J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
Evaluation of an Activated Sludge Process Combined with Powdered Activated Carbon for the Treatment of Oil Refinery Wastewater J.C. Campos 1,, C.R.A. Machado1, J. M. S. Couto1, P.L.Florido2, A.C.F.P.Cerqueira2, V.M.J.Santiago2 1
School of Chemistry, Federal University of Rio de Janeiro, Av. Athos da Silveira Ramos, 149 room E-206, ZIP Code 21941-909, Rio de Janeiro, Brazil; 2Research Center of PETROBRAS, Av. Horácio Macedo, 950, Zip Code 21.941-915, Rio de Janeiro, Brazil Received April 26, 2013; Accepted March 25, 2014
Abstract: The work described herein is a study of the use of powdered activated carbon (PAC) in the activated sludge process (PACT™ - Powdered Activated Carbon Treatment) for the treatment of wastewater from oil refinery. In the first step, two carbon samples were chosen out of four different types of activated carbon based on the results of tests that evaluated the removal efficiency of recalcitrant organic substances. Next, experiments in a continuous bioreactor were conducted at laboratory scale, which simulated the sludge activated process. Two different operating conditions were tested by varying carbon replacement and sludge age. The best results were achieved by activated sludge and Norit (SAE Super 94009-7) PAC under the following reactor conditions: HRT, 24 h; SRT, 30 d; carbon replacement, 150 mgPACL-1affluent; and 4,5 gPAC.L-1reactor. Under these conditions, the removal efficiencies were 98% for COD and 99% for phenol, and there was an overall decrease in Ceriodaphnia dubia chronic toxicity. Additionally, the PACT process was more stable compared to the control reactor process (activated sludge without PAC). Keywords: Oil refinery wastewater; activated sludge; powdered activated carbon
Introduction In petroleum refineries, the combination of complex processes generates a complex mix of wastewaters. According to Lesage et al (2008), the main refinery processes that generate wastewater are storage, desalination, fractionation, thermal and catalytic cracking, reforming, polymerization, alkylation, isomerization and solvent refining. The wastewater composition may include aliphatic and aromatic hydrocarbons, phenolic compounds, sulfur, mercaptans, oil, solvents and chloride. Wastewater treatment using a conventional process such as activated sludge can be hampered by the presence of recalcitrant organic compounds; as a result, tertiary treatments for wastewater polishing are needed. The addition of powdered activated carbon (PAC) to the activated sludge process may result in a higher quality of treated wastewater, a more stable system and the generation of reusable water. The PACT™ process was developed by DuPont™ in the early '70s. This process combines the use of PAC with the activated sludge process, in which PAC is directly added to the aeration tank so that biological oxidation and physical adsorption occur simultaneously. The main feature of this process is that it can be integrated into existing activated sludge systems at relatively low costs (Eckenfelder, 2000). The addition of PAC confers several advantages to the process, including system stability during load shocks; reduction in priority refractory pollutants; removal of color, odor and ammonia; improvement in sludge settling; and reduction or elimination of biological inhibition (Eckenfelder, 2000; Sher et al, 2000; Metcalf & Eddy, 2003). These effects are the result of additional organic biodegradation due to decreased toxicity or inhibited adsorption onto PAC and the resulting degradation of slowly biodegradable substances due to the increased time of exposure between the adsorbed substances on the PAC and the biomass (Aktas & Çecen, 2007). These substances remain in the system during the sludge aging process (solid retention time), whereas in the absence of PAC, the substances would remain in the system only during the hydraulic retention time (HRT) (Eckenfelder, 2000; Metcalf & Eddy, 2003).
Corresponding: E-Mail:
[email protected]. Tel./Fax +55 21 25627640
24
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) Equation 1 illustrates how the main activated sludge operational parameters are related.
Xpac
Xci SRT , HRT
(1)
where Xpac is the accumulated concentration of PAC in the reactor (equilibrium powdered activated carbon – MLSS content), in mgPAC.L-1reactor; Xci is the PAC concentration dosage (the carbon replacement), in mgPAC.L-1affluent; SRT is the solid retention time (sludge age), in days; and HRT is the hydraulic retention time, in days. Meidl (1997) notes that, by merging biological and physical treatment into a single process step, the system is able to buffer toxic loads that might otherwise impair a straight biological system and reduce the amount of carbon needed by a straight adsorption treatment system. However, the author presents case studies using PACT™ in the treatment of petrochemical wastewater, landfill leachate and highly contaminated groundwater. Recent studies demonstrating PACT™ application have addressed the following topics: cheese whey wastewater (Orozco et al, 2010), hexavalent chromium (Orozco et al., 2011) and landfill leachate (Zhao et al., 2012). Lesage et al (2008) and Satyawali and Balakrishnan (2009) discuss the introduction of PAC in MBR (Membrane Bioreactor) systems. The authors note that the addition of PAC could structure the biomass, modify the cake deposit on the membrane and reduce membrane fouling by decreasing protein and carbohydrate concentrations. This study evaluated the efficacy of combining the activated sludge process with PAC in the treatment of oil refinery wastewater. The removal of organic substances and the acute and chronic toxicities of the treated wastewater were compared to those of wastewater treated using an activated sludge system without PAC.
