United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-067 March 1983
Project Summary
Effect of Trace Metals and Sulfite
Oxidation on Adipic Acid
Degradation in FGD Systems
J, B. Jarvis
The adipic acid degradation rate was
measured in a bench-scale flue gas
desulfurization (FGO) system de-
signed to simulate many of the impor-
tant aspects of full-scale FGDsystems.
Results show that the adipic acid degra-
dation rate depends on the sulfite oxi-
dation rate, the adipic acid concentra-
tion, the presence of manganese in
solution, and temperature. The degra-
dation rate is also affected by pH, but
only when manganese is present Adipic
acid degradation products identified in
the liquid phase include valeric, butyric,
propionic, succinic, and glutaric acids.
When manganese was present the
predominant degradation products were
succinic and glutaric acids. Analysis of
solids from the bench scale tests shows
large concentrations of copreciprtated
adipic acid in solids with high sulfite
concentrations. By contrast low quan-
tities of copreciprtated adipic acid were
found in solids with low sulfite and
high gypsum concentrations.
This Project Summary was develop-
ed by EPA's Industrial Environmental
Research Laboratory, Research Tri-
angle Park. NC, to announce key find-
ings of the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
The addition of adipic acid to flue gas
desulfurization (FGD) wet scrubbers has
been shown to benefit both S02 removal
and limestone utilization. Adipic acid has
the effect of buffering scrubber solutions,
thereby improving liquid-phase mass trans-
fer. The use of adipic acid was first
proposed by G.T. Rochelle* and has been
tested by EPA in pilot systems at its
Industrial Environmental Research Labo-
ratory (Research Triangle Park), at TVA's
Shawnee test facility, Paducah, KY, and at
Springfield City Utilities' Southwest Power
Plant Springfield, MO. Results from these
previous test programs show that adipic
acid is effective as a scrubber additive at
concentrations of 700-2000 ppm. How-
ever, overall mass balances on these pilot
systems revealed unaccounted-for losses
of adipic acid. These losses were presumed
to be the result of chemical degradation of
adipic acid.
Further testing at Shawnee indicated
that adipic acid degradation could be
quenched by operating below pH 5.1.
Since these unexpected but favorable re-
sults could be significant to the future
application of adipic acid as a scrubber
additive, independent verification was
desired. Radian was contracted by EPA to
conduct a systematic study of the effect of
scrubber operating conditions on adipic
acid degradation. This investigation in-
volved two distinct areas of experimen-
tation: (1) a study of the effects of
scrubber operating variables on the adipic
acid degradation rate (the experiments
were performed using a bench-scale FGD
unit designed to simulate many of the
important characteristics of full-scale FGD
systems); and (2) in conjunction with the
bench-scale testing, analytical procedures
were developed to identify and quantify
adipic acid degradation products (these
"Rochelle, G T., Process Synthesis and Innovation in
Flue Gas Desulfurization, EPRI FP-463-SR, July
1977.
-------
techniques were used to analyze many
samples produced during bench-scale
testing). Photomicrographs of the solids
produced during bench-scale operation
were analyzed to determine the effect of
adipic acid on particle size and morphology.
A bench-scale FGD unit was constructed
to determine the effects of operating vari-
ables on the adipic acid degradation rate
The following variables were investigated:
SC>2absorption/oxidation rate, adipic acid
concentration, pH, trace metals (manganese,
iron), limestone and limestone plus fly ash
(versus reagent grade CaCOa), boiler
(versus synthetic) flue gas, temperature,
thiosulfate (as an additive), and sulfite
concentration.
A total of 43 bench-scale runs were
performed, 21 of which were baseline
tests in which the variables were limited to
the S02 absorption/oxidation rate, the
adipic acid concentration, and pH. CaCOa
was used as the alkaline species for all
baseline tests. The ranges for the variables
in the baseline tests were:
Variable
S02 absorption rate,
g SOz/hr
Percent solids oxidation
(a measure of the total
absorbed S02 that is
oxidized to sulfate)
Adipic acid concentration,
ppm
PH
Range
2.07 - 7.86
9.7- 100
876-10080
4.6 - 5.5
The resulting adipic acid degradation
rates were 15.4-600 mg/hr. A computer
program was used for a statistical analysis
of the baseline test data.
