Environmental Engineering Reference
In-Depth Information
attempting to apply partitioning to urban sediments,
however, another problem is that the environment is
rarely stable. Pulses of sediment are constantly being
released by phases of construction (Hollis 1988).
This sediment may be contaminated topsoil, or may
simply supply silt and clay particles, which provide
binding sites for contaminants already present in the
environment. These may remain on impermeable
urban surfaces until the fi rst storm transports them
by storm sewers to receiving streams. Morrison
et al.
(1988, 1995) studied processes in gullypots, coining
the term “biochemical reactors” and found that spe-
ciation is constant until the onset of a storm event,
when acidic rainwater fl ushing through the gully pot
leads to increased solubility of trace elements and
their consequent removal in solution, adding signifi -
cantly to the fi rst fl ush effect. Adding the possibility
of fi rst fl ush to the mobilization of sediments by
street cleansing activities and gully pot emptying
(Ellis & Revitt 1982; Morrison
et al
. 1995), a great
deal of contaminated material can eventually reach
receiving waters by storm sewers (Harrison
et al.
1985; Anderberg & Stigliani 1994; Charlesworth &
Foster 1999).
Sand
Clay
Organic matter (coating)
Fe and Mn oxides (coating)
Mineral
Fig. 5.6
Diagram of typical sediment (mineral + organic
matter + oxides).
Adapted from Federal Interagency Stream - Restoration
Working Group (1998).
crystalline structure of minerogenic materials, as
shown in Fig. 5.6, making sediments all the more
important in the treatment process.
There have been many studies of such speciation
of trace elements in urban aquatic sediments, which
have sought to identify the trace elements that were
potentially the most likely to be released into solu-
tion, should environmental conditions change. Many
such studies have found (e.g. Gibbs 1973) that PAPs
are transported in discrete phases, refl ected by Tessier
et al.
(1979) in their sequential extraction protocol
(see also Charlesworth
et al.
, and Bortoluzzi
et al.
,
both this volume).
Sutherland (2000) and Charlesworth & Foster's
(1993) studies in urban lakes have shown that the
dominant binding site can vary both with the indi-
vidual element and with time. The former found Pb
to be potentially in the most available form, whereas
the latter found Cu and Zn were mainly associated
with exchangeable sites. Other studies identifi ed
other elements that were of concern; for example,
Revitt & Morrison (1987) found that 59% of Cd in
association with particulates in stormwater could be
considered bioavailable. These data show the range
of inorganic contaminants that have the potential,
should environmental conditions change, to cause
problems upon release (Bird
et al.
200
3).
However, many of these phases are operationally
defi ned and thus it is diffi cult to apply them back to
the environment on the one hand, but also it is dif-
fi cult to compare different studies of sequentially
extracted heavy metals because different analytical
methods tend to be used (Sutherland 2000). When
5.5.1 Site specifi city
Table 5.1 shows the wide range of concentrations
of trace elements found in association with urban
aquatic sediment. This refl ects differences in the
layout of the individual city: traffi cked areas, pedes-
trianization, distribution of green space, treatment of
watercourses, etc. As was discussed by Charlesworth
& Lees (1999), and further improved by Wong
et al
.
(2006), processes unique to the urban environment
impact signifi cantly on the transport, deposition, and
storage of urban sediment. These processes include,
but are not exclusive to the frequency of street cleans-
ing, climate which determines the intensity and fre-
quency of rainfall, the type of industry present, and
the structure of the hard-engineered sewer system. As
a result, Charlesworth & Lees (1999) have shown
that the frequency distribution of urban sediment
samples changes as they pass from source to deposit.
The source groups have a highly positively skewed
distribution, the transported group less so, and the
distribution curve of the deposited group becomes
near normal (Fig. 5.7). This refl ects, fi rstly, the great
variability in sources of these elements within urban
