Competitive Stripping of Multi-Organic Pollutants from Contaminated Water in Bubble Column Semi-Batch

Air stripping for removal of Trichloroethylene (TCE), Chloroform (CF) and Dichloromethane (DCM) from water were studied in a bubble column (0.073 m inside dia. and 1.08 m height with several sampling ports). The contaminated water was prepared from deionized water and VOCs. The presence of VOCs in feed solution was single, binary or ternary components. They were diluted to the concentrations ranged between 50 mg/l to 250 mg/l. The experiments were carried out in batch experiments which regard the bubble column as stirred tank and only gas was bubbled through stationary liquid. In this case transient measurements of VOC concentration in the liquid phase and the measured concentration profiles were modeled by bubble aeration model (BAM) to fit the experimental data fairly well. The results from batch experiments show that the removal efficiency of VOCs increases with increasing gas flow rate or gas holdup. It is found a pH=10 give the best removal rate, but all experiments were adjusted at pH=8 which allow to study other operating conditions. TCE is being removed faster than the other two components for all systems and a single component was removed faster than binary or ternary system. The K L a values were evaluated by fitting the BAM to the experimental data. It is found that K L a increased with increasing gas flow rate and TCE exhibits the highest K L a values.


INTRODUCTION
Ground water contamination by trace levels of volatile organic compounds (VOCs) has emerged as a major water quality issue in the last five decades.Interest in VOCs in drinking water began escalating in the mid-70s when it became possible to measure these compounds concentrations µg/l levels.Since then, it has been recognized that volatile compounds are finding their ways into ground water, which is often used as the main source of drinking water, Djebbar 1998.
VOCs are organic compounds with a tendency to migrate or diffuse from the aqueous phase (water) to the gaseous phase (air), or viceversa.They are generally highly volatile, with low molecular weight.VOCs are a widely used class of compounds employed in many types of industrial, commercial, agricultural, and household activities,DeMarco 1983.
These compounds find their way into the environment through industrial and municipal wastewater discharges, leachate from deficient, old waste disposal sites that received hazardous chemicals, illegal discharges, and leaks from above or underground storage tanks and pipes.Contamination of drinking water supplies may occur from any combination at relatively high concentrations, and several other compounds will be present at lower concentrations ,Dyksen et al., 1982.
A very prevalent water pollution problem pertains to contamination by VOCs that offer potentially very important health risks, when they have a certain propensity to evaporate away from the water into air.Many of these VOCs are only partially miscible with water but in general they all present certain solubility.The problem of reducing VOCs in water applies to ground waters, surface waters, and waste waters alike.The origin of the water has some important design implications that will be discussed later.This write-up will discuss the application of air stripping for the removal of VOCs from water.
Depending on the contaminant and site characteristics, volatilization can be an important mechanism for the loss of hazardous compounds from soils and liquid waste systems.Volatilization may decrease the concentrations of the wastes at a facility or site; however, because of the law of mass conservation, contaminants will subsequently be found in the atmosphere.Therefore, air emissions, may become hazardous air pollutants (HAPs) and result in short or longterm health effects.Volatilization can also be applied as a hazardous waste treatment process to clean up contaminated ground water through air stripping and to remediate soils by soil vapor extraction.If volatilization occurs, it will do so primarily as a function of the contaminants vapor pressure (if the compound is in relatively pure form) and the Henry's Law constant (if it is in aqueous solution), Richard, 1997.
One of the most useful methods to remove the VOCs from contaminated water by using the dispersion aeration (bubble column technique) which include flow of air bubbles in water, and mass transfer from liquid phase to gas phase will occur (VOCs move from the water to the air as the bubbles rise through the water).Several types of air stripper technologies are currently available for removal of VOCs from water.This evaluation focuses on the capabilities and limitations of several major established and emerging air stripper technologies that are potentially applicable for VOCs removal such as Packed tower Aeration, low profile aeration, bubble diffusion aeration, spray towers, aspiration ,Andrew et al., 1995.
Transfer of the VOCs from the water to the air can be improved by increasing the column length or by producing smaller bubbles.Its main advantages are that it is simple and that it can handle water having high levels of suspended solids, Dwight, 2001.

