- Email: [email protected]
Separation of colloidal polymeric waste using a local soil
Separation BYPurification Technology
ELSEVIER
Separation and Purification Technology 13 (1998) 161-169
Separation of colloidal polymeric waste using a local soil Nabil S. Abuzaid *, Muhammad H. ALMalack, Aarif H. El-Mubarak Research Institute, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia
Received 1 April 1997; received in revised form 17 November 1997; accepted 26 November 1997
Abstract The use of a local soil as a destabilizer for an emulsified colloidal wastewater was investigated in this study. The soil was found to contain 68% quartz, 24% muscovite, and 8% hedenbergite and to have a surface area, a pore volume, and a pore diameter of 173 m’/g, 0.16 c3/g, and 40.5 A, respectively. While preliminary investigation of the pollutants in the raw wastewater revealed their poor solubility and settleability, the reduction of the supernatant COD increased with the increase in soil mass and time until equilibrium was reached (within 24 h). Furthermore, the lowest supernatant COD achieved was within the acceptable range set by the regulatory authority. While destabilization of the colloidal polymers by the soil was attributed to the adhesion enhanced by the large soil surface area and the existence of aluminum and iron oxides, sedimentation was believed to occur because of discrete and zone types of settling. A considerable portion of the removal efficiency was achieved in the first hour and in the time range of (6-24 h), resulting in removal efficiencies as high as 95%. While the loading rate and the capacity of the soil were inversely proportional to the soil mass, the findings of the study indicated that lower kinetics and higher equilibrium capacities are expected if the clay proportion in the soil is increased by separation of quartz. 0 1998 Elsevier Science B.V. Keywords: Aluminum;
Coagulation;
Colloids; Iron; Polymers; Soil; Wastewater
1. Introduction
Purification of industrial wastewater prior to discharge is necessary to keep our environment clean. However, industrial wastewater treatment is becoming very costly, particularly if chemicals are used. Therefore, research on less costly treatment methods is crucial. Polymers-containing wastewater are difficult to treat because of their complex nature. This difficulty worsens when those polymers are neither soluble nor settleable. Soluble waste can be treated by adsorption [ 1 ] or biodegra* Corresponding author. Tel: 00 966 3 8604944; Fax: 00 966 3 8602266; e-mail: [email protected] 1383-5866/98/%19.000 1998 Elsevier Science B.V. All rights reserved S1383-5866(97)00069-S
PII
dation [2] while for settleable waste, sedimentation is the most used process. For emulsified suspended polymeric waste, on the other hand, the constituents of the wastewater need to be coagulated and flocculated before it can be clarified. Colloidal dispersion may be stabilized by electrostatic repulsion between particles, arising from ions that are either adsorbed onto or dissolved out of the surface of the solid. Destabilization of those colloids is essential in order to bring them in contact and aggregate. Coagulants can destabilize colloidal particles by four distinct mechanisms: double layer compression; charge neutralization; enmeshment in a precipitate; and interparticle bridging [ 31.
162
N.S. Abuzaid et al. /Separation
and Purification Technology 13 (1998) 161-169
Separation of colloidal industrial wastes was usually undertaken by the use of chemical coagulants. Shut’ko [4,5] investigated the treatment of petroleum wastewater by aluminum-containing coagulants. The findings of the study revealed that aluminum-containing coagulants were effective for the separation of petroleum wastewater, particularly for effluents with high suspended matter. Coagulation of pulp and paper mill effluents using alum and chlorinated copperas was carried out by Soetopo [6,7]. In his study, it was reported that significant reduction of suspended solids, biological oxygen demand (BOD), chemical oxygen demand (COD) and color was achieved. Wastewater from a textile plant was effectively treated by alum coagulation resulting in disperse dye reduction of 85% [8,9]. However, the cost involved in the purification of the aforementioned industrial wastewater by chemical coagulants should not be overlooked. As mentioned earlier, cheaper alternatives should be found and investigated. Clays or soils containing clay are used prevalently as containment liners in hazardous landfill sites and in slurry walls at hazardous waste sites, with the primary objective of retarding the migration of contaminants from the site [lo]. A major mechanism of organics soil interaction involves a partition process that depends almost exclusively on solute water solubility [ 10,111. Bentonitic clays have been used as coagulant aids for waters with low turbidity. Their use was found to reduce the amount of coagulant dosages and improve floes characteristics by weighting them leading to rapid settlement [ 121. Coagulation followed by sedimentation of polymeric wastewater by the addition of clay containing soil was investigated in this study. The soil used is a local soil found in the Eastern Province of Saudi Arabia named Khoweldi by the local people. The polymeric wastewater under investigation is discharged by a chemical factory located in the Eastern Province of Saudi Arabia. The factory manufactures emulsion polymers such as homopolymers, copolymers, terpolymers, styrene acrylics and pure acrylics. The discharged wastewater consists of very condensed colloidal polymers. Biodegradation of the aforementioned wastewater
in typical activated sludge processes was recommended by Nemerow [ 131. However, for the case under study, the average COD of the discharged wastewater was very much higher than the maximum allowable limit of wastewater effluents to the biological treatment plant of the industrial city. Therefore, pretreatment is necessary before the wastewater is allowed to be discharged. On the basis of the above discussion, the main objective of this research is to investigate the efficiency of the Khoweldi soil for the purification of the polymeric colloidal wastewater described earlier. After characterization of the soil was performed, different doses of the soil were added to the wastewater samples, and the supematant chemical oxygen demand (COD) was monitored with time until equilibrium was achieved.
