EnvironmentalEngineering[ ! Vet.4, No. 3 / September2000 pp. 161 ~164
Experimental Verification of Collision Efficiency Factor in DAF By M o o y o u n g Han*, Wontae Kim*, Yiseon Hart* and Seok Dockko**
Abstract The collision efficiency factor of bubble and particle (~,;,) in dissolved air flotation (DAI') can be calculated theoretically by trajectory analysis, which takes into account both hydrod~yaaamicsand inter-particle forces. To determine the theoretically optimum particle size lbr any given bubble size, a collision efficiency diagram for DAF was developed where collision efficiency is contoured on a plane of particle and bubble sizes lbr diffi:rcnt conditions of particle zeta potential. A set ofexpe~ments tested the validity of the suggested collision efficiency diagram, and examined whether pretreatment is important and why slight coagulant overdosing and shorter flocculation times are generally prefen~d in DAY, both current, accepted practices. Batch DAt" reactors were used and kaolin samples were prepared from jar tests using different alum dosages and floceulation times. The particle size distribution, particle zeta potential, and turbidity removal in each experiment were measured, as were bubble size and zeta potential. The results agreed well with the predictions of the collision efficiency diagram and explained current practice. A collision efficiency diagram identifies the prctreatment goal, i.e., tailoring of the optimum characteristics r~Nuired of particles (zeta potential and size) under existing operational bubble characteristic.
Ken'words: efflciemT diagram, DAE flocculation time, particle size distribution, turbidity removal, zeta potential.
Introduction Although dissolved air flotation (DAd,') is gaining acceplance in the field of water and wastewatcr treatment, progress in theoretical de.'scription has been slow and the mechanisms of the process are poorly understood. In this rt~arch, a collision efficiency diagram is proposed on the basis of theoretical considerations, and the validity of such a diagram is tested by experunents. The collision efficiency of bubble and particle (%p) can be calculated by trajectoD" analysis, which takes into account hydrodynamics and inter-particle forces. This analysis was derived using a similar method developed for differential sedimentation by tlan and Lawler (1991). Modeling results and sensitivity analyses for each gweming parameter in DAF have been previously reported (Han et af., 1997; IIan, 2000). The most imporlant parameters that affect collision efficiency were found to be the zeta potential and sizes of both parLicle and bubble. In this stud3,; a collision efficiency diagram is generated that shrews the distribution of collision efficiency in ten~s of the characteristics of both bubbles and particles. A set of experiments was pertbrmed to verify the colti-
sion efficiency diagram and some accepted, but theoretically unproven, common practices treed in DAE These practices are bascxl on the belief that pretreatmcnt is imrx)rtant, and that slight coagulant overdosing and shorter tlocculation times give the best DAF performance. Batch DAF reactor runs mad standard jar tests were used employing kaolin suspensions of different alum dosages and flocculation times. Particle size distribution, particle zeta potential, and degree of turbidity removal were measured for each experiment. Bubble siT~ and bubble zeta potential were also measured. The purpose of this paper is to verify and explain both the experimental results and current DAb" practices from a theoretical basis. 'I1ae ramifications of, and applications for. a collision efficiency diagwam are discths~d.
Collision Efficiency Diagram A collision efficiency" (c%) diagram was developed on a plane of particle and bubble size where contours of ~x~ value~ are drawn fbr three particle zeta potentials (+10, 0, -10 mV), as shown in Figure 1. In developing this diagram, the zeta potential of bubbles was assumed to be 25 mV from
* Seoul National Universi~; Seouk 151-742~Korea(e-mail:
[email protected]) ** Wc~suk University,Woanjoo-Ktm,Chtmbuk, Korea The manuscript for this paper was submitted for review on August 11, 2000. Vot. 4, No. 3/Seplember 2000
- 161
Mooyoung I[an, Wonlae Kim, Yiseon Han and Seok Dock!~
under this condition. From this diagram, it is possiblc m identi~ the goal of pretreatment ff~rDAF in terms &the size and zeta potential of particles ibr a given bubble gcneration system (bubble size and zeta potential).
previous measurements by the authors (Hart and Dockko, 1999), although theory suggests the magnitude of the bubble zeta potential is not a sensitive parameter in this range. The density of particle and bubble is taken as 2.0 and 0.00017 m~/cm ~, respectively. The ionic strength of liquid is taken as 0.000l M. At positive pmicle zeta potential that may result from coagulant overdose in the coagulation mechanism of charge neutralization (Figure l a), the overall colliskm efficiency becomes higher. The relationship between the sizes of particle and bubble that result in the highest collision efficiency is suggested. The most collisions occur when the particle size is the stone or slightly larger than the bubble size. Vv'hen particle zeta potential trends towards zero (Figure ib), collision efficient; becomes lower than in the case of positive zeta potential, but the relationship between collision efficiency and bubble and particle sizes is unchanged. IIowevcr, at negative particle zeta potential (Figure lc), which represents zero or a low coagulant dosage, the collision efficiency was ve~' low for any combination of particle and bubble size. A poor removal efficiency is expected
a) Zp. lOmV
~!:~:..(~--~ "," ,"/.~:/,7"
~
~
-~::::~:i~
60.r~
N.oo
.:.:,'§ ~.~7".~::~:
