aeruginosa added at the same concentration yielded the same result, showing a further potential application of biosurfactants in the area of effluent treatment and waste minimization. The process modification by the addition of biossurfactant can contribute to processes intensification, reducing from 38.5 to 51.7% the area occupied by the secondary sedimentation tank and other sludge treatment units and sludge disposal costs. However, the cost of biosurfactant production must be reduced to enable the modification of the proposed process.
Acknowledgements The authors would like to thank Cenpes/Petrobras for the support provided to this research project (Cooperation Agreement Nº 0050.0064561.11.9), to researchers Fernanda Ribeiro do Carmo Damasceno and Antonio Carlos de Oliveira Machado for having contributed in the biosurfactant production and technician João Paulo Garuzi Luz Machado for having assisted in the conduction of experiments.
References [1] MA, H.; ZHANG, S.; LU, X.; XI, B.; GUO, X.; WANG, H.; DUAN, J. Excess sludge reduction using pilot-scale
lysis-cryptic
growth
system
integrated
ultrasonic/alkaline
disintegration
and
hydrolysis/acidogenesis pretreatment. Bioresource Technology, v. 116, p. 441-447, 2012. [2] FOLADORI, P.; ANDREOTTOLA, G.; ZIGLIO, G. Sludge reduction technologies in wastewater treatment plants. IWA Publishing, London, 2010. [3] METCALF & EDDY. Wastewater engineering: treatment and reuse. 4ª Ed., McGraw-Hill, 2003. [4] VON SPERLING, M. Activated sludge and aerobic biofilm reactors. 1st Ed., IWA Publishing, London, 2007. [5] POMBO, F. R.; MAGRINI, A.; SZKLO, A. Technology roadmap for wastewater reuse in petroleum refineries in Brazil. In: Elzbieta Broniewicz. (Org.). Environmental Management in Practice. InTech, v. 1, p. 425-448, 2011. [6] MAHMOOD, T.; ELLIOTT, A. A review of secondary sludge reduction technologies for the pulp and paper industry. Water Research, v. 40, p. 2093-112, 2006. [7] ALEXANDRE, V. M. F.; DE CERQUEIRA, A. C. F. P.; SANTIAGO, V. M. J.; CAMMAROTA, M. C. C. Bioproducts for sludge reduction in activated sludge systems treating oil refinery wastewater. Oil & Gas Science and Technology – Rev. IFP Energies Nouvelles, 2015. [8] GINESTET, P. Comparative evaluation of sludge reduction routes. IWA Publishing, 2007. [9] REIS, R. S.; PEREIRA, A. G.; NEVES, B. C.; FREIRE, D. M. G. Gene regulation of rhamnolipid production in Pseudomonas aeruginosa – A review. Bioresource Technology, v. 102, p. 6377–6384, 2011. [10] MAIER, R. M.; SOBERÓN-CHÁVEZ, G. Pseudomonas aeruginosa rhamnolipids: biosynthesis and potential applications. Applied Microbiology and Biotechnology, v. 54, p. 625-633, 2000. [11] STARK, O.; KALOS, G. Reduction of excess sludge by application of surface-active substances biochemical desintegration. Retrieved May, 2015, from: www.stark-consult.com. [12] ARAÚJO, L. V.; FREIRE, D. M. G.; NITSCHKE, M. Perspectives on using biosurfactants in food industry. In: NaimKosaric and Fazilet Vardar Sukan (Ed.). Biosurfactants: Production and Utilization— Processes, Technologies, and Economics. CRC Press, p. 295-312, 2014.
[13] WANG, X.; GONG, L.; LIANG, S.; HAN, X.; ZHU, C.; LI, Y. Algicidal activity of rhamnolipid biosurfactants produced by Pseudomonas aeruginosa. Harmful Algae, v. 4, p. 433-443, 2005. [14] BROOKES, A. Immersed membrane bioreactors for produced water treatment. Thesis (Ph.D.), Cranfield University, 2005. [15] SANTOS, A. S.; SAMPAIO, A. P. W.; VASQUEZ, P. S.; SANTA ANNA, L. M.; PEREIRA Jr, N.; FREIRE, D. M. G. Evaluation of different carbon and nitrogen sources in production of rhamnolipids by a strain of Pseudomonas aeruginosa. Applied Biochemistry and Biotechnology, v. 98-100, p. 1025-1035, 2002. [16] SANTA ANNA, L. M. M.; FREIRE, D. M. G.; SEBASTIAN, G. V.; MENEZES, E. P.; ROSADO, A. S.; ALVES, T. L. M.; PEREIRA JR, N. Production of biosurfactants from Pseudomonas aeruginosa PA1 isolated in oil environments. Brazilian Journal of Chemical Engineering, v. 19, p. 159-166, 2002. [17] ARAÚJO, L. V.; ABREU, F.; LINS, U.; SANTA ANNA, L. M. M.; NITSCHKE, M.; FREIRE, D. M. G. Rhamnolipid and surfactin inhibit Listeria monocytogenes adhesion. Food Research International, v. 44, p. 481-488, 2011. [18] NOVAK, J. T.; CHON, D. H.; CURTIS, B. A.; DOYLE, M. Biological solids reduction using the Cannibal process. Water Environment Research, v. 79, p. 2380-2386, 2007. [19] APHA, AWWA, WEF. Standard Methods for the Examination of Water & Wastewater. 21ª Ed., Washington, 2005. [20] RAJ, S. E.; BANU, J. R.; KALIAPPAN, S.; YEOM, I.; KUMAR, S. A. Effects of side-stream, low temperature phosphorus recovery on the performance of anaerobic/anoxic/oxic systems integrated with sludge pretreatment. Bioresource Technology, v. 140, p. 376-384, 2013. [21] UYSAL, A.; TÜRKMAN, A. Effect of biosurfactant on 2,4-dichlorophenol biodegradation in an activated sludge bioreactor. Process Biochemistry, v. 40, p. 2745-2749, 2005. [22] SPONZA, D. T.; GÖK, O. Effect of rhamnolipid on the aerobic removal of polyaromatic hydrocarbons (PAHs) and COD components from petrochemical wastewater. Bioresource Technology, v. 101, p. 914-924, 2010. [23] BENINCASA, M.; ABALOS, A.; OLIVEIRA, I.; MANRESA, A. Chemical structure, surface properties and biological activities of the biosurfactant produced by Pseudomonas aeruginosa LBI from soapstock. Antonie Van Leeuwenhoek,v. 85, p. 1-8, 2004.