Methods This study used wastewater from an oil refinery located in Rio de Janeiro State (RJ, Brazil). Three different sample types were used, as shown in Table 1. The wastewater samples were refrigerated (4 °C) before use. Table 2 presents the characterization of the samples collected for this work. All of the parameters were measured according to the methodology described in the Standard Method for the Examination of Water and Wastewater (APHA, AWWA, WEF, 2005), which will be described in the Analytical Methodology section. Table 1. Origin and use of the wastewaters in this study. Wastewater (Stream)
Industrial Process Source
“D”
Catalytic Cracking Unit and the Lubricant Distillation Tower Affluent to Biological Process (aerated lagoon) in Treatment Industrial Wastewater Treatment Station
“L”
Step of study which the stream was used -
“D” and “L” mixture
-
Continuous system experiments – Activated Sludge Process and PACT Process
“B”
Effluent to Biological Process in Treatment Industrial Wastewater Treatment Station (biotreated stream)
Adsorption experiments
The results obtained from the characterization of the wastewaters from the refinery revealed that the bio-treated wastewater is consistent with the required patterns of discharge into bodies of water, as specified by the INEA at the Rio de Janeiro State Institute of the Environment (NT202-R10, 1986 and DZ205-R6, 2007). The characterization values for streams D and L showed large variations, a common feature of industrial wastewaters.
25
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) Table 2. Characterization of samples used in this study. For stream B, n=2 samples; for streams D and L, n=10 samples. Parameters COD - Chemical Oxygen Demand (mg O2.L-1) TOC - Total Organic Carbon (mg C.L-1) Absorbance at 254 nm (cm-1) Phenol (mg Phenol.L-1) Total ammonia nitrogen (mg [N-NH3].L-1) Chloride (mg Cl-.L-1) Dissolved Reactive Phosphorus (mgP.L-1)
Stream B 80-90 0.36-0.39 3.0-3.5 -
Stream D 1100-1800 250-520 4.55-10.50 114-400 15-20 90-155 0.08-0.20
Stream L 230-920 75-200 1.20-8.30 13-22 2.0-8.5 15-35 0.60-1.40
Parameters (-) were not measured
PAC Choice Step To select the PAC to be used in the bioreactor, experiments measuring adsorption isotherms were performed according to the method established in ASTM 3860-98. Bio-treated wastewater from the Industrial Wastewater Treatment Station in the refinery, which contains an aerated lagoon as a biological treatment system, was used. The objective of this experiment was to evaluate the removal of organic substances, which are resistant to biological treatment, by PAC adsorption. In this step, absorbance at a wavelength of 254 nm was measured because it can detect aromatic organic substances (APHA, AWWA, WEF, 2005). The PAC samples tested are shown in Table 3, as are the source, brands and nomenclature used in this work. The adsorption tests were performed in triplicate. Table 3. Description of the PAC samples used in this work PAC A B C D
Source bituminous coal wood bituminous coal wood
Brand Norit (SAE Super 94009-7), The Netherlands Carbomafra (118CB AS), Brazil Crossfilter (PWI 125-7 – Calgon), USA Brasilac (kapa L), Brazil
The resulting absorbance values were used in the Freundlich isotherm equation, presented below, to determine the other parameters. qe = Kf Ce1/n, (2) where qe is expressed by adsorbate mass unit per adsorbent mass unit, Ce is the adsorbate equilibrium concentration and Kf, and n are experimental constants. Kf is related to the adsorption capacity of the adsorbent by the adsorbate, and n is related to the adsorption bonding strength (Eckenfelder, 1999). For selected PACs, a textural characterization was performed according to the parameters described in the Analytical Methodology section. Continuous system experiments – Activated Sludge and PACT A bioreactor was configured according to the Eckenfelder model, which is suitable for refractory wastewater (Eckenfelder, 2000). Figures 1 depicts the apparatus. The aeration chamber volume was 2 L, and the decanter volume was 1 L. Sludge was withdrawn daily to maintain the specified sludge age within the aeration tank. The sludge accumulated in the decanter was manually returned to the aeration tank each day. The PAC was replaced every day; the total carbon added in a day corresponded to the value specified by the operation stage. Phosphorus was adjusted to maintain the BOD:N:P (100:5:1) ratio, through the addition of sodium dibasic phosphate. Sodium carbonate was added to correct alkalinity to achieve nitrification conditions of 7.14 mg of CaCO 3 per mg of oxidized N (Eckenfelder, 2000). These experiments were divided into two stages. For each stage, the wastewater used as the feed was a mixture of two streams from the oil refinery: the affluent of the Industrial Wastewater Treatment Station (Stream L) and the stream from the Catalytic Cracking Unit and the Lubricant Distillation Tower (Stream D). The purpose of the mixture was to maintain feed the COD at a constant level and to ensure the presence of substances that cause chronic toxicity, which were present in stream D only. Table 4 shows the different stages of the continuous system, operating conditions and operating time. As can also be observed in Table 4, the SRT (Solid Retention Time) and Xci values
26
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) changed throughout the stages. These changes were made to achieve better efficiencies in the treatment and decrease the cost of replacement of the carbon.