Results and Conclusions
The results showed that the adipic acid
degradation rate was a function of the
overall S02 oxidation rate and the adipic
acid concentration. The degradation rate
was found not to depend on pH, at least
over the range tested, when manganese
was not present The sulfite ion concen-
tration was also included in this analysis;
however, it was generally a function of pH
and was also found not to influence the
adipic acid degradation rate. A weighted
least squares analysis was used to corre-
late the significant variables. For data at
50°C, the resulting correlation, represent-
ing the best fit of the experimental data, is:
adipic acid
degradation rate,
mg/g S02 removed = 0.00308
i adipicacid v 05542
(concentration, ppmj
Figure 1 plots the above correlation.
I percent oxidation I
A literature search indicated that certain
trace elements, notably manganese and
iron, catalyzed sulfite oxidation. Since the
degradation rate of adipic acid was shown
to depend on sulfite oxidation, several
bench-scale runs were performed to eval-
uate the effects of manganese and iron on
adipic degradation. The effects of lime-
stone (vs. reagent grade CaCOa) and fly
ash were also evaluated, since these mate-
rials are sources of trace metals.
The presence of manganese, at concen-
trations of 2-21 ppm, significantly re-
duced the adipic acid degradation rate.
Reductions of 42.5 - 85.4 percent below
the adipic acid degradation rates predicted
from the baseline correlation were ob-
served for forced oxidation runs. Iron was
found to be essentially insoluble under
wet scrubber operating conditions: no
effect on adipic acid degradation was
observed.
The results of the forced oxidation runs
in which manganese was present are
shown in Figure 2, which plots adipic acid
degradation rate constant Kd versus pH.
Rate constant Kd, used here, is simply the
adipic acid degradation rate normalized by
the total sulfite oxidation rate and the
average adipic acid concentration. Kd is a
convenient method of comparing the re-
sults of runs at different adipic acid concen-
trations and levels of oxidation. Values of
Kd for the baseline runs without manga-
nese were 1.24 - 1.62 for the range of
variables shown in Figure 2. The re-
duction in degradation due to manganese
is evident in Figure 2: the values observed
for Kd are 0.2 - 0.93.
Figure 2 also shows a dramatic effect of
pH on the adipic acid degradation rate in
the presence of manganese. In the baseline
runs, in which no manganese was present
no pH effect was observed At a given pH,
the degradation rate also appears to be a
function of the manganese concentration.
The data supporting this result however,
are somewhat limited.
The concentration of manganese which
could be held in solution varied with both
pH and the level of oxidation. Generally
speaking, more dissolved manganese was
observed when operating under forced
oxidation conditions than under natural
oxidation conditions. At a given level of
oxidation, the concentration of dissolved
manganese increased with decreasing
pH. A survey of the available literature
indicates that manganese can remain in
solution only in the +2 oxidation state.
Further, Mn+2 is probably not oxidized to
the 44 or +7 oxidation states in scrubber
solutions. Since less manganese could be
held in solution during natural oxidation
tests, the precipitate is most likely MnSOa.
This possibility, however, was not verified
by analysis.
One natural oxidation test was per-
formed in which significant concentrations
of manganese (c*13 ppm) remained in
solution. The resulting adipic acid degra-
dation rate was 59 percent below that pre-
dicted by the baseline correlation. This
indicates that manganese is also effective
in reducing the adipic acid degradation
rate under natural oxidation conditions if
sufficient quantities of manganese can be
held in solution.
Several forced oxidation runs were per-
formed in which Springfield limestone,
rather than CaCOa, was used as the source
of alkalinity. The adipic acid degradation
rates in the limestone tests were signifi-
cantly below those in similar baseline tests
with CaCOa. This limestone was analyzed
for trace metals, and a significant quantity
of manganese was found. Some of this
manganese dissolved in the scrubber liquor
during the bench-scale tests. The resulting
adipic acid degradation rates were about
the same as those in runs with CaCOa in
which manganese was added to the hold
tank. In natural oxidation tests with lime-
stone, however, much less of the man-
ganese in the limestone went into solution.
In these tests the degradation rates were
not significantly different from that pre-
dicted from the baseline correlation.
