Optimizing dissolved air otation design and saturation

L.A. Fris, C.W. Gallina, R.T. Rodrigues and J. Rubio

Departamento de Engenharia de Minas-PPGEM-Laboratrio de Tecnologia Mineral e AmbientalUniversidade Federal do Rio Grande do Sul, Av. Osvaldo Aranha 99/512, 90035-190, Porto Alegre, RS, Brazil (E-mail: laferis@vortex.ufrgs.br ; jrubio@vortex.ufrgs.br; http://www.lapes.ufrgs.br/Laboratorios/ltm/ltm.html) Abstract Dissolved air otation (DAF) of iron hydroxide precipitates at working pressures lower than 3 atm, using modied otation units to improve the collection of fragile coagula, was studied. Conventional DAF otation was studied as a function of saturation pressure in the absence and presence of surfactants in the saturator. Without surfactants, the minimum saturation pressure required for DAF to occur was found to be 3 atm. But, by lowering the air/water surface tension in the saturator, DAF was possible at a saturation pressure of 2 atm. This behavior was found to occur in both batch and pilot DAF operation tests and almost complete recovery of the precipitates was attained. Results are explained in terms of the minimum energy which has to be transferred to the liquid phase to form bubbles by a cavity phenomenon. Further, studies were conducted changing equipment design and feed bubbles size distribution (mixing micro and midsized bubbles). Thus, bubbles entrance position in the collision-adhesion zone (capture zone) was compared to bubble entrance position in the water ow inlet below the oating bed. A mushroom type diffuser was used for the capture zone experiment and better performance was obtained. Results are explained in terms of different mass transfer phenomena in the collection zone and in the separation zone. Finally, results obtained with the use of a column otation cell working as normal DAF and with a wide bubble size range are presented. Results indicate good performance and some gains in process kinetics with middle size bubbles. Keywords Dissolved air otation; process optimization; bubbles

Water Science and Technology Vol 43 No 8 pp 145157 IWA Publishing 2001

Introduction

The use of flotation as a solid/liquid or liquid/liquid (or both) separation process is widely used in domestic wastewater treatment. Modern flotation devices and schemes exhibit both high throughput and efficiency. It is believed that this process could be incorporated as a clean technology to treat industrial wastewater and acid drainage waters in a variety of industrial fields, including industries where the process is not presently used for water treatment (Matis, 1995; Rubio, 1998; Da Rosa et al., 1999). Dissolved Air (Pressure) Flotation (DAF), is a well-established separation process that uses micro-bubbles as a carrier phase to remove solids, ions, macromolecules, fibers and other materials, and with the reduction of BOD (Biological Oxygen Demand), COD (Chemical Oxygen Demand) and for sludge thickening (Ives and Bernhardt, 1995). Flotation offers process advantages over such processes as filtration, precipitation, or adsorption onto natural and synthetic sorbents. The advantages include better treated-water quality, rapid startup, high rate operation, and a thicker sludge, shorter sludge residence time in a secondary clarifier stage and flexibility in changing the operation parameters as the result of active sludge separation. In DAF bubbles are formed by a reduction of pressure of a water stream previously saturated with air (or other gas) at pressures usually higher than 3 atm. In industrial practice, the supersaturated water is forced through needle valves, and clouds of bubbles having 0.020.10 mm in diameter are subsequently produced just down-stream of the constriction. Disadvantages include the high cost of water saturation required for bubble formation, as well as the problem of breakage (rupture) of the coagula and small (or even colloidal)

145

particles in the collection zone. In general, micro-bubbles are required and because they are usually aggregated colloids rather than dispersed ones, high shear rates must be avoided to obviate destruction of the aggregates. In the present study, DAF was conducted using working pressures of less than 3 atmospheres. This was made possible by lowering the air/fluid interfacial tension in the saturator, and subsequently in the nozzle where the bubbles are generated (Fris and Rubio, 1999). The work was conducted in batch and small pilot units using Fe(OH)3 as a fragile colloidal precipitate. Then, cell design (location of diffusers) and bubbling features affecting the capture of particles by mid-sized bubbles were studied.
Experimental
Materials and reagents

L.A. Fris et al.

