Environ. Sci. Technol. 2005, 39, 2307-2314
Breakage, Regrowth, and Fractal Nature of Natural Organic Matter Flocs PETER JARVIS, BRUCE JEFFERSON, AND SIMON A. PARSONS* School of Water Sciences, Building 39, Cranfield University, Cranfield, Bedford MK40 0AL, United Kingdom
The growth, breakage, regrowth, and fractal nature of flocs was investigated by use of a laser diffraction particle sizing device. A range of coagulants were investigated for the coagulation of natural organic matter (NOM) and compared to other coagulated systems. The results showed NOM floc structural characteristics varied in steady-state size depending upon which coagulant was used. When compared to other systems, the order of floc size was Fe precipitate > Fe-NOM > latex (in NaCl solution). Floc regrowth after exposure to high shear was limited for all of the flocs under investigation other than for latex in an inert electrolyte. This highlighted differences in the internal bonding structure of flocs, with the results suggesting that physical bonds have a capacity to re-form after breakage. Fractal dimension analysis by small-angle laser light scattering (SALLS) had limited applicability to large flocs that dominated all of the systems under investigation, but the degree of compaction increased as flocs were broken in high shear. This provided a possible mechanistic reason for the irreversible breakage seen.
Introduction Coagulation and flocculation remains the most common process used for the removal of turbidity particles and natural organic matter (NOM) at water treatment works (WTW). While the major removal mechanisms of these pollutants during coagulation and flocculation have been well studied, little thought is generally given to the fundamental floc operational parameters. These include physical properties such as floc size, compaction, and strength. One property that may have a significant impact at a WTW is the potential for flocs to regrow after being broken. Unit processes at WTW are generally designed to minimize floc breakage; however, often in practice this is not the case, with regions of high shear being prevalent (1). This may include areas around the impeller zone of flocculating tanks or transfer over weirs and ledges. A capacity to regrow may improve subsequent floc removal process efficiency if broken flocs are subsequently allowed to re-form during further particle-particle contact. The relative breakage and regrowth of different flocculated systems have previously been compared by use of floc strength and recovery factors (2-4) which may be calculated as follows: * Corresponding author phone: +44 1234 754841; fax: +44 1234 751671; e-mail:
[email protected]. 10.1021/es048854x CCC: $30.25 Published on Web 02/16/2005
2005 American Chemical Society
d(2) × 100 d(1)
(1)
d(3) - d(2) × 100 d(1) - d(2)
(2)
strength factor ) recovery factor )
where d(1) is the average floc size of the plateau before breakage, d(2) is the floc size after the floc breakage period, and d(3) is the floc size after regrowth to the new plateau. An increased value of strength factor indicates flocs that are better able to withstand shear and thus should be considered stronger than a suspension with a lower factor. Likewise, an increase in the recovery factor shows flocs that have better regrowth after high shear. By use of model spherical particles, such as latex beads, it has been previously shown that aggregates formed at a low velocity gradient in salt solutions that are then broken into smaller aggregates on exposure to an increased shear field will re-form to their initial size if the original velocity gradient is subsequently reapplied (5). This behavior is known as reversible breakage and is also observed during cyclical breakage and regrowth of activated sludge flocs (6). However, in most instances where conventional metal coagulants and polymers are used for the aggregation of small particle suspensions (such as kaolin and latex beads), irreversible breakage is usually seen, such that the initial floc size is never subsequently achieved after breakage. Differing degrees of recoverability have been seen depending upon the coagulant under investigation. Monodisperse polystyrene beads coagulated with aluminum sulfate have been shown to have variable regrowth depending on the intensity of the breakage shear (5). For the coagulation of kaolin particles, alum and polyaluminum chloride have been shown to have the poorest regrowth, reaching only a third of their previous size after shear; poly(diallyldimethylammonium) chloride (polyDADMAC) showed near complete regrowth; and a copolymer of acrylamide and cationic monomer showed total regrowth (4). From these studies there is a suggestion that the particular coagulant used and therefore the coagulation mechanism involved has a considerable impact on floc regrowth potential. There has been no previous work showing the regrowth potential of NOM flocs with different coagulants so an understanding of these properties may provide an important addition to the well-studied field of NOM coagulation and removal. Since Mandelbrot introduced the concept of fractal theory in the 1970s, the application of fractal geometry is now a well-established means of describing the complicated structure of floc aggregates (7-11). Flocs are examples of mass fractal objects as both the internal and surface structure of the aggregate exhibit fractal properties as opposed to surface fractals where the fractal relationship is only evident on the outside of the particle. Mass fractals may be summarized by the relationship between their mass M, a characteristic measure of size L, and the mass fractal dimension Df:
M ∝ LDf
(3)
For Euclidean objects, the value of Df will be 1 for a line, 2 for a two-dimensional planar shape, and 3 for a compact three-dimensional shape (12). Fractal objects take noninteger values of Df and are therefore said to show non-Euclidean dimensionality. Recent work has successfully used smallangle laser light scattering (SALLS) for the determination of fractal dimension for the aggregation of model primary particles such as latex (8) and aluminum oxide (9). Application VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2307
of SALLS to more complex natural systems such as activated sludge have been more difficult to interpret due to the fact that these flocs are thought to have two or more characteristic fractal dimensions as a result of multilevel structure (10). Furthermore, in the past relatively little work has investigated the fractal properties of NOM flocs, and the application of SALLS to large NOM flocs has not yet been fully explored. Understanding the changes in aggregate compaction during floc growth, breakage, and regrowth may help to explain floc structural changes seen during these periods. The overall objective of this paper was to compare the structural properties of NOM flocs coagulated with ferric sulfate, alum, or high molecular weight cationic polymer by use of an on-line laser diffraction particle sizing instrument. NOM flocs were then compared to model floc systems composed of monodisperse latex and coagulant precipitate. Floc aggregates were characterized in terms of their size, fractal dimension, breakage, and regrowth potential. This work also investigated the effect of shear exposure time on the recoverability of flocs.
Materials and Methods (A) Suspension Preparation and Coagulation Optimization. The NOM-rich water source was a reservoir in the north of the U.K. The water was of high color and dissolved organic carbon (DOC) (10 mg L-1 as C), low turbidity (0.5 NTU), and low alkalinity (