Figure 1. Schematic of the reactor used in the experiments. (1-Feed container; 2-Peristaltic pump; 3-Aeration tank; 4-Airdiffision system; 5-Decanter; 6-Underflow sludge; 7-Sludge outlet; 8-Sliding baffle; 9-Effluent outlet; 10-Air compressor)
Table 4. Experimental conditions in the continuous system. Stage
System
Operational conditions Period of operation (days) HRT = 24 h AS-control SRT = 15 d PACT A Xci*= 300 mg.L-1 75 I PACT B Xpac** = 4.5 g.L-1 CODfeed = 1000 mg.L-1 F/M = 0,67 kgCOD.kg-1VSS.d-1 HRT = 24 h AS-control SRT = 30 d PACT A Xci* = 150 mg.L-1 70 II Xpac** = 4,5 g.L-1 CODfeed = 1000 mg.L-1 F/M = 0,5 kgCOD.kg-1VSS.d-1 *mg PAC per liter of affluent; **g PAC per reactor volume; AS-control, control reactor: activated sludge without PAC; PACT A and B: activated sludge with PAC A or B
During stage I, the PAC A and B processes and the “control” bioreactor (activated sludge process without PAC) were performed. In the second stage, because PAC A showed better efficiency than PAC B in stage I, only PAC A and the control processes were performed. The efficiencies of the systems were quantified by the removal of organic matter, through COD (Chemical Oxygen Demand), TOC (Total Organic Carbon) and BOD (Biochemical Oxygen Demand) analyses, as well as an absorbance analysis at a wavelength of 254 nm. Additionally, ammonia nitrogen, phenol, MLSS (Mixed Liquor Suspended Solid), MLVSS (Mixed Liquor Volatile Suspended Solid), carbon content, optical microscopy, dissolved oxygen, pH and total alkalinity were monitored. The efficiency goal established by the refinery to treat the wastewater was a COD of 150 mg.L-1. Analytical methodology COD, TOC, absorbance (254 nm), phenol, ammonia nitrogen and chloride were measured in samples taken from the affluent and effluent of the reactors. The total alkalinity, pH, dissolved oxygen, MLSS, MLVSS, and Xpac concentration in the PACT reactors were measured in samples taken from inside the reactors. All of the analyses were performed according to the methodology presented in APHA (2005), except for Xpac, which was measured using methodology developed by Siemens, called the "Zimpro Method” (Meidl, 1997). Optical Microscopy of Biomass in Bioreactors. To monitor the biomass characteristics, some microscopic observations were performed on
27
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) samples taken from inside the reactors. A trinocular Quimis microscope, model C7885K, was used. CAP Textural Analysis. Table 5 presents a list of the textural analyses performed in CAPs. Table 5. Textural characterization performed for carbon activated carbons. Analysis
SEM (Scanning Electronic Microscopy)
Description Measurement of gas adsorption capacity for forming a monomolecular layer on the surface of the adsorbent Textural characterization that allows assessment of the shape and texture of the adsorbent and the presence of mineral impurities.