146

FeCl3.6H2O was employed as solute to form the hydroxo-precipitates, which served in this work as the fragile colloidal precipitate. Dowfroth 1012 (a commercial frother) and Sodium Oleate (an anionic collector and frother) were used to modify the solution/air surface tension. Solution pH was adjusted using NaOH.
Methods: dissolved air otation

DAF tests were carried out using batch and pilot units (Figures 1 and 2). Figure 3 shows the DAF modified unit endowed with the two different types of diffusers. Figure 4 illustrates a column type of cell (0.3 m diameter, 1.4 m high) which includes:

Figure 1 Dissolved Air Flotation (DAF) pilot unit. Cell dimensions: 0.46 m high, 0.880.22 m length

Figure 2 Batch Dissolved Air Flotation (DAF) unit. 1.5 L otation cell and a 2.2 L saturator

L.A. Fris et al.

Figure 3 DAF cell showing both conventional and mushroom types diffusers

Figure 4 Column otation cell operating with micro- and mid-sized bubbles

feed, located 6 cm from the top, via an special feeder endowed with various entrance holes; an entrance for the recycled water from saturator, with micro-bubbles; an external air bubble generator operating with a surfactant/frother to control the bubble size. In the presence of surface-active reagents micro-bubbles (0.10.6 mm) can be obtained (Yoon et al., 1992; Yoon and Luttrell, 1994). In this cell recycled water is pumped through a nozzle flow constrictor for air suction. Process efficiency was evaluated by measuring the residual content of iron (using atomic absorption) and the surface tension was monitored with a ring tensiometer (Krss model 8431). Table 1 shows all operational parameters and conditions used in the experiment.

147

Table 1 Experimental parameters and conditions in the otation of iron hydroxide

Rates

Pilot units Parameters DAF Column

Batch unit DAF

Saturation pressure (atm) L.A. Fris et al.

148

23 20 30 0.05 0.16 48.2 57 2.9 4 0.1 0.2 30 65 30.5 10 18 Precipitate dispersion, Qe Sodium oleate (SO) at the saturater Fe+3 SO at the saturater SO in the precipitate dispersion Dowfroth 1012 at the otation cell

3 30 0.08 57 3.4 4 65 10 20 27

1.8 4 20 0.09 0.22 59.3 57 30 15 13

Recycle ratio (%) Air/Solid rate at 25C, as (kg.kg1) Interfacial tension in the saturated recycle (mN.m1)* pH Supercial ow rates, (m3.m2.h1) Feed ow rate (L.min1) Concentration (mg.L1)

Residence time (min) * Water interfacial tension = 73 mN.m1

Table 2 Effect of working pressure on removal of Fe(OH)3 by DAF (batch) without surfactants in the saturator. Other conditions the same as in Table 1. 1 atm = 1.13.105 Pa
CFe (mg.L1) Flotation rate (cm.s1)

Pressure (atm)

Initial

Final

Removal (%)

1.8 2 3 4

30.6 29.9 29.6 29.9

30.6 29.9 2.1 9 1.0

0 0 3.0 96.5

0 0 0.13 0.14

Table 3 Effect of working pressure on removal of Fe(OH)3 by DAF (batch) with sodium oleate addition in the saturator. Other conditions the same as in Table 1. 1 atm = 1.13.105 Pa
CFe (mg.L1) Flotation rate (cm.s1) as (kg.kg1)

Pressure (atm)

Initial

Final

Removal (%)

1.8 2 3 4

29.8 29.3 29.8 30.4

3.7 2.8 2.5 0.9

87.7 90.3 91.5 97.2

0.075 0.11 0.13 0.14

0.09 0.10 0.16 0.22

Results and discussion

Batch studies

Table 2 shows that, without surfactants, the minimum saturation pressure required for DAF to occur was 3 atm. However, by lowering the air/water surface tension, DAF was possible at 2 atm. Table 3 shows that a minimum air/solid ratio of 0.10 was required to achieve flotation rates of about 0.11 cm.s1. Micro-bubbles are formed by the precipitation of gas from the saturated water. The excess of pressure over the friction loss provides the energy for bubble nucleation. When bubbles are formed in pure water, where little pressure transfer exists, large bubbles are produced. This is the reason why DAF, operating at less than 3 atm in the absence of

frothers, does not provide sufficient energy to overcome attrition and nucleation to yield bubbles. However, a decrease in solution/air interfacial tension provides low energy nucleation sites and minute bubbles are generated. Results obtained in this study and in a previous work (Fris and Rubio, 1999) show that the minimum energy, F, to be transferred to the liquid phase to form bubbles by a cavity phenomenon (arising from the liquid turbulence) is given by the following equation (Takahashi et al., 1979): F =
16 3

L.A. Fris et al.

