892 PCBs AND ASSOCIATED AROMATICS
The furnace volume, V A · x e∴⋅ ⋅⎛
⎝⎜⎞
⎠⎟t
QV
TTT
T
1
()ln
em m(10)that is , the residence time, t, is the time spent by the gas
above a temperature, T.
In partial plug flow kinetics, there is some longitudinal
mixing which allows some of the gas to exit the hot zone
along a ‘fast path’. If the fast path is assumed to be half of
the mean residence time, t f , then,tV
QT TT
f em Tm
2 ⋅⋅⋅⎛
⎝⎜⎞
()⎠⎟ln.
(11)A plot of Eqs. (10) and (11) are shown in Figure 40.
The mean residence time of a material has been calcu-
lated on the assumption that it is passing through a 113,000 L
incinerator with a gas velocity of 400,000 L/min, a maximum
temperature of 1600°C and an exit temperature of 650°C.
A 99.9999% combustion efficiency is a mandated
requirement for an incinerator burning PCBs. The PCB
contaminated material introduced into the combustion
zone must, according to 40 CFR 761.40, be maintained for
2- seconds at 1200°C ( 100°C), that is, 2200°F ( 200°F),
with 3% excess oxygen in the stack gas, or, 1.5 sec resi-
dence time at 1600°C ( 100°C), that is, 2900°F ( 200°F)
with 2% excess oxygen in the stack gas.
The figure indicated that a 99.9999% destruction and
removal efficiency could be obtained for tetrachlorobenzene in
the example incinerator when the mean residence time is 0.7 sec
or less. In the case of pentachlorobiphenyl, the residence time
would have to be less than about 1.1 sec. If either the tetrachlo-
robenzene or the PCB passed through the hot zone on a fast-
path, the model predicts that the required destruction efficiency
would not be achieved since the curves are predominantly above
the fast-path temperature-time curve. Note that a residence time
of 2.0 sec at 2200°F (1200°C) predicts that a 6–9’s DRE might
be achieved for both substances, but that these conditions are a
lower limit for the incinerator to achieve this level of destruc-
tion. A temperature of 2900°F (1600°C) for 1.5 sec would seem
to be a much surer set of conditions to achieve 6–9’s destruction
because then both of the curves for the compounds of interest
would lie substantially below the mean residence time and fast-
path residence time curves in the model.Toxic Equivalency Factors and Human Exposure Many of
the PCB congeners yield several different PCDF isomers upon
pyrolysis. Consequently, in addition to the discussion above on
the toxicity of individual PCB isomers present in Aroclor 1016,
the following consideration of PCDF isomer toxicity is made
of evaluate the risk of generation of PCDFs in a worst case
scenario. The analysis of PCDFs in individuals exposed to the
Yu-Cheng poisoning showed that the most persistent congeners
were the 2,3,7, 8-tetra- 1,2,3,7,8-penta-, 2,3,4,7,8-penta-, and1,2,3,4,7,8-hexachlorodibenzofuran. Mixtures of these PCDFs
with 6 of the most toxic PCBs found in the contaminated oil
were prepared as a reconstituted mixture and tested to deter-
mine the relative potencies of PCBs and PCDFs (Safe et al.^65 ).
It was shown that PCDFs are from 680 to 2,210 times more
toxic than the PCBs. It was therefore suggested that PCDFs
were the major etiologic agents in the Yusho and Yu-Cheng
poisonings (Table 23).
The relative potencies of PCB and PCDF congeners
were measured in terms of their enzyme induction. For
example, ED 50 values are defined as the concentration of
chemical which results in 50% of the maximal enzyme
induction observed in the dose-response procedure used for
each PCDF. The smaller the value for the induction potency
or the receptor binding affinity of an isomer, the more toxic
the compound. Table 23 lists the relative potencies of some
of the most toxic PCDFs. There are considerable differences
in the observed activities of PCDFs which differ with respect
to Cl substitution at the C-2 and C-3 or C-7 and C-8 posi-
tions. In all cases the C-3 or C-7 substituted isomers are the
most active. PCDF isomers which differ only in their Cl sub-
stitution at C-1 (or C-9) and C-4 (or C-6) positions show
that the enzyme induction and receptor binding activities of
the C-1 isomers were all lower than the corresponding C-4
substituted isomers.
The toxicity of PCDFs in animals has been reviewed by
several authors and while there is very little isomer specific
toxicity data for congeners other than 2,3,7,8-TCDF, the
generalization has been made that there is a close similarity
between the biological activity of PCDFs and that of the
structurally similar PCDDS. This is illustrated in Table 18,
which shows the relative potency of some PCDF isomers
relative to TCDD.
Nagayama et al.^67 noted that strains of mice can be sepa-
rated into two groups after treatment with 3-methylcholanthrene
(3-MC) according to whether or not they were AHH respon-
sive. Similarly, the toxicity of PCDFs and PCDDs to mice
has been shown to correlate with AHH response (Nagayama
et al.^68 ). The authors therefore attempted to determine the pres-
ence of a similar susceptibility in humans by correlating the
AHH activity in the lymphocytes of Yusho patients and their
skin symptoms (Nagayama et al.^69 ).
Interpretation of the data given above leads to the con-
clusion that when PCBs are pyrolyzed, only the formation of
PCDF isomers which contain 4 or more Cl substituents need
be of concern. In particular, tetra- and higher chlorinated
congeners with lateral substitution are the most significant.
It should be noted that PCDDs are not generated at all by the
pyrolysis of PCBs (Table 24).
The four mechanisms of ring closure illustrated in Figure
20 can be used to predict the isomer distribution of PCDF
products although not their relative amounts. Table 25 indi-
cates the PCDF isomers with more than three chlorines which
can be obtained from Aroclor 1016. The potentially most
toxic mixture of compounds which could be derived from
Aroclor 1016 in a worst case scenario would, from the above
prediction of the isomers formed, be of very low toxicity. The
PCDF isomers which could theoretically be produced wouldC016_003_r03.indd 892C016_003_r03.indd 892 11/18/2005 1:12:36 PM11/18/2005 1:12:36 PM