2. Materials and methods The soil used in this study was dried in an oven at 105°C for about 3 h, allowed to cool at room temperature, ground and stored in a desiccator prior to experimental work. The surface area, pore volume, and the pore diameter of the Khoweldi soil were determined by measuring nitrogen adsorbed on a degassed sample at different relative pressures and liquid nitrogen temperature using the theory developed by BET. The procedure involves measuring the nitrogen physically adsorbed as a monolayer on the surface of the soil at the liquefaction temperature of nitrogen. X-ray diffractometer (XRD) is an indespensable technique for investigation of the crystal structure of the solid matter. In this study, XRD analysis was performed for the soil on a Philips PW 1700 automated diffractometer with a monochromator and a spinner. Diffraction patterns were generated on a vertical goniometer attached to a broad focus X-ray tube with a copper target operating at 45 kV and 30 mA. The soil specimens were powdered with a pestle and mortar. A homogeneous sample of this powder was packed into a sample holder and scanned from 4-80” 28 at a speed of 0.01 2&‘/s. Metal content was determined by transferring 1 g of the Khoweldi soil to a 50 ml beaker. The sample had 7 ml of aqua regia ( HC1:H,NOJ, 3: 1)
N.S. Abuzaid et al. / Separation and Purification Technology I3 (1998) 161-169
added to it. The content was heated at 60°C for 1 h and then at about 120°C to near dryness (total time of digestion was about 4 h). Subsequently, the heater was cooled, the content filtered and the volume was raised to 50 ml using double distilled water. This aliquot was used directly for the determination of metal content in the soil using inductively coupled plasma (model 3580 OES, ARL, Switzerland). Further dilutions were made to bring higher concentrations into the dynamic range. Three exposures were averaged and the RSD was calculated. A 20 1 sample was collected from the wastewater effluent of the factory and stored in a fridge to prevent any biological activity. From the stored sample 1 1 of the wastewater was taken and it was divided into two portions. One portion was filtered using 0.45 urn filter paper. The filtered and unfiltered portions of the wastewater were analyzed for COD in order to know the soluble and total COD, respectively. The COD was analyzed in accordance with the procedure outlined in the Standard Methods [ 131. The pH of the wastewater was measured using a digital pH meter (ORION, Model 701A) and found to be 7.4. The settleability of the polymeric colloids in the wastewater was tested in order to check the extent of stability of the colloids. A 1 1 volume measuring cylinder was filled with the wastewater and monitored for a period of 24 h for COD. The settleability test was conducted according to the procedures outlined in the Standard Methods [ 131. For the coagulation study, jar tests were conducted using a multiple stirring apparatus. Seven batches were run at the same time, each was filled with 500 ml of the wastewater. The wastewater in the seven beakers was analyzed for COD. Six different masses of the Khoweldi soil (0.5, 1, 1.5, 2, 2.5 and 3 g) were added to the aforementioned beakers. The seventh beaker was run without soil to serve as a blank. In the beginning, the beakers were mixed at high speed (100 rpm) for 1 min. Subsequently, the mixing speed was lowered (40 rpm) and continued for another 10 min. The speed was further reduced to 8 rpm and the mixing process continued for another 10 min. After the mixing stage, COD of the supernatant was monitored with time.
163
3. Results and discussion Identification of the soil surface area, pore volume, and pore diameter, resulted in values of 173 m2/g, 0.16 c3/g, and 40.5 A, respectively. These results, particularly the surface area, show that the Khoweldi soil has a relatively good sorption ability. Since the most important soil constituent for the purpose of coagulation is clay, X-ray diffraction (XRD) tests were conducted, mainly to investigate the existence of clay in the Khoweldi soil. The results of the XRD analysis showed that the Khoweldi soil contained 68% quartz, 24% muscovite, and 8% hedenbergite. The XRD pattern for the Khoweldi soil is given in Fig. l(a) while the patterns of its constituents i.e. quartz, muscovite, and hedenbergite are shown in Figs. 1(b)-(d), respectively. Analysis of metal content in the soil has shown concentrations as high as 67450 and 39708 ppm of aluminum oxide and iron oxide, respectively. This is a very important finding because those oxides can form coagulants that helps with destabilization process of the polymeric colloids. The COD of the polymeric wastewater discharged by the factory was found to be 13500 mg/l, while the COD of the filtered sample (soluble COD) was found to be 130 mg/l. This clearly shows that the soluble COD is less than 1% of the total COD which is negligible and thus will not be considered in further work. Actually, this finding has motivated the main objective of this work which is destabilization of the emulsified colloids in the polymeric wastewater and not treatment of the soluble constituents. The high COD of the bulk sample emphasizes the need of pretreatment before the effluent can be discharged to the biological treatment plant of the industrial city. It should be remembered that the COD limit for industrial wastewater discharges to the aforementioned plant is 1000 mg/l. The settleability test of the wastewater revealed that even after a detention time of 24 h the colloidal waste was totally in suspension. This stressed the stable nature of the colloids under study and emphasized the need for a destabilizer to flocculate the colloids and subsequently have them clarified. In other words, conventional sedimentation tanks
164
N.S. Abuzaidet al. /Separation and Purification Technology 13 (1998) M-169
Fig. 1. X-ray diffraction patterns of the Khoweldi soil (a) and the constituents; quartz (b), muscovite (c), and hedenbergite (d).
