PCBs AND ASSOCIATED AROMATICS 889
FIGURE 31
FIGURE 32
FIGURE 33
FIGURE 34
FIGURE 35
(13)
FIGURE 36
represent the same species as in reaction (3) or reactions (9),
(10), (11) or (13).
The unimolecular dissociation of dioxin takes place
when the molecule is heated sufficiently.
This reaction represents the disruption of the ring system
with an hydroxyl radical to yield non-dioxin products.
A major reaction pathway is expected to be the bimo-
lecular decomposition of phenoxy radicals with molecular
oxygen when oxygen is present in excess, as is usually the
case in incineration, as opposed to pyrolysis in a reductive
atmosphere.
The production of fuel radicals by hydrogen abstraction
from fuel is likely to be slower than the reaction between
hydroxyl radicals and the phenoxy radicals of reaction (8)
but this reaction will dominate especially in areas where
there are fuel-rich pockets in the incineration process.
The decomposition of fuel into products describes a
wide range of reactions.
Schaub and Tsang derived estimates of the rate constants
for each of the steps in the reaction sequence by comparing
literature values for similar chemistries and calculated four
situations governed by different mole fractions. The first
case, for example, assumed the condition that no O 2 or fuel
is available in the region in which the reaction mechanism
is taking place. This was modeled so that the formation of
dioxin could be exaggerated.
In the second case, concentrations were calculated on
the basis of there being oxygen available for reaction but at
an initial mole fraction which corresponded to about 60%
of the value needed for stoichiometric combustion. Under
normal conditions incinerators are run with excess oxygen
or at least a stoichiometric amount. This condition deem-
phasizes the rate of loss of chlorophenoxy radical in reaction
11 and consequently biases the calculation in favor of dioxin
formation. The third scenario models a fuel-rich pocket in
a post-combustion zone environment where no oxygen is
available for the reaction sequence.
Case four is similar to case three but includes the avail-
ability of oxygen (Figures 37, 38, 39).
The correlation between combustion temperature and
the concentration of dioxin in the example scenarios is illus-
trated in the above figures. The three curves were not plotted
together because the axes extend over significantly differ-
ent ranges. While the absolute concentrations predicted by
the model are relatively unimportant, it is significant that,
under post-combustion mixing, in an intermediate tempera-
ture zone, unburned chlorophenol is predicted to react with
hydroxyl radicals to produce PCDDs. Clearly, high tempera-
tures, and a sufficient quantity of air and fuel are able to
provide conditions which will combust chlorophenols into
products which are not PCDDs. Conditions can occur during
operation of the incinerator in which the amount of PCDD
produced is extremely small but, most importantly, some
conditions will allow the formation of PCDDs at concen-
trations which are 5 to 10-orders of magnitude higher than
those produced under ‘designed’ operating conditions.
The rate of destruction of organic material is deter-
mined from the kinetic differential equation which relates
concentration at a given moment of time to the rate constant
for the reaction:
d
d
C
t
−⋅kC
(1)
where
C concentration at time, t
k rate constant.
Upon integration,
t
k
C
C
(^10)
⋅
⎛
⎝⎜
⎞
⎠⎟
ln
(2)
where ,
C 0 initial concentration at time t 0.
C016_003_r03.indd 889C016_003_r03.indd 889 11/18/2005 1:12:35 PM11/18/2005 1:12:35 PM