X-Ray diffraction
Elemental analysis of the adsorbent constituents
Humidity content Ash content
Water absorption Intrinsic characteristic of the adsorbent precursor
BET surface area
Equipment ASAP 2000-Micrometrics TM1000 Tabletop Microscope
Diffractometer Flex II Desktop X-Ray Rigaku Incubator Gehaka G4023D Muffle furnace Quimis
Toxicity Assays The toxicity assays were performed during the continuous bioreactor stage. For acute toxicity, Danio rerio assays were performed according to the methodology proposed by the Brazilian Association of Technical Standards (ABNT, 2004). This methodology requires 10 fish (aged 4-12 months) to be placed in 2000 mL glass aquaria containing the negative control (reconstituted water) or sample (pH adjusted to 6.5-7.5 at 25±2 °C and aerated for at least 12 h to saturate dissolved oxygen and stabilize pH). The LC(I)50 values for Danio rerio were determined using the Trimmed Spearman– Karber method, version 1.5 (USEPA, 1990). Each assay was replicated four times. All of the Danio rerio assays were considered valid if mortality in the control assays was ≤ 5%. For chronic toxicity, assays were performed using the microcrustacean Ceriodaphnia dubia, in which the observed effect was the reproduction of the test organisms (ABNT, 2005). The results of chronic toxicity tests were statistically analyzed using the Williams test and Toxtstat 3.3 software (Gulley et al, 1991). The lowest-observed-effect concentration (LOEC), the no-observed-effect concentration (NOEC) and the chronic value (ChV) were calculated. Statistical significance was assumed to be at the p=0.05 level. The ChV was calculated as the geometric mean of the LOEC and the NOEC.
Results and Discussion PAC Choice Step Table 6 shows the results of adsorption experiments using bio-treated wastewater as feed. The COD values and absorbance at 254 nm were 90 mg.L-1 and 0.3851 cm-1, respectively. Table 7 shows the values of parameters Kf and n, taking into account the absorbance as the adsorbed pollutant parameter. All of the correlations presented R 2 >0.95. The results show that the PACs with higher affinity in the adsorption of recalcitrant organic substances from biological treatment effluent were carbon A and B. These two types of carbon presented higher K f values, which showed that they had higher recalcitrant organic substance adsorption rates. However, the values of n were higher for PACs C and D, showing that they had higher adsorption bonding strengths. Table 6. Results of the adsorption isotherm at equilibrium: medium values of absorbance at 254 nm and reduction of absorbance. -1
PAC dosage (g.L ) 0.1 0.3 0.5 0.8 1.0 3.0 5.0
Absorbance at 254 nm in cm-1, (reduction values in %) PAC A PAC B PAC C PAC D 0.1499 (61.1) 0.2368 (38.5) 0.2454 (36.3) 0.2831 (26.5) 0.0845 (78.1) 0.1173 (69.5) 0.1921 (50.1) 0.1704 (55.7) 0.0520 (86.5) 0.0659 (82.9) 0.1374 (64.3) 0.1317 (65.8) 0.0472 (87.7) 0.0374 (90.3) 0.1187 (69.2) 0.1061 (72.4) 0.0465 (87.9) 0.0298 (92.2) 0.1113 (71.1) 0.0975 (74.7) 0.0450 (88.3) * 0.1147 (70.2) 0.0931 (75.8) 0.0444 (88.5) * 0.1148 (70.2) 0.0930 (75.9)
*experiment was not performed
28
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
Table 7. Values of Kf and n from the Freundlich model. PAC A B C D
Kf (cm-1/gPAC) 312 153 35 39
n 0.73 0.56 1.06 1.49
PAC Characterization BET surface area Table 8 presents the surface area and porosity characterization of adsorbents PAC A and PAC B. Table 8. Surface area and porosity characterization of adsorbents PAC A and PAC B. Parameters Surface area (m².g-1)
BET area Micropore area External area
Micropore volume (cm³. g-1) Micropore (Å)
PAC A 958.16 722.23 235.93 0.347 28.06
PAC B 726.68 560.59 166.08 0.266 25.6
PAC A, which is from a bituminous source, has a higher BET area than PAC B. However, they are both good adsorbents because they have BET areas larger than 600 m².g-1 (Machado, 2000). Meanwhile, the micropore area, external area and micropore volume values associated with PAC A were higher than those of PAC B by approximately 20-30%. This finding indicates that PAC A is a better adsorbent than PAC B. Eckenfelder (2000) presents the properties of several PACs, including the BET area. For bituminous CAPs, the BET area was superior, ranging from 950-1050 m².g-1. For vegetal source PACs, the BET ranged from 600 to 650 m².g-1. Scanning Electronic Microscopy (SEM) Figures 2 (a, b and c) and 3 (a, b and c) show photos of the adsorbents taken using a Scanning Electronic Microscopy.