14000.0 13000.0 12000.0 11000.0 lCOCO.O 9000.0 f
8000.0 7000.0
8 0
6000.0 6000.0 4000.0 3000.0 2000.0 1000.0 III 0.0
4.0
0.0
12.0
111
III
16.0
20.0
Time
(hr)
II, 24.0
II1 26.0
,,, 32.0
36.0
Fig. 2. Relationship between the supematant COD and time of sedimentation at different soil masses.
165
N.S. Abuzaid et al. / Separation and Purtjication Technology 13 (1998) 161-169
will not be able to remove the constituents of the wastewater under study. Fig. 2 depicts the relationship between the supernatant COD and time of sedimentation under the effect of six soil masses. It should be noted that the COD measurements started after the mixing stage. The figure shows that at any time the supernatant COD is inversely proportional to the amount of soil used. It is also shown that the lowest supernatant COD achieved was 675 mg/l which is acceptable by the standards of the industrial city. This value was achieved at a soil mass of 3 g after a detention time of 24 h. The aforementioned result shows that the problem of the wastewater discharges from the factory can be solved by the use of the Khoweldi soil as a destabilizer. It should be noted from Fig. 2 that equilibrium with time was reached within 24 h under all of the soil masses. It is worth mentioning that negligible COD reduction was found in the beaker with no soil. The interaction between the Khoweldi soil and the colloidal polymers in the wastewater may best be interpreted in terms of surface adhesion because of the large soil surface
area in general, and the existence of aluminum and iron oxides. When aluminum and iron oxides dissolve in water they are expected to become aluminum and iron hydroxides (coagulant agents). Regarding sedimentation, there are two possible mechanisms; discrete settling in which each particle settles by it self which is characteristic of the coarse portion of the soil, and zone settling in which fine particles settle as a blanket. The second mechanism is supposed to enhance sedimentation of the colloidal polymers by sweeping them by the settling zone. Removal of colloids in this manner is frequently referred to as a sweep-floe coagulation [ 31. This is of course in addition to the aforementioned polymers-particle adhesion process. Fig. 3 shows the temporal effect of different soil masses on the COD removal efficiency. The highest removal efficiencies (88-95%) were achieved when the soil mass was 3 g, while the lowest values (50-72%) were achieved under a soil mass of 0.5 g. Fig. 3 has a useful application regarding the design and operation of this process. If the process is to be designed for a certain removal efficiency, two parameters can be manipulated, namely, soil mass
0.85 0.80 '0 s
0.75
1 B
0.70
E d
0.65
:: 0
0.60
0.0
4.0
8.0
12.0
16.0
20.0
24.0
28.0
32.0
36.0
Time (hr) Fig. 3. The effect of different
soil masses on the COD removal
efficiency
as a function
of time.
166
N.S. Abuzaidet
al. /Separation
and Purification
Technology 13 (1998) 161-169
100.00 95.00
1
ii cz
A
A
90.00
A 85.00
A
A A
55.00 50.00
I 0.00
0.50
I
I
I
I
I
1.00
1.50
2.00
2.50
3.00
3.50
Soil Mass (g)
Fig. 4. Relationship between the ratio of COD removed during the first 1 h to the total COD removed and soil mass.
and detention time. For example, a removal efficiency of 70% can be achieved at detention times of 30 min, 8, 13 and 22 h, and soil masses of 2, 1.5, 1, and 0.5 g, respectively. It should be understood that the decision on which pair of these parameters to adapt in order to reach a given removal efficiency is mostly related to economy which is outside the scope of this work. In Figs. 3 and 2 it is clear that in the first 1 h a considerable portion of the removal efficiency was achieved under each of the soil masses and the ratio of the removal efficiency reached to the total removal efficiency increased with the increase in soil mass. Furthermore, this removal efficiency did not increase considerably in the first 6 h for all the masses used. During the time period (6-24 h), a considerable increase in the removal efficiency was achieved. The increase in the aforementioned period was inversely proportional to the soil mass. However, at equilibrium (within 24 h), the total removal efficiency attained was proportional to the soil mass. The kinetics of this process is very complicated because of the heterogeneity in both
the wastewater and the soil. However, in an attempt to explain the aforementioned findings the ratio of COD removed during the first 1 h to the total COD removed versus soil mass are plotted in Fig. 4. The figure generally shows that the percentage of the COD removed in the first 1 h to the total COD removed increases with the increase in soil mass up to a soil mass of 2.5 g after which the aforementioned percentage did not change. In order to explain this behavior, one should recall that the soil used was not pure clay but a mixture of mainly, muscovite and quartz. The non clay portion was higher in quantity than clay and naturally coarser and heavier. So, it is expected that most of the early COD removal (within 1 h) occurs mostly because of the non clay portion and clay starts settling with the attached polymers at a later time (after 6 h as was shown in Fig. 4). At a low soil mass, the amount of coarse constituents is low, resulting in a comparatively low percentage of early COD removal which will give chance to the fine constituents (clay) to remove more COD with time. However, at a high soil mass, the
