Page 347 - Fundamentals of Water Treatment Unit Processes : Physical, Chemical, and Biological
P. 347

302                            Fundamentals of Water Treatment Unit Processes: Physical, Chemical, and Biological



                                                                                                    p
              z is the packing factor, indicating how the monomers are  For the Euclidian model, D F ¼ 3 and z ¼ p=(3 2) ¼ 0:7405;
                packed (dimensionless)                         the latter applies for close-cluster packing. The value of
              D F is the parameter that characterizes the fractal dimension  z depends also on the shapes of the monomers, i.e., if other
                of aggregates with respect to its three-dimensional  than spherical, which is likely. The usual assumption for the
                geometry; see Box 11.1 (dimensionless)         traditional model, however, as depicted in Figure 11.6a, is that
                                                               z ¼ 1, which means that there is no pore space due to packing
                                                               effects (which, of course, cannot be true). The diameter, d,of
                                                               the fractal aggregate is a pseudo dimension, since a fractal, by
                                                               definition, is difficult to characterize. Dimensions that could
                  BOX 11.1   FRACTAL DIMENSION, D F
                                                               serve include (1) the ‘‘hydraulic’’ diameter (based upon the fall
              The fractal dimension, D F , seen as the exponent in  velocity with diameter calculated by Stoke’s law) and (2) the
              Equation 11.9 has become a common parameter      diameter calculated from the radius of gyration (Lee et al.,
              among those who wish to pursue the idea of floc geom-  2000, p. 1990; Chakraborti et al., 2000, p. 3969). For fractals,
              etry in terms of fractal theory, as developed largely since  i.e., D F < 3, the lower values represent large, highly branched,
              the about the mid-1980s. Gregory (1989, p. 215)  and loosely bound structures (Chakraborti et al., 2000,
              explained  this  parameter  as  follows:  A  solid  p. 3969). For reference, values given for coagulation of min-
              three-dimensional body has a mass, which depends on  erals with 4.5 mg=L alum and 1 mg=L polymer (Purifloc A-23,
              the third power of some characteristic length (such as the  Dow Chemical) were D F (illite)   1.49, D F (montmorillonite)
              diameter of a sphere), so that a log–log plot of mass    1.79, D F (calcite)   1.65, and D F (silt)   1.37.
              against size should give a straight line with a slope of  Table 11.3 gives D F values with descriptions of suspensions
              three. When such plots are made for aggregates, how-  for lake water and a montmorillonite suspension after alum
              ever, lower slopes are found, with non-integer values.  coagulation by charge neutralization, and ‘‘sweep floc,’’ respect-
              The slope of the line is known as the fractal dimension,  ively (Chakraborti et al., 2000, p. 3969). The initial suspension
              D F . In three-dimensional space, D F may take values  was without coagulant. The charge neutralization stage was
              between 1 and 3, the lower value representing a linear  defined by the coagulant dosage required to give a floc zeta
              aggregate and the upper one an aggregate of uniform  potential for a minimum settled water turbidity before restabili-
              density or porosity. Generally intermediate values are  zation (and higher turbidity). For lake water, this was for zeta
              found, and the lower the fractal dimension, the more  potential   1 mV and for the montmorillonite suspension, zeta
              ‘‘open’’ or ‘‘stringy’’ the aggregate structure. The earli-  potential   15 mV (with alum dosages 3 and 2 mg=L, respec-
              est attempts at computer simulation of aggregation were  tively). The sweep-floc stage was defined as the minimum alum
              based on the random addition of single particles to  dose that resulted in a settled water turbidity   1ntu(14mg=L
              growing clusters, which gives D F ¼ 2.75, indicating a  for lake water and 20 mg=L for the montmorillonite suspension).
              fairly compact aggregate structure. An alternative model  As seen in Table 11.3, the fractal dimension parameter decreases
              is for cluster–cluster aggregation which is more like real  with increasing fractal size, indicating a looser, more spread-out
              flocculation which leads to a much lower value of D F ,  structure as corroborated by in situ photographs. Also, as
              i.e., D F ¼ 1.75, indicating a fairly ‘‘open’’ structure.  described, the larger aggregates are more irregular in structure,
                                                               with the primary particles for the sweep floc being surrounded



                       TABLE 11.3
                       Descriptions of Floc at Three Stages of Coagulation
                                       Coagulation
                       Suspension        Stage         D F                Description of Suspension

                       Lake water    Initial        2.93   0.20  Heterodisperse
                                      suspension
                                     Charge         2.57   0.20  Small flocs, irregular in shape
                                      neutralization
                                     Sweep floc      2.12   0.50  Large aggregates of many primary particles surrounded
                                                                 by gel-like alum floc
                       Montmorillonite  Initial     2.71   0.20  Heterodisperse
                                      suspension
                                     Charge         2.51   0.20
                                      neutralization
                                     Sweep floc      2.39   0.30
                       Source: Adapted from Chakraborti, R.K. et al., Environ. Sci. Technol., 34(18), 3969, 2000.
   342   343   344   345   346   347   348   349   350   351   352