(a) (b) Figure 2. PAC A SEM micrographics: (a) 100 X, (b) 500 X, (c) 1000 X
(c)
PAC A has finely divided particles and an intense black color. It has a uniform structure and rounded particles. Large impurities were not observed because there were no variations in particle size or structure. This structural characteristic may be related to the source of the raw material (bituminous source). The PAC B particles were diverse in size and had a blade shape, which is different from the particles of PAC A. This difference may be attributed to the fact that PAC B is from a vegetal source and, more specifically, a wood source.
29
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
(a) (b) Figure 3. PAC B SEM micrographics: (a) 100 X, (b) 200 X, (c) 1000 X
(c)
X-Ray Diffraction Figure 4 presents the X-Ray diffraction intensities for both PAC A and PAC B. The presence of quartz and carbon in both PACs was verified. Mineral impurities, such as sodium, potassium and calcium, were not present. For PAC B, a specific peak for sulfur was observed, which may be present in its associated raw material.
Figure 4. X-Ray diffraction for both PAC A and PAC B. Humidity and volatile content Table 9 presents the humidity and volatile content values for PACs A and B. It was found that PAC A had a higher quantity of carbon than did PAC B. Ultimately, the PAC characterization results, in which PAC A showed a greater adsorption area and higher carbon content, may explain the better performance of PAC A compared to PAC B in the tests shown in the PAC choice step section. Table 9. Volatile content and humidity values for PACs A and B AC
Volatile content (%) A 81 B 43
Humidity content (%) 7.7 13.8
Continuous system experiments: activated sludge and PACT Stage I results. Figure 5 illustrates the total COD results obtained during the monitoring period of the 3 reactors (PACT A, PACT B and activated sludge). During the first ten days, all of the reactors operated without PAC so that the sludge could acclimate and reach steady state. Therefore, the first results obtained from the control reactor were an average of the results obtained from the other three reactors. Furthermore, within the graphic, there is another chart that illustrates the results in further detail. 30
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
1350
total COD (mg.L-1)
1200 1050 300
900
250
750 600 450
200
Affluent
150
Control reactor
100
PACT A
50
PACT B
0
300
0
10
20
30
40
50
60
70
80
150 0 0
10
20
30 40 50 Time of operation (d)
60
70
80
Figure 5. Results of total COD monitoring during the Stage I for control, PACT A and PACT B reactors. Operational conditions: HRT = 24 h; SRT = 15 d; Xci = 300 mg.L -1 and Xpac = 4.5 g.L-1. Except for a few data points at the end of monitoring, both reactors with CAP achieved the goal of 150 mg COD.L-1. In the PACT A process, the minimum COD value of the effluent reactor was 12 mg.L-1, reaching a removal rate of 98%. The PACT A process generated better quality effluent than PACT B. The control bioreactor showed the worst performance, with COD values in the range of 100 to 200 mg.L-1 for the treated wastewater. Additionally, it was verified that the COD levels of the effluent from the control reactor fluctuated, indicating low stability in situations where there is variation in the feed or an operational problem. Figure 6 illustrates the TOC in the reactors. The results of this analysis show that the removal percentages in the PACT reactor were higher than those of the control reactor. Both PACs showed similar TOC removal efficiencies. 250 225
200 175
TOC (mg.L-1)
150
Affluent
125
Control reactor
100
PACT A
75
PACT B 50 25 0 0
10
20
30
40
50
60
70
80
Time of operation (d)
Figure 6. Results of TOC monitoring during stage I for the control, PACT A and PACT B reactors. Operational conditions: HRT = 24 h, SRT = 15 d, Xci = 300 mg.L -1 and Xpac = 4.5 g.L-1. The TOC removal efficiencies were in the range of 50 to 90% for the control reactor, 95 to 99% for the PACT A reactor and 80 to 95% for the PACT B reactor. These results show that the operation of the PACT A process was more stable. Figure 7 presents the results of absorbance at 254 nm. Generally, the TOC results confirmed the results observed in total COD and TOC monitoring. However, it was observed that there was a further reduction of substances absorbing at 254 nm (most likely aromatic organic substances) in the PACT A reactor; thus, PAC A had the best performance.
31
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
4.4 4 3.6 Abs 254 nm (cm-1)
3.2 2.8
Affluent Control reactor PACT A PACT B
2.4 2 1.6
1.2 0.8
0.4 0 0
10
20
30 40 50 Time of operation (d)
60
70
80
Figure 7. Results of absorbance at 254 nm monitored during stage I for the control, PACT A and PACT B reactors. Operational conditions: HRT = 24 h, SRT = 15 d, Xci = 300 mg.L-1 and Xpac = 4.5 g.L-1.