As far as
published studies are concerned, the main interest is concentrated on design and scale-up, fluid dynamics and regime analysis and characteristic parameters, especially gas holdup, bubble characteristics, mass transfer coefficient and heat transfer coefficient., the effects of superficial gas velocity, liquid properties, operating conditions , column dimensions, gas distributor design , solid type and concentrations are presented,

2.OBJECTIVES
The main objective of this work is to investigate competitive removal of identified VOCs (TCE, CF and DCM) pollutants from water by air stripping batch wise, and to investigate stripping of dangerous VOCs from wastewater by air using bubble column from liquid phase to gas phase through single, binary and ternary systems of VOCs in water, also to determine the volumetric mass transfer coefficient K L a for a particular component as single and in the presence of other components (binary and ternary systems).

3.MATHEMATICAL MODEL
For a VOC with a Henry's law coefficient much greater than approximately (1.6*10 -4 atm.m 3 /mole at 25 °C), however, the gas-film resistance is negligible compared with the liquid-film resistance ,Mackay and Leimonen, 1975.
K L a values for VOCs are not easily predicted because they are highly system dependent.However, a relative value of K L a can be computed because K L a is proportional to the VOCs diffusion coefficient in water, which is proportional to the molar volume at its normal boiling point ,Wilke and Chang, 1955.
As mass transfer occurs between water and air in a bubble, the VOC concentration in a bubble increases as the bubble moves through the water column to the water surface.This phenomenon was considered by Matter-Muller et al., 1981 and Roberts et al, 1983, who developed a kinetic model for bubble transfer with the following assumptions; the system is isothermal; mass-transfer coefficients are constant and independent of time and location, if hydraulic conditions are stable; equilibrium follows Henry's law; complete mixing occurs in the water; the change of VOC concentration in water during the retention time of a bubble in the water is insignificant compared with concentration change in a bubble; and although the upward velocity, surface area, and volume of a bubble depend on the actual depth of the bubble, use of the mean values over that total depth is a acceptable .
The bubble-transfer model is based on the continuity equation for a VOC in a bubble rising through the water ,Matter-Muller et al,.1981.
Integrating Eq. (1) yields a relationship between water depth and the mole fraction of VOC in the bubble phase.
(2) When Eq. ( 2) applied for the entire bubble rise for typical unsteady-state gas transfer, C L becomes a function of the aeration time (t), according to:

EXPERIMENTAL SET UP
Fig. 1 the experimental set up which consists mainly of a pixglass bubble column 73 mm inside dia.and 1.08 m height with several sampling ports, the column; a feed pump, a feed tank, a stock solution tank, an air compressor and air rotameter.The distributor plate is a 1/8 in metal sheet having 50 orifices of 0.5mm in diameter arranged on a square pitch.mm polyethylene.The liquid line is a plastic pipe with nominal diameter of 1/2 inch.The column has a two 3/4 in overflow pipes, and all fittings are either PVC or galvanized steel.The gas holdup was measured according to the volume fraction of the gas phase occupied by the gas bubbles, which defined as the volume of gas phase divided by the total volume.

MATERIALS
VOCs used for the experiments were trichloroethylene TCE, chloroform CF and dichloromethane DCM.Alternatively DCM is known as methylene chloride.All these chemicals were of purity greater than 99.6 % and they were used without further purification.In this study deionized water was used for the preparation of the feed solution, i.e. aqueous solution of VOCs.Due to a low solubility of these VOCs in water, high purity methanol was used as a co-solvent which assisted in obtaining homogeneous solutions of VOCs in water.
The physical-chemical properties of the three VOCs used in the present work are summarized in Table 1.

PREPARATION OF FEED SOLUTION
Contaminated water was used as feed solution.This was prepared by dissolving VOCs in deionized water.However, VOCs chosen for this study are of little solubility in water and consequently the aqueous solutions produced were heterogeneous.
Fatima et al, 2009, suggested the use of a co-solvent before mixing with water.Their procedure is simple, in which 1gm of a VOCs e.g TCE or CF or DCM is dissolved in 10 mL of high purity methanol e.g > 99.6% purity and the mixture is added to 10L of deionized water.This step yielded a homogeneous aqueous solution of a single component of VOC of concentration 100 mg/l.Employing this technique, aqueous solutions of single VOC of initial concentrations of 50,100,150,200 and 250 mg/l were prepared for the experiments.By a similar method binary and ternary solution were prepared with initial concentration of 50,150 and 250 mg/L of each component.The use of methanol proved to be an effective solvent which assisted in the dissolution of VOCs in water with the result that homogenous aqueous solution were obtained.Methanol concentration is too low to have an effect on desorption performance and it´s volatilization during bubble aeration processes can be neglected according to blank test.