3 ( Po Pa )3 ( joules)

where

= air / water surface tension (Nm 1 ); Pa = atmospheric pressure (atm or Pascal units);
Po = saturation pressure (atm or Pascal units) Thus, the energy required to generate micro-bubbles will be smaller with lower air/liquid interfacial tensions and with higher pressure differences of the liquid phase with respect to the atmosphere. Thus, by lowering the air/liquid interfacial tension, the smaller will be the liquid/solid attrition, the faster will be the flow fluid velocity and the bubble formation. Dupre et al. (1998) reported that DAF users observe a reduction in the diameters of the bubbles when using polyelectrolytes. However, energy transfer phenomena in the bubble formation was not considered. It is suggested that the use of polymers or surfactants in the pressure vessel will drastically reduce energy costs in DAF operations. Moreover, DAF operations at working pressure lower than 3 atm might also be safer and the saturation rapid and efficient. The surfactant or polymer can be the same collector or flocculant used to improve DAF kinetics via enhancement of the bubble/particle interaction (should the reagent increases particle hydrophobicity). The amount required of these reagents is small and will not influence operational costs.
Pilot scale studies

Conventional DAF cell. Table 4 shows results obtained with surfactant addition in the saturator at two different diffusers positions. One is in the collection (inclined recycled inlet) zone (the conventional), and another is in the separation (mushroom type) tank. In the first case, bubbles act as nuclei for the attachment of the precipitates, providing a transport mechanism for the movement toward the separation zone. Here, the aggregates should withstand resistance to settling, depending mainly on the air/solids ratio. The mushroom type of diffuser yielded better results than did the conventional diffuser because losses caused by the aggregates settling are smaller (see Table 4). Instead, a full white water bed of bubbles below the precipitates guaranteed good performance. Moreover, no lifting to the surface is required (the particles already enter at the top of the cell). However, some dead areas were observed when no bubbles were injected in the recycled inlet. At working pressures less than 3 atmospheres, when the air-to-solids ratio was low, none of the diffusers in the absence of surfactants yielded exceptionally good results. Using the conventional diffuser, not many particles were lifted due to the low capture (collisions and bubble-particles adhesion) of precipitates by bubbles. More, in the separation zone, the floated aggregates settled quite rapidly. However, when the recycle inlet flux was divided equally into two parts (50% in each diffuser), the performance improved somewhat, even at 2 or 2.5 atmospheres (see Figure 5 and Table 5). This modification enhances the collection of fragile coagula by bubbles, avoiding aggregate rupture and settling because the mushroom type of diffuser creates a cloudy like bubble bed, below the floating particles.

149

L.A. Fris et al. Figure 5 Removal of ferric hydroxide by DAF at different working pressures. Conditions as in Tables 4 and 5 Table 4 Removal of Fe(OH)3 by DAF (pilot unit). Conditions: P=3atm, 20% recycle ratio, sodium oleate addition only at the saturator. Other conditions see Table 1. CD = conventional diffuser; MD = mushroom type diffuser
CFe (mg.L1) Removal (%) T (min) CD MD CD MD

0 15 30 45

57.2 1.5 1.3 1.1

48.4 1.3 0.4 0.6

0 97.2 97.6 98.0

0 99.2 98.8 99.0

Table 5 Removal of Fe(OH)3 by DAF using both diffusers (pilot unit) with sodium oleate addition only at the saturator, 20% recycle ratio. Other conditions the same as in Table 1. 1atm = 1.13.105 Pa.
T (min) CFe (mg.L1) Removal (%)

CD

MD

CD

MD

0 15 30

28.5 1.3 0.8

30 0.9 0.8

0 95.4 97.1

0 97.0 97.3

Table 6 Removal of Fe(OH)3 by column otation, operating as DAF unit (30% recycle ratio) and with mid-sized bubbles (14 L.min1 recycle ow rate for middle size bubbles generation). Other conditions the same as in Table 1.
CFe (mg.L1) Removal (%)

T (min)