Stage II results. At this stage, carbon replacement was decreased (from 300 to 150 mgPAC.L-1 affluent) to reduce the carbon cost, and consequently, SRT was increased (from 15 to 30 d) to maintain the same amount of accumulated carbon in the reactor (4.5 g.L-1). At this stage, only PACT A was run, as it had the best results in the previous stage. Figures 8 shows the results obtained from monitoring the total COD, as well as an inset graph with further detail. 1200
total COD (mg.L-1)
1050 350
900
300
750
250
600
150
Affluent
200
Control reactor
100
PACT A
50
450
0
0
300
10
20
30
40
50
60
70
80
150
0 0
10
20
30 40 50 Time of operation (d)
60
70
80
Figure 8. Results of total COD monitoring during the Stage II for the control and PACT A reactors. Operational conditions: HRT = 24 h, SRT = 30 d, Xci = 150 mg.L -1 and Xpac = 4.5 g.L-1.
On three of the days during the operating period, the effluent from the PACT A reactor did not reach the target of 150 mg COD.L-1. The total COD values for the control reactor effluent were, most of the time, far from the target. After each change of the affluent supply (days 19, 26, 40 and 53), it could be observed that the sensitivity of the control process was greater than the PACT A process, as it responded more quickly to changes in the operating system. This finding was verified by reestablishing the quality effluent in each process. In the PACT A process, the effluent was
32
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) reestablished faster than it was in the control reactor, as illustrated in the inset of Figure 8. The range of COD removal efficiencies in PACT A was 81 to 98%, while that of the control reactor was in the range of 55 to 90%. The analyses of the absorbance at 254 nm resulted in the following ranges: for the affluent, 8.688 to 10.550 cm-1; for the control reactor effluent, 1.073 to 3.722 cm-1; and for the PACT A effluent, 0.0804 to 0.6050 cm-1. Even with half of the carbon replaced daily, the COD and absorbance results were similar during stages I and II. This outcome indicates that it is possible to achieve the refinery goal (COD effluent = 150 mg.L-1) using less carbon and thus reducing the cost of the process. Figure 9 shows the results for the analyses of phenol in the affluent and effluent from both reactors, as well as the concentration required by legislation (0.2 mg.L-1) (Rio de Janeiro State, Brazil). Phenol was monitored only during the first 30 days of the experiment. 300
Affluent Control reactor
250
Phenol (mg.L-1)
PACT A 8
200
Control reactor PACT A legislation
7
6
150
5 4
100
3 2
50
1 0 0
5
10
15
20
25
30
0
0
5
10
15 20 Time of operation (d)
25
30
35
Figure 9. Phenol monitoring results during the Stage II for the control and PACT A reactors. Operational conditions: HRT = 24 h, SRT = 30 d, Xci = 150 mg.L -1 and Xpac = 4.5 g.L-1.
The variability in the composition of wastewater from the refinery accounts for the variation in the phenol concentration in the affluent ranging from 115 to 300 mg.L-1, as shown in Figure 9. The inset graph of Figure 9 shows that only the effluent from the PACT A reactor reached the required concentration for disposal (0.2 mg.L-1) after 27 days of operation. It was assumed that, during this period, the biomass acclimatized to the high phenol concentration, which was increased by the synergistic effect of the PAC in the mixture. Biomass Monitoring. It was observed that, at the end of each stage, the control reactor, consisting of activated sludge only, settled in an uneven manner, creating gelatinous sludge and a large amount of filamentous bacteria flocks. On these occasions, the SVI (Sludge Volumetric Index) values were greater than 200 mL.g-1, while for the PACT reactors (A or B, regardless of stage), the SVI values were in the range of 40 - 60 mL.g-1. Figure 10 (a, b, and c) shows some microscopic images taken during the operation of the reactors. The quantification of the accumulated carbon in the reactors shows that the amount of PAC in the reactor was consistent with the theoretical carbon content (eq. 1). These values are presented in Table 10, which also presents average values for the amount of solids in the reactors at different stages.