N.S. Abuzaid et al. /Separation and PuriJcation Technology 13 (1998) 161-169
167
2000.0 0.0
4.0
8.0
12.0
16.0
20.0
24.0
28.0
32.0
36.0
Time (hr) Fig. 5. The variation of the amount of COD removed per mass of soil with time under different oil masses
comparatively high coarse constituents are responsible for higher percentage of early COD removals reducing the amount of COD to be removed by the fine portion. Above a soil mass of 2.5 g, the percentage of early COD removals reaches a fixed value of 95%. This means that 5% of the removable COD cannot be removed by the coarse portion of the soil regardless of quantity. From Fig. 4 it can be concluded that if efforts were made to increase the clay quantity in the Khoweldi soil by separation of quartz, both the kinetics and equilibrium capacity would be affected. As a result, kinetics may be lowered while higher equilibrium capacities are expected to be attained due to the comparatively large surface area of clay. In other words, less soil masses would be needed but at the cost of time. The temporal variation of the amount of COD removed per mass of soil with different masses of soil is depicted in Fig. 5. From the figure it is obvious that the rate of increase of the aforementioned ratio with time increases with the decrease in soil mass. At equilibrium (after 24 h) the amount
of COD removed per mass of soil can be defined as soil capacity. From the figure it is clear that the soil capacity is inversely proportional to the soil mass, i.e. higher colloids to soil mass ratios results in higher soil capacities. From Fig. 5 it can be concluded that both the kinetics and the capacity of soil coagulation for the waste under study increase with the decrease in soil mass. The effect of the variation in soil masses on kinetics was explained earlier while its influence on soil capacity will be explained soon. The relation between soil capacity and supernatant COD is depicted in Fig. 6. This type of illustration is very common in the adsorption studies and named an isotherm [ 11. The data of Fig. 6 demonstrated two distinct trends. Up to a supernatant concentration of 2500 mg/l, a linear relation between capacity and COD exists. Above the aforementioned COD, another linear relation is depicted. However, the slopes of the two trends are dramatically different. In fact, the slope of the later trend is larger than that of the first one. This kind of behavior is known in the activated carbon
168
N.S. Abuzaid et al. /Separation
and Purification
Technology 13 (1998) 161-169
20000.0
A
A A
4000.0
!I 500.0
A
A
looo.o
lsoo.o
2000.0
2500.0
3000.0
3500.0
4000.0
Residual COD (I@) Fig. 6. Relationship between soil capacity and supematant COD.
adsorption literature [3] and attributed to the complexity and heterogeneity of the wastewater which is typical of the waste under study.
4. Conclusions The Khoweldi soil was found to destabilize the colloidal waste resulting from a polymeric industry, and the reduction of the supernatant COD increased with soil mass and time until equilibration which was reached in 24 h. The lowest supernatant COD achieved was acceptable to the regulatory authority standards. The destabilization of the colloidal polymers was attributed to the adhesion enhanced by the large soil surface and the existence of aluminum and iron oxides. However, sedimentation was believed to occur because of discrete and zone settling. A considerable portion of the removal efficiency was achieved in the first 1 h and in the time range of (6-24 h), and the highest removal efficiency
achieved was 95%. Furthermore, the output of the study indicated that lower kinetics and higher equilibrium capacities are expected if the clay proportion in the Khoweldi soil is increased by separation of quartz. Additionally, the loading rate and the capacity of the soil were found to be inversely proportional to the soil mass. Finally, the above mentioned findings have colloidal polymeric important implications; wastewater discharges with high COD concentrations are amenable to purification by a cheep local material (soil). Accordingly, the results obtained can be extended to wastes and soils of similar properties to those investigated in this study.
Acknowledgment The authors thank the Research Institute, Ring Fahd University of Petroleum and Minerals, for providing support to this research.
N.S. Abuzaid et al. 1 Separation and Purification Technology 13 (1998) 161-169
References [I] N. Abuzaid, G. Nakhla, J. Envir. Sci. Technol. 28 (1994) 216. [2] G. Nakhla, I. Harazin, Envir. Technol. 14 (1993) 751-760. ]3] L. Benefield, J. Judkins, B. Weand, Process Chemistry for Water and Wastewater Treatment, End ed., Prentice-Hall, NJ, 1972. [4] A.P. Shu’ko, Neftepererab. Neftechim. (USSR) 8 (5) (1986). [5] A.P. Shu’ko, Chem. Abstr. 105, 158243~ (1986). [6] R.S. Soetopo, Berita Selulosa (Indon) 20(2) (1984) 46. [7] R.S. Soetopo, Chem. Abstr. 105, 11414a (1986).
[8] 0. Colak, B. Arikan, Doga:Mohendislik
169
Cevre Bilimleri (Turk.), lO(2) (1986) 185. [9] 0. Colak, B. Arikan, Chem. Abstr. 105, 1582565 (1986). [lo] C.T. Chion, S.E. Porter, D.W. Schmediting, Envir. Sci. Technol. 17 (1983) 227. [ 111 D.R. Garbarini, L.W. Lion, Envir. Sci. Technol. 20 (1986) 126331269. [ 121 N.L. Nemerow, Industrial Water Pollution: Origins, Addison-Wesley, Treatment, Characteristics, and Reading, UK, 1978. [ 131 American Public Health Association. Standard Methods; for the Examination of Water and Wastewater, 15th ed., 1980.