S1*

S2*

S3*

S1

S2

S3

0 15 30 45

65.0 2.2 2.3 2.2

64.9 2.0 2.0 2.1

65.0 1.9 1.7 1.6

0 96.7 96.5 96.5

0 96.7 96.9 96.8

0 97.2 97.4 97.6

150

*S1 = column cell working as DAF, with micro-bubbles; S2 = column cell working with mid-sized bubbles only; S3 = column cell working as DAF mixed with mid-sized bubbles

Nevertheless, Figure 5 shows better performance with working pressure of 3 atm than at 2 and 2.5 because of the higher air-to-solids ratio. Within this approach, Bauer et al. (1998) showed the advantages of the counter current dissolved air flotation filtration process (COCO-DAFF), where the recycled water is introduced, in a DAF cell, down stream from the flocculated water inlet. Thus, an even bubble blanket field in the separation zone (tank) is formed, through which all the aggregates to be floated must pass. Here, the recycle inlet is moved away from the conventional position, eliminating the floc damage by the recycle. Column cell. Table 6 summarizes the results obtained in the column flotation cell, working as a DAF cell, with micro-bubbles, and/or with mid-sized bubbles. High separation values were obtained in all cases despite some turbulence and high Fe3+ concentration. The feed entered at the top smoothly and the capture of the particles occurred efficiently by the rising bubbles. No rupture of the colloidal precipitates (or aggregates) was observed and a low split was ensured by the concentrate flow-rate. It is believed that flotation separations in such devices, using mid-sized bubbles with high air volume and micro-bubbles with high surface area, have great potential as a clean technology to treat water and wastewater (Rubio, 1998; Da Rosa et al., 1999; Rubio and Tessele, 1997; Parekh and Miller, 1999).
Conclusions

L.A. Fris et al.

The dissolved air flotation modifications presented in this study appear to have a great potential in industrial wastewater treatment. The optimization of the bubble generation stage by lowering the air/water interfacial tension with the use of a surfactant (collector/frother) in the saturator enables DAF operation (with a high efficiency) at saturation pressures of about 2 atmospheres. Since the saturation stage accounts for about 50% of the total operating energy costs and considering the low cost of the surfactant, this modification, in particular, appears to be a good process alternative. The mushroom type of diffuser improved DAF performance by decreasing the settling of bubble-particles aggregates. A column cell enhanced the bubble column depth, the air-to-solids ratio and process throughput when using mid-sized bubbles.
Acknowledgements

The authors thank all colleagues responsible for the friendly atmosphere and to all institutions (FINEP, CNPq, FAPERGS, among others) which support research in Brazil.
References
Da Rosa, J., Dias de Souza, M.L., Rodrigues R.T. and Rubio J. (1999). Waste waters treatment by non-conventional flotation. Global Symposium on Recycling, Waste Treatment and Clean Technology REWAS99, TMS-Inasmet. L. Gaballah, J. Hager and. R. Solozabal (eds.). San Sebastian-Spain. Proceedings, pp. 21232132. Dupre, V., Ponasse, M., Aurelle, Y. and Secq, A. (1998). Bubble formation by water release in nozzles-II. Influence on various parameters on bubble size. Wat. Res., 32(8), 24982506. Fris L.A. and. Rubio, J. (1999). Dissolved air flotation (DAF) performance at low saturation pressures. Filtration and Separation, 36(9), 6165. Ives, K.J. and Bernhardt, H.J. (eds.) (1995). Flotation processes in water treatment and sludge treatment. Wat. Sci. Tech., 31(34). Matis, K.A. (ed.) (1995). Flotation Science and Engineering. Rubio, J. (1998). Environmental applications of the flotation process. Effluent Treatment in the Mining Industry, University of Concepcin-Chile. S.H.Castro, F. Vergara and M. Sanchez (eds.). chapter 9, pp. 335364. Parekh, B.K. and Miller, J.D. (eds.). (1999). Advances in Flotation Technology.

151

Rubio, J. and Tessele, F. (1997). Removal of heavy metal ions by adsorptive particulate flotation. Minerals Engineering, 10(7), 671679. Yoon, R.H. and Luttrell, G.H. (1994). Microcel column flotation scale-up and plant practice. 26th Annual Meeting Can. Min. Proc., CIM, paper 12. Yoon, R.H., Luttrell, G.H., Adel, G.T. and Mankosa, M.J. (1992). The application of MicrocelTM column flotation to fine coal cleaning. Coal Preparation, 10, 177188.
L.A. Fris et al.
152