ANALYTICAL TECHNIQUES
Water samples drawn from bubble column during the course of experiments were analyzed to determine VOCs concentration.Single component samples were analyzed using the UV-spectrophotometer system at 190-1100 wavelength spread .The injection volume of samples was 5 ml.For the two-component systems binary systems the spectro-photometric method is also used but the technique is different from that used for single-component systems.The bivariate calibration method presented by Pedrol.et.al ,1997, was applied for the simultaneous determination of two VOCs components in a mixture, particularly when the two components have close wavelengths.

AIR STRIPPING EXPERIMENTS
All stripping experiments were carried out under isothermal conditions of temperature ranging within 22-28 C° average 25C°.The experiments were operated as semi-batch wise by allowing air to bubble through non flowing water.The experiments involved desorption of single component solutions.To investigate competitive desorption, binary and ternary experiments were done for this purpose.The removal efficiency of VOCs and values of the mass transfer coefficient K L a could be determined.
In these experiments, the stripping air was allowed bubble through stationary water at different rates of flow.In each run about 4L of wastewater aqueous solution of VOCs were charged into the column.Different flow rates of air were used for the experiments.The pH of solution was varied covering five values 3, 5, 7, 8, and10 and pH was adjusted by sodium hydroxide or hydrochloric acid.After initialization the air flow water samples were withdrawn at discrete time intervals through a sampling port in the middle of the bubble column.The samples were immediately kept in sealed vials 10mL volume and placed in a cold box to prevent volatilization.The sparging process was terminated when the VOC concentration in the column was low or relatively constant.In most of the runs steady state   The removal rate or removal efficiency for each run was calculated as the difference between initial and residual concentrations.The results shows that a pH =8 was the most favorable value for recovering VOCs from wastewater consequently.

RESULTS AND DISCUSSIONS
A series of desorption tables and curves were obtained experimentally for air stripping from water of single, binary and ternary systems in a bubble column.In all experiments the overall mass transfer coefficient (K L a) was obtained by Bubble Aeration Model (BAM), the results were obtained by solving the model by using STATISTICA software computer program.
The mass transfer coefficient (K L a) assumed as an unknown parameter, was obtained from the model using the experiment data.The experimental and predicted curves are presented as plots of removal factors (C L /Co) versus time.
Several experiments were carried out to investigate the effect of different experimental parameters on the removal efficiency of VOCs from water.The parameters studied were pH, air flow rate and the initial concentration of solutions for single, binary and ternary systems of VOCs.

1-Effect of pH
The effect of pH on the removal of TCE ,CF and DCM was studied in single system as shown in figures 2, 3 and 4 respectively plotting the concentration ratio C L /Co versus time at various pH values while the other variables remained constant.From these three figures it can be seen that concentration ratio decreases sharply at the beginning of the runs, then it began to decrease slowly with time and it was found that the best pH value for the removal of all components to be 10 highly alkaline but all experiments were adjusted at pH=8 for laboratory purposes to control experiments easily and to be easy to study another operation conditions.

2.Effect of Air Flow Rate
The removal of VOCs from water was studied at different gas flow rates 1, 2, 3, 4, 6 l/min with the pH adjusted at 8, and for three systems single, binary and ternary.The results are presented in Figs. 3, 4, and 5, as plots of C L /Co versus time with air flow rate as parameters.It can be seen that high removal occurs with higher gas flow rates.This is due to the fact that the hydrodynamics of bubble columns depend on superficial gas velocity U G .If U G increases hinder liquid flow indicating that the Liquid is getting more agitated and hence more removal.At low gas velocity, bubbles are small, and rise uniformly without much interaction with the liquid.On the other hand, at higher gas velocity, large fast-rising bubbles appear which disrupt the system contents, resulting in an increase in the liquid dispersion, Moustiri et al., 2001.Thus increasing the gas flow rate increases the removal of volatile solutes by gas transfer in the range of interest for both surface and bubble aeration ,Paul et al. 1985.
The first point of significance from these plots is the systematic increase in the extent of VOCs removal with increase in air flow rate.The general shape of these curves is as expected based on the solution of the mass transfer equation BAM, it was an exponential removal rate with time.The effect of air flow rate is depicted in figures above and as found with single the binary and ternary the removal efficiency was increased with increasing air flow rate for all components in the mixture.Similarly, TCE was removed faster than the other components.