ELSEVIER
Separation and Purification Technology 13 (1998) 161-169
Separation of colloidal polymeric waste using a local soil Nabil S. Abuzaid *, Muhammad H. ALMalack, Aarif H. El-Mubarak Research Institute, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia
Received 1 April 1997; received in revised form 17 November 1997; accepted 26 November 1997
Abstract The use of a local soil as a destabilizer for an emulsified colloidal wastewater was investigated in this study. The soil was found to contain 68% quartz, 24% muscovite, and 8% hedenbergite and to have a surface area, a pore volume, and a pore diameter of 173 m’/g, 0.16 c3/g, and 40.5 A, respectively. While preliminary investigation of the pollutants in the raw wastewater revealed their poor solubility and settleability, the reduction of the supernatant COD increased with the increase in soil mass and time until equilibrium was reached (within 24 h). Furthermore, the lowest supernatant COD achieved was within the acceptable range set by the regulatory authority. While destabilization of the colloidal polymers by the soil was attributed to the adhesion enhanced by the large soil surface area and the existence of aluminum and iron oxides, sedimentation was believed to occur because of discrete and zone types of settling. A considerable portion of the removal efficiency was achieved in the first hour and in the time range of (6-24 h), resulting in removal efficiencies as high as 95%. While the loading rate and the capacity of the soil were inversely proportional to the soil mass, the findings of the study indicated that lower kinetics and higher equilibrium capacities are expected if the clay proportion in the soil is increased by separation of quartz. 0 1998 Elsevier Science B.V. Keywords: Aluminum;
Coagulation;
Colloids; Iron; Polymers; Soil; Wastewater
1. Introduction
Purification of industrial wastewater prior to discharge is necessary to keep our environment clean. However, industrial wastewater treatment is becoming very costly, particularly if chemicals are used. Therefore, research on less costly treatment methods is crucial. Polymers-containing wastewater are difficult to treat because of their complex nature. This difficulty worsens when those polymers are neither soluble nor settleable. Soluble waste can be treated by adsorption [ 1 ] or biodegra* Corresponding author. Tel: 00 966 3 8604944; Fax: 00 966 3 8602266; e-mail: [email protected] 1383-5866/98/%19.000 1998 Elsevier Science B.V. All rights reserved S1383-5866(97)00069-S
PII
dation [2] while for settleable waste, sedimentation is the most used process. For emulsified suspended polymeric waste, on the other hand, the constituents of the wastewater need to be coagulated and flocculated before it can be clarified. Colloidal dispersion may be stabilized by electrostatic repulsion between particles, arising from ions that are either adsorbed onto or dissolved out of the surface of the solid. Destabilization of those colloids is essential in order to bring them in contact and aggregate. Coagulants can destabilize colloidal particles by four distinct mechanisms: double layer compression; charge neutralization; enmeshment in a precipitate; and interparticle bridging [ 31.
162
N.S. Abuzaid et al. /Separation
and Purification Technology 13 (1998) 161-169
Separation of colloidal industrial wastes was usually undertaken by the use of chemical coagulants. Shut’ko [4,5] investigated the treatment of petroleum wastewater by aluminum-containing coagulants. The findings of the study revealed that aluminum-containing coagulants were effective for the separation of petroleum wastewater, particularly for effluents with high suspended matter. Coagulation of pulp and paper mill effluents using alum and chlorinated copperas was carried out by Soetopo [6,7]. In his study, it was reported that significant reduction of suspended solids, biological oxygen demand (BOD), chemical oxygen demand (COD) and color was achieved. Wastewater from a textile plant was effectively treated by alum coagulation resulting in disperse dye reduction of 85% [8,9]. However, the cost involved in the purification of the aforementioned industrial wastewater by chemical coagulants should not be overlooked. As mentioned earlier, cheaper alternatives should be found and investigated. Clays or soils containing clay are used prevalently as containment liners in hazardous landfill sites and in slurry walls at hazardous waste sites, with the primary objective of retarding the migration of contaminants from the site [lo]. A major mechanism of organics soil interaction involves a partition process that depends almost exclusively on solute water solubility [ 10,111. Bentonitic clays have been used as coagulant aids for waters with low turbidity. Their use was found to reduce the amount of coagulant dosages and improve floes characteristics by weighting them leading to rapid settlement [ 121. Coagulation followed by sedimentation of polymeric wastewater by the addition of clay containing soil was investigated in this study. The soil used is a local soil found in the Eastern Province of Saudi Arabia named Khoweldi by the local people. The polymeric wastewater under investigation is discharged by a chemical factory located in the Eastern Province of Saudi Arabia. The factory manufactures emulsion polymers such as homopolymers, copolymers, terpolymers, styrene acrylics and pure acrylics. The discharged wastewater consists of very condensed colloidal polymers. Biodegradation of the aforementioned wastewater
in typical activated sludge processes was recommended by Nemerow [ 131. However, for the case under study, the average COD of the discharged wastewater was very much higher than the maximum allowable limit of wastewater effluents to the biological treatment plant of the industrial city. Therefore, pretreatment is necessary before the wastewater is allowed to be discharged. On the basis of the above discussion, the main objective of this research is to investigate the efficiency of the Khoweldi soil for the purification of the polymeric colloidal wastewater described earlier. After characterization of the soil was performed, different doses of the soil were added to the wastewater samples, and the supematant chemical oxygen demand (COD) was monitored with time until equilibrium was achieved.
