Geoscience Reference
In-Depth Information
samples; total organic carbon (TOC); dissolved organic carbon (DOC); nitrate (NO
3
-
) and
phosphate (PO
4
3-
) ions; UV absorption at 254 nm, 340 nm, and 450 nm; dissolved oxygen
levels (DO); and ammonia (NH
3
).
Correlations of Pearson's coefficient (
r
or
r
2
) values of between 0.77 and 0.98 have
been reported between BOD and peaks T and A
T
(Reynolds and Ahmad,
1997
; Ahmad and
Reynolds,
1999
; Baker,
2001
; Ahmad et al.,
2002
; Hudson et al.,
2008
; Hur et al.,
2008
).
The correlations of peaks A
C
and C (which represent humic and fulvic-like fluorescence)
with the 5-day BOD were lower (
r
2
= 0.72-0.77) (Baker,
2001
; Hudson et al.,
2008
). The
strong correlation between the 5-day BOD and the tryptophan-like fluorescence at around
340-350 nm is expected, as BOD is an indirect measure of microbial growth. Furthermore,
the tryptophan-like fluorescence has been associated with wastewater microbial activity
(Reynolds,
2002
; Elliott et al.,
2006
; Hudson et al.,
2008
), whereas fulvic and humic-like
fluorescence has been reported to represent the nonreadily biodegradable organic fraction
of wastewaters (Reynolds,
2002
).
Reported correlations of Pearson's coefficient for peak T and the COD and DOC values
of wastewaters typically range from values between 0.42 and 0.97 (Reynolds,
2002
; Vasel
and Praet,
2002
; Lee and Ahn,
2004
; Wu et al.,
2006
). This is unsurprising given the well
established chemical relationships between TOC and DOC. The wide range of correlations
that have been observed for fluorescence (especially peak T) and DOC, COD, and TOC are
explained by the ratio of refractory DOM to labile DOM, of fluorescent and nonfluorescent
character. Correlations can exhibit greater variation depending on the ratio of both humic/
fulvic-like/tryptophan-like material and luorescent
/
nonluorescent DOM. For this reason,
some investigators have found it useful to examine the T/C ratio in relation to wastewater
and effluent samples. Baker (
2001
) found that the T/C ratios for rivers (1.0) were far lower
than that of untreated sewage (2.7-31). A comprehensive data set representing the T/C
ratios of waste, rivers, and drinking and deionized water is shown in the recent review by
Henderson et al. (
2009
). Limited data are available regarding the relationship with peak
C, although a correlation of 0.87 was reported by Bari and Farooq in
1985
. Overall, for
both BOD and COD measurements, much stronger correlations have been found when
sewage-derived fluorescence dominates the system under investigation. For example, Wu
et al. (
2006
) noted that the strength of correlation between COD and peak T in river water
increased when sewage-derived DOM was more dominant (correlation coefficients rising
from 0.5 to 0.9). Strong correlations have also been found between peak T and PO
4
3-
, NO
3
-
(Baker and Inverarity,
2004
), total Kjeldahl nitrogen (N
k
), and NH
3
(Vasel and Praet,
2002
).
These relationships are generally considered to be indirect because wastewater-derived pol-
lution is typically characterized by high phosphate and nitrate ions when advanced nutri-
ent removal is not undertaken at the wastewater treatment plant. Relationships between
ammonia and peak T are not always observed owing to the high removal rate of ammonia
during sewage treatment. However, ammonia can be present in significant quantities during
pollution events, as observed by Baker et al. (
2003
).
The monitoring of DOM levels through a treatment works allows process optimization,
and it has been estimated that as much as 40% of energy costs could be saved through