[24] VASILEVA-TONKOVA, E.; SOTIROVA, A.; GALABOVA, D. The effect of rhamnolipid biosurfactant produced by Pseudomonas fluorescens on model bacterial strains and isolates from industrial wastewater. Current Microbiology, v. 62, p. 427-433, 2011. [25] KULAKOVSKAYA, T. V.; KULAKOVSKAYA, E. V.; GOLUBEV, W. I.ATP leakage from yeast cells treated by extracellular glycolipids of Pseudozyma fusiformata. FEMS Yeast Research, v. 3, p. 401-404, 2003. [26] PODELLA C.W., HOOSHNAM N., KRASSNER S.M., GOLDFELD M.G. Yeast protein-surfactant complexes uncouple microbial electron transfer and increase transmembrane leak of protons. Journal of Applied Microbiology, 106, 1, 140–148, 2008. [27] JIANG, B.; LIU, Y. Roles of ATP-dependent N-acylhomoserine lactones (AHLs) and extracellular polymeric substances (EPSs) in aerobic granulation. Chemosphere, v. 88, p. 1058–1064, 2012. [28] LIWARSKA-BIZUKOJC, E.; BIZUKOJC, M. Effect of selected anionic surfactants on activated sludge flocs. Enzyme and Microbial Technology, v. 39, p. 660–668, 2006. [29] GRADY, C. P. L.; DAIGGER, G. T.; LOVE, N. G.; FILIPE, C. D. M. Biological wastewater treatment. 3rd ed, CRC Press, USA, 2011. [30] HENKEL, M.; MÜLLER, M. M; KÜGLER, J. H.; LOVAGLIO, R. B.; CONTIERO, J.; SYLDATK, C. et al. Rhamnolipids as biosurfactants from renewable resources: Concepts for next-generation rhamnolipid production. Process Biochemistry, v. 47, p. 1207-1219, 2012.
Figure Captions
Figure 1. COD removal efficiency (A), biomass concentration (VSS) (B), VSS/TSS ratio in mixed liquor (C) and sludge volumetric index (SVI) (D) in Control and Test bioreactors under different rhamnolipid addition conditions.
Figure 2. Sludge disposal difference between bioreactors (Control - Test) and rhamnolipid concentration throughout the operation time.
Figure 3. Sludge Disposal (A) and yield coefficient (B) in Control and Test bioreactors under different rhamnolipid concentrations.
Table Table 1 – Example of sizing of the secondary sedimentation tank surface area for activated sludge systems with and without addition of biosurfactant. Parameters
Control
Test
Average influent flow (m3/h)a
500
500
Average return sludge flow (m3/h)b
500
500
3,558 ± 183
1,717 ± 203
TSS in the reactor (g/m3)c
Design based on solids loading rate Influent TSS load to the secondary sedimentation tank (kg/h)
3,558
1,717
3.0
3.0
1,186
572.3
Average solids loading rate (kg TSS/m2.h)d Required surface area (m2) Reduction of occupied area (%)
51.7 Design based on hydraulic loading rate
Range of sludge volume index values (mL/g)c
300-400
50-100
Very poor
Good
v0 (m/h)e
5.6
9.0
k (m3/kg)e
0.73
0.35
v (m/h)e
1.6
2.6
Hydraulic loading rate (m3/m2.h)e
1.6
2.6
312.5
192.3
Settleability
Required surfacearea (m2) Reduction of occupied area (%) a
38.5
Typical value of a large oil refinery, b recirculation ratio - r = 1, cvalues obtained in this study in period with
addition of 50 mg/L biosurfactant, d [3],e [4].