High air flotation efficiency of multiphase flow pump with the addition of dodecyl dimethyl benzyl ammonium chloride (DDBAC)

Recently, centrifugal multiphase pump–dissolved air flotation (CMP-DAF) has become an increasingly popular alternative to DAF that can achieve more stable performance and higher removal efficiency, and this method is widely used in sewage treatment. However, the nonuniformity of the bubble size and low adherence of the floc particles and bubbles, as well as the complicated raw water quality, pose great challenges to CMP-DAF, which does not meet the standards of water supply and drainage in practical use. In the present study, the surfactant dodecyl dimethyl benzyl ammonium chloride (DDBAC) was utilized as a flotation agent to further improve the flotation efficiency of the CMP-DAF process. DDBAC at a dosage of 0.2 mg/L was introduced to the air flotation of raw water to construct a flotation enhanced CMP air flotation system. The results showed that the average turbidity decreased to 0.433± 0.017 NTU, and effluent floc particles were present at 1,053 cnt/mL with an acceptable removal rate of 96.20%. In addition, 34.0% and 30.1% of UV254 and CODMn were removed, respectively. These results imply that DDBAC can increase the collision efficiency of bubble particles by reducing the diameter of the bubbles, which is conducive to forming larger flocs, and enhancing the shear resistance of the bubble–floc particles, thus improving the air flotation efficiency.

In the present study, dodecyl dimethyl benzyl ammonium chloride (DDBAC) was utilized as flotation agent to further increase the removal efficiency for different pollutants during the CMP-DAF process. First, we systematically investigated the removal rate of different pollutants before and after the addition of DDBAC. Second, the change in bubble size before and after the addition of DDBAC was studied. Finally, we analyzed the removal mechanism of DDBAC-assisted CMP-DAF (DCMP-DAF), which is shown in Figure 1.

Raw water quality
The raw water was supplied from the Yellow River, which has operated stably for ten years. The water quality parameters are shown in Table 1.

Pilot plant and test method
The pilot-scale experiment setup ran continuously and was located at the Quehua Water Plant in Jinan, which belongs to the National Science and Technology Major Special Pilot Base. The experimental setup is shown in Figure 2. The experimental setup was composed of a mixing tank, flocculation chamber, contact zone, floating zone and tube settler. The overall size of the floating zone and the tube settler was 2.6 m × 0.8 m × 4.3 m. The test lasted for 20 days from December 20 to January 10, when the raw water was characterized by low temperature and low turbidity as shown in Table 1. The operating pressure, reflux ratio, and vacuum of the CMP were 0.5 MPa, 20% and À0.01 MPa, respectively.

Test process
The two chemicals (poly-aluminum-ferric chloride (PAFC), DDBAC) are added to the raw water. Then, the raw water runs through the mixing tank, flocculation chamber, and contact room and enters the floating zone. Subsequently, the water passes through the tube settler separation zone and flows into the reflux tank. Water and air are sucked into the CMP, and a uniform mixture of water and air is formed. The mixture passes through the expansion tube and is released from the gate valve in the contact chamber.

Analysis methods
Turbidity is measured by a portable turbidimeter (Hach, 2100N). The number of floc particles is determined using a benchtop particle counter. COD Mn is measured by the potassium permanganate method. UV 254 is investigated spectrophotometrically (Hach, DR5000). Total organic carbon (TOC) is determined by a Shimadzu total organic carbon analyzer (Shi-madzu, TOC-VCPH). The average bubble particle size is measured according to Stokes' theory, and calculated by Equations (1) and (2)   where H is water height in the cylinder, T is the time that the bubble floated to the surface of the water completely, d is the diameter of the microbubble, μ is the viscosity of the liquid or slurry, V is the rise velocity of the microbubble, g is the gravitational acceleration, ρ is the density of the liquid-microbubble mixture, and ρ g is the density of the microbubble gas.