3-Competitive Desorption:
To show the effect of operating conditions on desorption experiments of a component by the presence of other components in the mixture; TCE is taken as a case.Fig. 6 and Table 3 illustrate the removal rate of TCE as a single system, as a binary system first with CF and second with DCM and as ternary system TCE-CF-DCM.It can be seen from the curves of experiments, TCE is being removed at a faster rate in a single system in comparison with binary and ternary cases.For example, after 700 s of operation, 82% of TCE is removed in the single case, whereas 68% is removed in the presence of CF and 70% is removed in the presence of DCM.In the case of ternary system, 66% of TCE is removed over the same period of time and operating conditions.Another example taken near the end of stripping operation (nearly after 2000 s), we note that 99% of TCE is removed in the single case, 96% and 97% with CF and DCM respectively, and 93%in the ternary case.It is clear that TCE in all these cases is the components with highest tendency for removal.The closeness of the removal CF and DCM is possibly due to the closeness of their Henry's constant and their molecular weight.

4-Evaluation of the Volumetric Mass Transfer Coefficient (K L a)
The overall liquid phase mass transfer coefficient for each compound (as single, binary, and ternary systems) was evaluated by applying the experimental data to the aeration bubble model Eqs. 4 and 5 using STATISTICA program.The results of calculations are presented in Tables 4, 5 and 6, for all systems investigated.Fig. 7 presents the results for single component systems which show that K L a increases with increasing gas flow rate or gas holdup for the three VOC compounds.This result indicates the dependency of K L a on bubbling behavior created by gas flow through the liquid phase.A large number of small size bubbles observed at high gas flow rate caused the creation of large interfacial area for mass transfer and in turn this led to higher values of K L a.As can be seen from Fig. 7, TCE exhibits the highest value of K L a in comparison with other two VOC compound and the order is:          .
The slope of the relationship between In(C L.t /C L.O ) versus t of Eq. (4) is expressed as :(4) Because ф in Eq. (4) is a function of (K L a) b , (K L a) b of any VOC i can be obtained from Eq. (5) as rearranged to solve for (K L a ) b,i(5)
probably due to difference in diffusion coefficients of VOC compounds in the liquid phase.K L a is directly proportional to the Henry's law constant and molecular diffusion in the liquid ,cited by Wilke and Change, 1955, i.e. is the diffusivity coefficient of VOC in the liquid phase.And from Table3, we noted that experiments of VOCs with mechanical stirring and found that K L a values increased when stirring speed increased.Mechanical stirring has improved the interfacial area by producing large number of small size bubbles.Figs.8, 9and 10 present the results for binary systems.As can be seen, the highest value of K L a is exhibited by the component with the highest value of Henry's constant , we note that CF and DCM exhibit weak competition, i.e. their K L a values are close and this is probably due to the closeness of their diffusion coefficients.This discrepancy in the results is possibly due to variations in experimental conditions, particularly temperature which is difficult to control.The results shown in Fig.11 for the ternary system indicate weak competition among the three VOC compounds, particularly at low gas flow rate.At higher air rate, value.The effect of flow rate on K L a value is in agreements with Matter-Muller et al., 1981, who carried out batch desorption.

Figure 4 .
Figure 4. Effect of air flow rate on the removal of CF single system pH=8, C Lo =250 mg/l, T ≈ 25°C.

Figure 5 .
Figure 5.Effect of air flow rate on the removal of TCE in TCE-CF binary system, C Lo = 250 mg/l, pH=8, T ≈ 25°C.

Figure 6 .
Figure 6.Competitive desorption of TCE in presence of CF or DCM in binary system and CF-DCM in ternary, Q G =2 l/min, C Lo =250 mg/l, T ≈ 25°C.

Figure 7 .
Figure 7. Effect of air flow rate on K L a single system component C Lo =250 mg/l , pH= 8, T≈ 25°C.

Figure 8 .
Figure 8.Effect of gas flow rate on K L a in binary system (TCE-CF) component, C Lo =250 mg /l , pH= 8, T =25°C.

Table 1 .
Physical and chemical properties of VOCs.

Table 2
summarizes the operating conditions used in the experiments.

Table 3 .
Competitive desorption of TCE in presence of CF or DCM in binary system and in presence of CF-DCM in ternary system, Q G =2 l/min, C o =250 mg/l, T≈25°C.

Table 1
the D l values of DCM and CF is rather close and the diffusion coefficient in a solution could be different from that in pure solvent.The observed effect of gas flow rate on Mechanical stirring has improved the interfacial area for mass transfer by producing large number of small size bubbles.
K L a is in agreements with Matter-Muller et al., 1981 who carried out batch desorption experiments of VOCs with mechanical stirring and found that K L a values increased with stirring speed.

Table 4 .
K L a and gas holdup for single experiments.
 