33
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
(a)
(b)
(c) Figure 10. (a) Biomass microscopy of the control reactor in Stage I showing excess filamentous bacteria, 100 X. (b) Biomass microscopy of PACT A in Stage I, 200 X. (b) Biomass microscopy of PACT B in Stage I, 200 X. Table 10. Carbon content and mean biomass concentration in the bioreactors
Stage I II
Carbon content (mg MLVSS.L-1) 4550 4450
PACT reactor (A or B) Theoretical carbon content (mgMLVSS.L-1) 4500 4500
Biomass (mg MLVSS.L-1) 1800 2500
Control reactor Biomass (mg Biomass MLSS.L-1) (mgMLVSS.L-1) 1200 1000 2750 2500
Toxicity assays. At stage II, samples for the acute and chronic toxicity assays were taken on the same day. Tables 11 and 12 present the results of the acute and chronic toxicity tests, respectively. Additionally, the COD and absorbance at 254 nm results for these samples are presented below. Table 11. Acute Toxicity (Danio rerio assays) for the affluent and effluent from reactors in stage II Parameter
Affluent
LC 50 – 24 h (%) Total COD (mg.L-1) Abs 254 nm
9.47 1019 9.46
Effluent from Control Reactor 65.98 241.7 1.29
Effluent from PACT A Reactor 70.10 123.5 0.332
The acute toxicity results did not differ significantly between the reactors, unlike the chronic toxicity assay results, which revealed that the effluent from the PAC A reactor presented a low chronic toxicity. This finding demonstrates that the PAC is essential to decreasing chronic toxicity. 34
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014)
Table 12. Chronic Toxicity (Ceriodaphnia dubia assays) for affluent and effluent from the reactors in stage II. Parameter NOEC (%) LOEC (%) ChV (%) Total COD (mg.L-1) Abs. 254 nm
Affluent
Effluent from the Control Reactor 0.78 1.56 1.10 153 1.691
0.39 0.78 0.55 1028 3.875
Effluent from PACT A reactor 25 50 35.5 24 0.387
Conclusions The main conclusions of this study are summarized below: • Among the four PACs used in the experiments for obtaining adsorption isotherms, the results show that the PACs with higher affinities for the adsorption of recalcitrant organic substances from biological treatment effluent were carbon A [bituminous carbon - Norit (SAE Super 94009-7)] and B [wood activated carbon – Carbomafra (118CB AS)]; • In the PAC characterization analyses, PAC A showed a greater adsorption area and a higher carbon content, which may explain the better performance of PAC A compared to that of PAC B in the bioreactor experiments; • In the PACT process, both types of carbon tested were effective in organic matter removal. A better efficacy was achieved by the PACT system than in the activated sludge process. However, PAC A was more effective than PAC B. Efficacy is based not only on the removal of organic matter but also on the ability to provide a more stable treatment system in adverse situations, such as variations in the feed or high concentrations of phenol; • During the monitoring period, the COD values of the PACT A effluent reached the target value required by the refinery (COD = 150 mg.L-1). Moreover, in Stages I and II, it was observed that the effluent COD reached values lower than 25 mg.L-1. In general, because of the greater system stability achieved by PAC A addition, better results were observed in the bio-treated effluent; • The biomass from the control reactor (activated sludge process) presented settling problems. In the microscopic investigations, it was verified that, in these periods, the sludge flocks had high concentration of filamentous organisms, hindering sedimentation; • The toxicity tests showed that the effluent from PACT A presented a less chronic toxic effect on the reproduction of the species (Ceriodaphnia dubia) when the tests were performed under the following operational conditions: HRT=24 h, SRT=30d and replacement dosage carbon = 150 mg.L-1. A less significant difference was observed in the acute toxicity tests performed using Danio rerio fish. Acknowledgments: The authors would like to thank Petrobras for the shipment of wastewater, the exchange of technical information and financial support.