Removal efficiency of turbidity
The effluent turbidity and removal rate of CMP-DAF and DCMP-DAF are shown in Figure 3. As obtained, the effluent turbidity of CMP-DAF varied from 0.519 ± 0.026 to 0.765 ± 0.03 NTU. When 0.2 mg/L DDBAC was added to the air flotation system, the average effluent turbidity decreased from 0.608 ± 0.058 NTU to 0.433 ± 0.017 NTU. In addition, it was also found that DCMP-DAF had better control of effluent turbidity and better impact resistance to the complicated raw water, which resulted in more stable effluent turbidity.
In contrast, the effluent turbidity trend of CMP-DAF corresponded to that of the raw water. When the turbidity of the raw water increased, the effluent turbidity of CMP-DAF increased and vice versa.
The difference between the effluent turbidity of CMP-DAF and DCMP-DAF was most likely induced by the addition of DDBAC. DDBAC is an amphiphilic surfactant, and its structure is shown in Figure 4(a) (Hu et al. ).
When DDBAC was introduced to the emulsion, the hydrophilic groups adhered to the surface of the floc particles, and the hydrophobic groups adhered to the surface of the nanobubbles (Figure 4(b)). Since the hydrophobic group Fortunately, DDBAC is an amphiphilic surfactant, and the addition of DDBAC in the DCMP-DAF system was conducive to enhancing the interaction between hydrophilic floc particles and bubbles, which is described in Figure 6.   Table 2.
As obtained from Table 2, the addition of DDBAC to the system could reduce the diameter of the bubbles in the air-floating separation zone. When the concentration of  When the content of DDBAC was 0.2 mg/L, the average diameter of the bubbles in DCMP-DAF was 39.03 ± 0.42 μm, which was 7.05 μm lower than that in CMP-DAF.
However, it was obvious that the diameter of the bubble remained unchanged as the content of DDBAC further increased. Rykaart & Haarhoff () proposed that the bubble generation process mainly went through two stages: (1) bubbles are generated and grown on the bubble core of dissolved gas water, which is released through high pressure; (2) in the air-floating separation zone, bubbles gradually grow as they coalesce. In the general process of model calculation, certain assumptions were made. For the traditional DAF and CMP-DAF, the bubble sizes generated ranged from 30 to 100 μm, which can be simplified as a spherical shape.
According to the theory of complementary nuclei, when the two phases of gas and liquid are not saturated, the saturation required to form bubbles is due to the lack of gas cavities in the solution (Xi et al. ). However, once bubbles are generated, new bubbles are hardly formed at the same position.
We can calculate the critical diameter of the bubbles using the following Equation (3) (Edzwald ): where d cd is the critical diameter of the bubble, σ is the surface tension of water, and ΔP is the pressure difference on both sides of the dissolved water releaser.
It can be seen from Equation (3) Table 2, after the addition of DDBAC, the bubble diameter decreased to a certain extent, which was believed to increase the collision efficiency and improve the air flotation efficiency.
When a small amount of DDBAC was added, the surface tension of the liquid was first reduced. According to the theory of complementary nuclei, the critical diameter of the generated bubble group was reduced when bubbles precipitated at the nucleation site (Xi et al. ). Since the hydrophobic end of the DDBAC molecule was more easily bound to the bubble, a layer of DDBAC molecules was adsorbed on the surface of the bubble, which was embedded in the water film around the bubble, as shown in Figure 8(a). The van der Waals force between bubbles was weakened and the hydrodynamic repulsion was strengthened; thus, the coalescence between bubbles was weakened, and the average particle size of the bubble was reduced (Sobhy & Tao ). In contrast, when the concentration of DDBAC molecules in water was too high, the adhesion mode of DDBAC changed, as shown in Figure 8(b). These bubbles gathered and adsorbed to each other and then formed micelles that were stable in water. The micelles had little effect on the aggregation of bubbles, and the particle size of the bubbles remained unchanged at a certain level, which was evidenced to have a slight influence on the removal efficiency for turbidity, floc particles and organic matter (Basarǒvá et al. ).