2. Materials and methods The soil used in this study was dried in an oven at 105°C for about 3 h, allowed to cool at room temperature, ground and stored in a desiccator prior to experimental work. The surface area, pore volume, and the pore diameter of the Khoweldi soil were determined by measuring nitrogen adsorbed on a degassed sample at different relative pressures and liquid nitrogen temperature using the theory developed by BET. The procedure involves measuring the nitrogen physically adsorbed as a monolayer on the surface of the soil at the liquefaction temperature of nitrogen. X-ray diffractometer (XRD) is an indespensable technique for investigation of the crystal structure of the solid matter. In this study, XRD analysis was performed for the soil on a Philips PW 1700 automated diffractometer with a monochromator and a spinner. Diffraction patterns were generated on a vertical goniometer attached to a broad focus X-ray tube with a copper target operating at 45 kV and 30 mA. The soil specimens were powdered with a pestle and mortar. A homogeneous sample of this powder was packed into a sample holder and scanned from 4-80” 28 at a speed of 0.01 2&‘/s. Metal content was determined by transferring 1 g of the Khoweldi soil to a 50 ml beaker. The sample had 7 ml of aqua regia ( HC1:H,NOJ, 3: 1)
N.S. Abuzaid et al. / Separation and Purification Technology I3 (1998) 161-169
added to it. The content was heated at 60°C for 1 h and then at about 120°C to near dryness (total time of digestion was about 4 h). Subsequently, the heater was cooled, the content filtered and the volume was raised to 50 ml using double distilled water. This aliquot was used directly for the determination of metal content in the soil using inductively coupled plasma (model 3580 OES, ARL, Switzerland). Further dilutions were made to bring higher concentrations into the dynamic range. Three exposures were averaged and the RSD was calculated. A 20 1 sample was collected from the wastewater effluent of the factory and stored in a fridge to prevent any biological activity. From the stored sample 1 1 of the wastewater was taken and it was divided into two portions. One portion was filtered using 0.45 urn filter paper. The filtered and unfiltered portions of the wastewater were analyzed for COD in order to know the soluble and total COD, respectively. The COD was analyzed in accordance with the procedure outlined in the Standard Methods [ 131. The pH of the wastewater was measured using a digital pH meter (ORION, Model 701A) and found to be 7.4. The settleability of the polymeric colloids in the wastewater was tested in order to check the extent of stability of the colloids. A 1 1 volume measuring cylinder was filled with the wastewater and monitored for a period of 24 h for COD. The settleability test was conducted according to the procedures outlined in the Standard Methods [ 131. For the coagulation study, jar tests were conducted using a multiple stirring apparatus. Seven batches were run at the same time, each was filled with 500 ml of the wastewater. The wastewater in the seven beakers was analyzed for COD. Six different masses of the Khoweldi soil (0.5, 1, 1.5, 2, 2.5 and 3 g) were added to the aforementioned beakers. The seventh beaker was run without soil to serve as a blank. In the beginning, the beakers were mixed at high speed (100 rpm) for 1 min. Subsequently, the mixing speed was lowered (40 rpm) and continued for another 10 min. The speed was further reduced to 8 rpm and the mixing process continued for another 10 min. After the mixing stage, COD of the supernatant was monitored with time.
163
3. Results and discussion Identification of the soil surface area, pore volume, and pore diameter, resulted in values of 173 m2/g, 0.16 c3/g, and 40.5 A, respectively. These results, particularly the surface area, show that the Khoweldi soil has a relatively good sorption ability. Since the most important soil constituent for the purpose of coagulation is clay, X-ray diffraction (XRD) tests were conducted, mainly to investigate the existence of clay in the Khoweldi soil. The results of the XRD analysis showed that the Khoweldi soil contained 68% quartz, 24% muscovite, and 8% hedenbergite. The XRD pattern for the Khoweldi soil is given in Fig. l(a) while the patterns of its constituents i.e. quartz, muscovite, and hedenbergite are shown in Figs. 1(b)-(d), respectively. Analysis of metal content in the soil has shown concentrations as high as 67450 and 39708 ppm of aluminum oxide and iron oxide, respectively. This is a very important finding because those oxides can form coagulants that helps with destabilization process of the polymeric colloids. The COD of the polymeric wastewater discharged by the factory was found to be 13500 mg/l, while the COD of the filtered sample (soluble COD) was found to be 130 mg/l. This clearly shows that the soluble COD is less than 1% of the total COD which is negligible and thus will not be considered in further work. Actually, this finding has motivated the main objective of this work which is destabilization of the emulsified colloids in the polymeric wastewater and not treatment of the soluble constituents. The high COD of the bulk sample emphasizes the need of pretreatment before the effluent can be discharged to the biological treatment plant of the industrial city. It should be remembered that the COD limit for industrial wastewater discharges to the aforementioned plant is 1000 mg/l. The settleability test of the wastewater revealed that even after a detention time of 24 h the colloidal waste was totally in suspension. This stressed the stable nature of the colloids under study and emphasized the need for a destabilizer to flocculate the colloids and subsequently have them clarified. In other words, conventional sedimentation tanks
164
N.S. Abuzaidet al. /Separation and Purification Technology 13 (1998) M-169
Fig. 1. X-ray diffraction patterns of the Khoweldi soil (a) and the constituents; quartz (b), muscovite (c), and hedenbergite (d).
