Page 510 - Fundamentals of Water Treatment Unit Processes : Physical, Chemical, and Biological

Adsorption 465

treatment plants for uses such as industrial cooling water, Equation 15.3, for a given HLR, if EBCT changes by some
irrigation of public parks, and other uses as may be dictated factor, L(packed bed) must change by the same factor. The
by water rights or market factors. relationship is illustrated in Example 15.2.
The issue of trace organics (found in the finished water of
New Orleans in 1974) was the stimulus for the expanded use
of GAC in drinking water. The motivation was to control Example 15.2 Relation between EBCT
and L(packed bed)
carcinogenic substances, not just taste and odor compounds.
To document the extent of carcinogenic substances, the USEPA
Given
conducted a survey of drinking water, i.e., the National
Data from the Pomona, California pilot plant (Section
Organics Reconnaissance Survey reported in 1975 (Symons
15.3.2.1), which had four GAC reactors in series, are
et al., 1975). The surveys led to the 1977 Interim Primary 2
EBCT ¼ 37 min; HLR ¼ 18.6 m=h (7.6 gpm=ft ).
Drinking Water Regulations giving impetus to the use of acti-
Required
vated carbon in controlling organics and a 1978 proposed
Calculate L(packed bed)
regulation published in the Federal Register that required acti-
vated carbon treatment under certain circumstances. The pro- Solution
posed regulation was highly controversial and was withdrawn Apply Equation 15.3, i.e.,
(Weber, 1984, p. 906). 37 min ¼ L(packed bed)=(18.6 m=h h=60 min),
15.1.5.1 Lore L(packed bed) ¼ 11.60 m
Concerning the design of GAC reactors, two criteria emerged Discussion
from the AWTR program, e.g., 4.9 HLR 24 m=h(2–10 A packed bed is not homogeneous and so the idea of
2
gpm=ft ) and the use of ‘‘empty-bed contact time (EBCT). ‘‘contact time’’ has no bearing on the performance of the
reactor. The value of L(packed bed) must, however, be
Both criteria have become established as a basis for sizing
longer than the ‘‘reaction zone’’; also the longer the
reactors. By definition,
L(packed bed), the longer the run time, i.e., before ‘‘break-
through’’ and column exhaustion. Thus L(packed bed) is
Q
(15:2) related conceptually to reactor performance (see also
HLR
A Section 15.2.3.1).
where
HLR is the hydraulic loading rate, also called ‘‘superficial 15.1.5.2 Science
velocity’’ (m=s) The ‘‘science’’ of adsorption as a water treatment process
3
Q is the flow into packed bed (m =s) evolved from the AWTR program. From the research spon-
2
A is the cross-sectional area of packed bed (m ) sored, several key ideas developed, which are given below.
By definition, EBCT ¼ V(packed bed)=Q. Dividing the 1. Isotherms were established as a methodology to
numerator and denominator on the right side by A(packed express capacity of activated carbon
bed) gives 2. Kinetics became a means to express the rate of
uptake of the adsorbate by the GAC
L(packed-bed)
(15:3) 3. Modeling by pilot plants became a means to take out
EBCT ¼
HLR the uncertainty in process design
4. Mathematical modeling, based on the materials bal-
where
ance principle and kinetics, became a means to better
EBCT is the detention time in portion of reactor packed
understand reactor design and operation
with media for empty condition (s)
L(packed bed) is the length of packed-bed portion of
reactor (m) 15.1.5.3 Practice
V(packed bed) is the volume of packed-bed portion of As a ‘‘technology’’ GAC treatment has emerged only since
3
reactor without the media, i.e., the ‘‘empty’’ bed (m ) about 1960, based on stimuli from the USPHS Advanced
Water Treatment Program (Anon., 1962). Examples of ter-
Recommended criteria for HLR and EBCT (Culp et al., tiary wastewater treatment have included plants at Pomona,
2
1978, p. 185) are 4.9 HLR 24 m=h(2–10 gpm=ft ), and California, c. 1966; South Lake Tahoe, c.1966; Colorado
10 EBCT 50 min, respectively. The EBCT has been Springs, c. 1968; Orange County Water District, California,
accepted in practice as a parameter for sizing a GAC reactor c. 1974; Denver, c. 1982. Water treatment examples have
and was considered a most important design parameter in included Nitro, West Virginia, c. 1963; Cincinnati, Ohio,
sizing a reactor (Hager and Flentj, 1965, p. 1445), with HLR c. 1992; Commerce City, Colorado, c. 1989. These and
having little effect (Culp and Culp, 1974, p. 236). As seen in other examples are reviewed in Section 15.4.3.