Page 380 - Fundamentals of Water Treatment Unit Processes : Physical, Chemical, and Biological

Rapid Filtration 335

12.2.3.1 Dual Media 12.2.3.3 Alternative Modes of Filtration
Although dual media of anthracite and sand is usually asso- Conventional filtration (coagulation-flocculation-settling-
ciated with the 1960s, Hansen (1936) mentions ‘‘renewed’’ filtration) has been the standard for practice from the time of
interest in the use of crushed anthracite as a filtering medium Fuller’s experiments—keeping in mind that the distinction
(see also McNamee et al., 1956, p. 805, who mention that between coagulation and flocculation was not well delineated
the use of anthracite dated back to 1914 in Cumberland, until about 1920 with Langelier’s design of paddle wheel
Maryland, and was in current use in 26 some states). He flocculators. The compelling rationale was that the filtration
mentions that for a 10 year period, crushed anthracite on top system could then handle higher seasonal increases in surface
of sand was used at the Marston plant in Denver. Its use was water turbidities.
favored partly because mudball and cracking problems were In 1968, Conley and Evers, working with the low turbidity
less. The widespread adoption of anthracite and sand as a (1-2 NTU) waters of the Columbia River, advocated the idea of
dual media was in the early 1960s when promulgated by coagulation-filtration (Conley and Evers, 1968). This process,
Walter Conley, who, at the same time, introduced mixed that is, coagulation-filtration, is called inline filtration, a term
media (anthracite, sand, and garnet) filter beds. By the early suggested by Cleasby (1984) in his American Society of Civil
1970s, dual media was common in practice. Mixed media Engineers (ASCE) Simon Freese Lecture at Boulder, Colorado.
was a proprietary product of Neptune Microfloct and
became widely used.
12.2.4 MODERN FILTRATION PRACTICE
12.2.3.2 Breaking the HLR Barrier
Modern practice is characterized by its focus on the process,
Fair (1963, p. 820) refers to the filtration velocity and bed that is, understanding what happens within the filter bed,
depth in the following statement: coupled with excursions from traditional guidelines of past
decades. This change has been ‘‘enabled’’ by theory with
Among the practices from which the profession was eventu- stimuli from federal mandates regarding turbidity and Giardia
ally liberated by clear-thinking operators and imaginative cysts. The ‘‘tool’’ has been the pilot plant, used widely in both
designers were a slavish adherence to 30-in. beds of non- design and operation (Logsdon, 1982).
uniform sand and constant rates of filtration of 2 gpm=sq ft
of bed surface.
12.2.4.1 The Federal Role
2
McNamee et al. (1956, p. 793) shed more light on 2.0 gpm=ft Filtration practice languished with small incremental improve-
rate noting that in the early years of rapid filtration, sizing of ments over the decades until 1974 when the Safe Drinking
filter beds was based on the filtration rate of 1.4 mm=sor Water Act (SDWA) was passed, the first direct foray into federal
2 regulation of drinking water. The turbidity standard adopted by
5.0 m=h (2.0 gpm=ft ), which was selected because high
quality water was associated with this rate. They noted, how- the states to this time was mostly the 5 Jackson Candle turbidity
ever, that most plants operated at peak hourly rates at perhaps units (JTU) based upon the 1962 Drinking Water Standards for
2
up to 3.5 mm=s or 12.5 m=h (5.0 gpm=ft ). Baylis was, most Interstate Carriers (USPHS, 1962, p. 6). The 1974 SDWA,
2
probably, the person who ‘‘broke’’ the 5.0 m=h (2.0 gpm=ft ) however, mandated a 1 NTU standard for drinking water,
filtration rate barrier. He described (Baylis, 1956) the operat- which provided the impetus for the industry to reassess its
ing experience from 1948 to 1955 in which he compared practices. Another impetus were the regulations, called the
performance of 10 of 80 filters with filtration rates varying ‘‘Filtration Rule,’’ published in the June 29, 1989 Federal Regis-
2
2
from 5.0 m=h (2.0 gpm=ft ) to 12.5 m=h (5.0 gpm=ft ). His ter (FR54:124:27486) which promulgated a 0.5 NTU standard
conclusion was that with the higher filtration rates there was (effective June 1993). The goal of many in the industry is a
no deterioration of filtrate quality and that the productivity per 0.1 NTU standard and some plants have set that as in internal
filter was higher with the higher rates. In commenting on the standard. In addition, the ‘‘Filtration Rule’’ required an overall
work of Baylis, Hudson (1956, p. 1146) noted that filtration 3 log reduction in Giardia cysts, administered by giving so
2
rates of 10.0 m=h (4.0 gpm=ft ) and even 25.0 m=h (10.0 many ‘‘credits’’ for filtration and so many for disinfection that
2
gpm=ft ) are possible without deterioration of water quality complied with criteria established by the ‘‘Rule.’’ For example,
provided proper grain size and depth of filter medium are with conventional filtration, a 2 log credit was given, and for
selected. This concept of filter bed design was perhaps 25 disinfection by chlorination a 1 log credit was given. The Filtra-
years ahead of notions that prevailed in practice. tion Rule was a guideline, but did not ensure complete safety of
In the 1960s, this higher rate was advocated for mixed product water. For example, during the April 1993 Milwaukee
media proprietary filters as an average rate and by 1980 outbreak of 403,000 cases of cryptosporidiosis (estimated) the
2
the 3.5 mm=s or 12.5 m=h (5.0 gpm=ft ) filtration rate plant was in compliance with standards.
was accepted widely in practice. State regulations moved
toward this higher rate by about the mid-1980s. Practice 12.2.4.2 Modern Practice
developed toward even higher rates, for example, 9.1 mm=s Two influences changed the character of practice during the
2
or 32.5 m=h (13.3 gpm=ft ) at the Los Angeles Aqueduct Plant 1980s: (1) theory became assimilated into practice and
(Kawamura, 1996). (2) pilot plants became the basis for design and an aid to