14000.0 13000.0 12000.0 11000.0 lCOCO.O 9000.0 f
8000.0 7000.0
8 0
6000.0 6000.0 4000.0 3000.0 2000.0 1000.0 III 0.0
4.0
0.0
12.0
111
III
16.0
20.0
Time
(hr)
II, 24.0
II1 26.0
,,, 32.0
36.0
Fig. 2. Relationship between the supematant COD and time of sedimentation at different soil masses.
165
N.S. Abuzaid et al. / Separation and Purtjication Technology 13 (1998) 161-169
will not be able to remove the constituents of the wastewater under study. Fig. 2 depicts the relationship between the supernatant COD and time of sedimentation under the effect of six soil masses. It should be noted that the COD measurements started after the mixing stage. The figure shows that at any time the supernatant COD is inversely proportional to the amount of soil used. It is also shown that the lowest supernatant COD achieved was 675 mg/l which is acceptable by the standards of the industrial city. This value was achieved at a soil mass of 3 g after a detention time of 24 h. The aforementioned result shows that the problem of the wastewater discharges from the factory can be solved by the use of the Khoweldi soil as a destabilizer. It should be noted from Fig. 2 that equilibrium with time was reached within 24 h under all of the soil masses. It is worth mentioning that negligible COD reduction was found in the beaker with no soil. The interaction between the Khoweldi soil and the colloidal polymers in the wastewater may best be interpreted in terms of surface adhesion because of the large soil surface
area in general, and the existence of aluminum and iron oxides. When aluminum and iron oxides dissolve in water they are expected to become aluminum and iron hydroxides (coagulant agents). Regarding sedimentation, there are two possible mechanisms; discrete settling in which each particle settles by it self which is characteristic of the coarse portion of the soil, and zone settling in which fine particles settle as a blanket. The second mechanism is supposed to enhance sedimentation of the colloidal polymers by sweeping them by the settling zone. Removal of colloids in this manner is frequently referred to as a sweep-floe coagulation [ 31. This is of course in addition to the aforementioned polymers-particle adhesion process. Fig. 3 shows the temporal effect of different soil masses on the COD removal efficiency. The highest removal efficiencies (88-95%) were achieved when the soil mass was 3 g, while the lowest values (50-72%) were achieved under a soil mass of 0.5 g. Fig. 3 has a useful application regarding the design and operation of this process. If the process is to be designed for a certain removal efficiency, two parameters can be manipulated, namely, soil mass
0.85 0.80 '0 s
0.75
1 B
0.70
E d
0.65
:: 0
0.60
0.0
4.0
8.0
12.0
16.0
20.0
24.0
28.0
32.0
36.0
Time (hr) Fig. 3. The effect of different
soil masses on the COD removal
efficiency
as a function
of time.
166
N.S. Abuzaidet
al. /Separation
and Purification
Technology 13 (1998) 161-169
100.00 95.00
1
ii cz
A
A
90.00
A 85.00
A
A A
55.00 50.00
I 0.00
0.50
I
I
I
I
I
1.00
1.50
2.00
2.50
3.00
3.50
Soil Mass (g)
Fig. 4. Relationship between the ratio of COD removed during the first 1 h to the total COD removed and soil mass.
and detention time. For example, a removal efficiency of 70% can be achieved at detention times of 30 min, 8, 13 and 22 h, and soil masses of 2, 1.5, 1, and 0.5 g, respectively. It should be understood that the decision on which pair of these parameters to adapt in order to reach a given removal efficiency is mostly related to economy which is outside the scope of this work. In Figs. 3 and 2 it is clear that in the first 1 h a considerable portion of the removal efficiency was achieved under each of the soil masses and the ratio of the removal efficiency reached to the total removal efficiency increased with the increase in soil mass. Furthermore, this removal efficiency did not increase considerably in the first 6 h for all the masses used. During the time period (6-24 h), a considerable increase in the removal efficiency was achieved. The increase in the aforementioned period was inversely proportional to the soil mass. However, at equilibrium (within 24 h), the total removal efficiency attained was proportional to the soil mass. The kinetics of this process is very complicated because of the heterogeneity in both
the wastewater and the soil. However, in an attempt to explain the aforementioned findings the ratio of COD removed during the first 1 h to the total COD removed versus soil mass are plotted in Fig. 4. The figure generally shows that the percentage of the COD removed in the first 1 h to the total COD removed increases with the increase in soil mass up to a soil mass of 2.5 g after which the aforementioned percentage did not change. In order to explain this behavior, one should recall that the soil used was not pure clay but a mixture of mainly, muscovite and quartz. The non clay portion was higher in quantity than clay and naturally coarser and heavier. So, it is expected that most of the early COD removal (within 1 h) occurs mostly because of the non clay portion and clay starts settling with the attached polymers at a later time (after 6 h as was shown in Fig. 4). At a low soil mass, the amount of coarse constituents is low, resulting in a comparatively low percentage of early COD removal which will give chance to the fine constituents (clay) to remove more COD with time. However, at a high soil mass, the