Nomenclature
ABNT – Associação Brasileira de Normas Técnicas ASTM - American Society for Testing and Materials BOD – Biochemical Oxygen Demand, in mg O2.L-1 Ce - adsorbate equilibrium concentration, in mgpollutant.L-1 ChV - chronic value (ChV), in % COD - Chemical Oxygen Demand, in mg O2.L-1 HRT – Hydraulic Retention Time, in h Kf - experimental constant Freundlich model, related to the adsorption capacity of the adsorbent by the adsorbate LC(I)50 – lethal dose for 50% tested organisms, in % LOEC - lowest-observed-effect concentration, in %
MBR – Membrane Bioreactor MLSS - Mixed Liquor Suspended Solid, in mg Solids.L-1 MLVSS - Mixed Liquor Volatile Suspended Solid, in mg Solids.L-1 n - experimental constant Freundlich model, related to the adsorption bonding strength N – nitrogen NOEC - no-observed-effect concentration, in % P- phosphorus PAC - powdered activated carbon PACT™ - Powdered Activated Carbon Treatment qe - adsorbate mass unit per adsorbent mass unit, in mgpollutant.g-1PAC SEM - Scanning Electronic Microscopy
35
J. Int. Environmental Application & Science, Vol. 9(1): 24-36 (2014) replacement), in mgPAC.L-1affluent Xpac - accumulated concentration of PAC in the reactor (equilibrium powdered activated carbon – MLSS content), in mgPAC.L-1reactor
SRT – Solid Retention Time, in days SVI - Sludge Volumetric Index, in mL.g-1 TOC – Total Organic Carbon, in mg.L-1 USEPA – US Environmental Protection Agency Xci - PAC concentration dosage (the carbon
References ABNT (Brazilian Association of Technical Standards). (2004).Aquatic ecotoxicology - Acute toxicity – Method of assays with fishes. NBR 15088. Rio de Janeiro, Brazil (in Portuguese). ABNT (Brazilian Association of Technical Standards). (2005). Aquatic ecotoxicology - Chronic toxicity – Method of assays with Ceriodaphnia spp. (Crustácea, Cladocera), NBR 13.373. Rio de Janeiro, Brazil. (in Portuguese). Aktas, O., Çecen, F. (2007) Bioregeneration of activated carbon: A review. International Biodeterioration & Biodegradation . 59, 257–272. ASTM 3860 (2003)- Standard Practice for determination of adsorptive capacity of activated carbon by aqueous phase isotherm Technique. Eckenfelder WWJ, (2000) Industrial Water Pollution Control, 3rd Ed., The McGraw-Hill, USA. Gulley, D.D., Boelter, A.M., Bergman, H.L. (1991). Toxstat 3.3. Laramie: University of Wyoming, Department of Zoology and Phisiology, Fish Physiology and Toxicology Laboratory, v.1. INEA – Rio de Janeiro State Institure of Environment – DZ (Guideline) 205, Revision number 6 (2007). Guidelines for Organic Load Control in Industrial Liquid Wastewaters – November 8th, 2007, 1-6 (in Portuguese). INEA – Rio de Janeiro State Institure of Environment – NT (Technical Note) 202, Revision number 10 (1986). Criteria and Standards for Discharge of Liquid Wastewaters. – December 12th, 1986, 1-4 (in Portuguese). Lesage N, Sperandio M, Cabassud C, (2008). Study of a hybrid process: Adsorption on activated carbon/membrane bioreactor for the treatment of an industrial wastewater. Chemical Engineering and Processing. 47, 303-307. Machado, C.R.A. Avaliação de processo de lodos Ativados combinado com carvão ativado em pó no tratamento de efluente de refinaria de petróleo (Evaluation of activated sludge process combined with powdered activated carbon in oil refinerey wastewater treatment). (2010) M.Sc. thesis, Chemical and Biochemical Process Tecnology Programe, Federal University of Rio de Janeiro, Brazil. (In Portuguese) Meidl JA, (1997) Responding to changing conditions: How powered activated carbon system can provide the operational flexibility necessary to treat contaminated groundwater and industrial wastes. Carbon. 35, 1207-1216 . Metcalf & Eddy Inc. (2003) .Tchobanoglous, G., Burton, F L., Stensel, D. Wastewater engineering: treatment and reuse, 4ª. Ed, McGraw Hill, USA. Orozco A.M.F., Contreras, E.M., Zaritzky, N.E. (2011). Effects of combining biological treatment and activated carbon on hexavalent chromium reduction; Bioresource Tech., 102, 2495-2502. Orozco A.M.F., Contreras, E.M., Zritzky, N.E. (2010) Dynamic response of combined activated sludge-powdered activated carbon batch systems. Chemical Engineering J., 157, 331-338. Satyawali Y., Balakrishnan M., (2009) Effect of PAC addition on sludge properties in an MBR treating high strength wastewater. Water Research, 43, 1577-1588. Sher MI, Arbuckle WB, Shen Z, (2000) Oxygen Uptake Rate Inhibition with PACT Sludge, J. Hazardous Materials, B73, 129-142 . Standard Methods for the Examination of Water and Wastewater. (2000) 21st Ed, American Public Health Association/American Water Works Association/Water Environment Federation, Washington DC, USA. USEPA. (1990) Trimmed Spearman-Karber (TSK) program version 1.5 ecological monitoring research division. Environmental Monitoring Systems Laboratory, USEPA, Cincinnati, OH. Zhao R, Novak JT, Goldsmith D, (2012) Evaluation of on-site biological treatment for landfill leachates and its impact: A size distribution study. Water Research, 46, 3837-3848 .
36