N.S. Abuzaid et al. /Separation and PuriJcation Technology 13 (1998) 161-169
167
2000.0 0.0
4.0
8.0
12.0
16.0
20.0
24.0
28.0
32.0
36.0
Time (hr) Fig. 5. The variation of the amount of COD removed per mass of soil with time under different oil masses
comparatively high coarse constituents are responsible for higher percentage of early COD removals reducing the amount of COD to be removed by the fine portion. Above a soil mass of 2.5 g, the percentage of early COD removals reaches a fixed value of 95%. This means that 5% of the removable COD cannot be removed by the coarse portion of the soil regardless of quantity. From Fig. 4 it can be concluded that if efforts were made to increase the clay quantity in the Khoweldi soil by separation of quartz, both the kinetics and equilibrium capacity would be affected. As a result, kinetics may be lowered while higher equilibrium capacities are expected to be attained due to the comparatively large surface area of clay. In other words, less soil masses would be needed but at the cost of time. The temporal variation of the amount of COD removed per mass of soil with different masses of soil is depicted in Fig. 5. From the figure it is obvious that the rate of increase of the aforementioned ratio with time increases with the decrease in soil mass. At equilibrium (after 24 h) the amount
of COD removed per mass of soil can be defined as soil capacity. From the figure it is clear that the soil capacity is inversely proportional to the soil mass, i.e. higher colloids to soil mass ratios results in higher soil capacities. From Fig. 5 it can be concluded that both the kinetics and the capacity of soil coagulation for the waste under study increase with the decrease in soil mass. The effect of the variation in soil masses on kinetics was explained earlier while its influence on soil capacity will be explained soon. The relation between soil capacity and supernatant COD is depicted in Fig. 6. This type of illustration is very common in the adsorption studies and named an isotherm [ 11. The data of Fig. 6 demonstrated two distinct trends. Up to a supernatant concentration of 2500 mg/l, a linear relation between capacity and COD exists. Above the aforementioned COD, another linear relation is depicted. However, the slopes of the two trends are dramatically different. In fact, the slope of the later trend is larger than that of the first one. This kind of behavior is known in the activated carbon
168
N.S. Abuzaid et al. /Separation
and Purification
Technology 13 (1998) 161-169
20000.0
A
A A
4000.0
!I 500.0
A
A
looo.o
lsoo.o
2000.0
2500.0
3000.0
3500.0
4000.0
Residual COD (I@) Fig. 6. Relationship between soil capacity and supematant COD.
adsorption literature [3] and attributed to the complexity and heterogeneity of the wastewater which is typical of the waste under study.
4. Conclusions The Khoweldi soil was found to destabilize the colloidal waste resulting from a polymeric industry, and the reduction of the supernatant COD increased with soil mass and time until equilibration which was reached in 24 h. The lowest supernatant COD achieved was acceptable to the regulatory authority standards. The destabilization of the colloidal polymers was attributed to the adhesion enhanced by the large soil surface and the existence of aluminum and iron oxides. However, sedimentation was believed to occur because of discrete and zone settling. A considerable portion of the removal efficiency was achieved in the first 1 h and in the time range of (6-24 h), and the highest removal efficiency
achieved was 95%. Furthermore, the output of the study indicated that lower kinetics and higher equilibrium capacities are expected if the clay proportion in the Khoweldi soil is increased by separation of quartz. Additionally, the loading rate and the capacity of the soil were found to be inversely proportional to the soil mass. Finally, the above mentioned findings have colloidal polymeric important implications; wastewater discharges with high COD concentrations are amenable to purification by a cheep local material (soil). Accordingly, the results obtained can be extended to wastes and soils of similar properties to those investigated in this study.
Acknowledgment The authors thank the Research Institute, Ring Fahd University of Petroleum and Minerals, for providing support to this research.
N.S. Abuzaid et al. 1 Separation and Purification Technology 13 (1998) 161-169
References [I] N. Abuzaid, G. Nakhla, J. Envir. Sci. Technol. 28 (1994) 216. [2] G. Nakhla, I. Harazin, Envir. Technol. 14 (1993) 751-760. ]3] L. Benefield, J. Judkins, B. Weand, Process Chemistry for Water and Wastewater Treatment, End ed., Prentice-Hall, NJ, 1972. [4] A.P. Shu’ko, Neftepererab. Neftechim. (USSR) 8 (5) (1986). [5] A.P. Shu’ko, Chem. Abstr. 105, 158243~ (1986). [6] R.S. Soetopo, Berita Selulosa (Indon) 20(2) (1984) 46. [7] R.S. Soetopo, Chem. Abstr. 105, 11414a (1986).
[8] 0. Colak, B. Arikan, Doga:Mohendislik
169
Cevre Bilimleri (Turk.), lO(2) (1986) 185. [9] 0. Colak, B. Arikan, Chem. Abstr. 105, 1582565 (1986). [lo] C.T. Chion, S.E. Porter, D.W. Schmediting, Envir. Sci. Technol. 17 (1983) 227. [ 111 D.R. Garbarini, L.W. Lion, Envir. Sci. Technol. 20 (1986) 126331269. [ 121 N.L. Nemerow, Industrial Water Pollution: Origins, Addison-Wesley, Treatment, Characteristics, and Reading, UK, 1978. [ 131 American Public Health Association. Standard Methods; for the Examination of Water and Wastewater, 15th ed., 1980.