Combined limestone and fly ash was
also tested under forced oxidation condi-
tions. The resulting adipic acid degra-
dation rate did not differ significantly from
that of limestone alone. In natural oxidation
bench-scale tests at Shawnee and Spring-
field, in which boiler flue gas with fly ash
was utilized, no significant effect of fly ash
was observed A separate run, evaluating
the effect of fly ash alone, was not per-
formed.
Manganese ions definitely decrease the
adipic acid degradation rate. Manganese
probably affects adipic acid degradation
through a change in the mechanism of
sulfite oxidation, which is thought to pro-
ceed through a free radical mechanism.
Manganese may catalyze the sulfite oxida-
tion reaction to reduce the concentration
of the free radical believed to be involved in
adipic acid degradation. Alternatively,
manganese may catalyze the sulfite oxida-
tion reaction utilizing a mechanism which
does not involve a free radical. The possible
existence of at least two competing sulfite
oxidation mechanisms, both of which could
be functions of pH, may help in interpreting
the dependence of the adipic acid degra-
-------
£
8"
I
I
I
160
ISO
140
130
120
110
100
90
80
70
60
SO
40
30
20
10
©
I
I
I
I
J
5000 rOOOO 75000
(Liquid-Phase Adipic Acid Concentration)
200OO
0.5542
25000
30000
35000
40000
(% Solids Oxidation)
1.185
Figure 1. Correlated relationship between the adipic acid degradation rate, the liquid-phase adipic acid concentration, and the percent solids
oxidation tor the baseline data set.
1.0
7 0.9
^ 0.8
!•'
S 0.6
*
| 0.5
i
0.3
0.2
0.1
0.0
R-34; 9 ppm Mn*
R-29 (limestone); 5 ppm Mn**;
-------
tration was a strong function of pH. To
obtain an independent determination of
the effect of the sulf ite ion concentration, it
was increased from 470 to 2600 ppm by
adding sodium ions. In this way, the effect
of different sulfite ion concentrations
could be evaluated at the same pH. A
comparison of the degradation rates from
these tests with the values predicted from
the baseline correlation indicates that the
sulfite ion concentration has little, if any
impact on the adipic acid degradation rate.
The effect of thiosulfate, a sulfite oxida-
tion inhibitor, was also evaluated. Even
under forced oxidation conditions, the
solids oxidation was only 3.4 percent The
resulting degradation rate was extremely
low, again emphasizing the important role
of sulfite oxidation on adipic acid degrada-
tion.
In addition to degradation, adipic acid
can be lost from FGD systems by co-
precipitation in the scrubber solids. The
amount of coprecipitated adipic acid in
solids generated in the bench-scale tests
is shown in Figure 3 which plots the ratio
of the solid- and liquid-phase adipic acid
concentrations versus percent oxidation.
Note that the solids adipic acid concentra-
tion used in Figure 3 refers only to co-
precipitated or occluded adipic acid; i.e.,
adipic acid in the liquid associated with the
wet filter cake is not included These
results show that the adipic acid concen-
tration in the solids is high at low oxidation
levels and decreases to nearly zero at high
oxidation levels.
Photomicrographs of the solids from
bench-scale testing showed that adipic
acid had an effect on both particle size and
morphology under natural oxidation con-
ditions. In general, increasing liquid-phase
concentrations of adipic acid caused both
a decrease in the calcium sulfite particle
size and an increase in aggregate structure.
Adipic acid can be lost from FGD systems
by three mechanisms: chemical degrada-
tion, coprecipitation in the scrubber solids,
and liquid losses with the filter cake solids.
The relative magnitude of these losses is
shown in Figure 4 in which estimates for
each of the three adipic acid losses are
plotted versus percent sulfite oxidation.
The bases for these estimates are:
• Chemical adipic acid degradation rate
constant Kd was assumed to be 0.5
M~1, typical of the degradation ex-
pected at pH 5.1 for systems in
which manganese is present at con-
centrations of about 20 ppm.
• The liquid-phase adipic acid concen-
tration was assumed to be 2000
ppm.
s,
3.50
3.25
3.00
2.75
2.50
2.25
2.00
7.75
7.50
7.25
7.00
0.75
0.50
0.25
0.00
30 data points
in square \ J
10
20 30
70
80
90 100
40 50 60
Percent Oxidation
Figure 3. Relationship of adipic acid content of bench-scale solids to percent oxidation.
18.0
16.0
14.0
§ '2.0
CO
10.0
1 "
? 6.0
I „
2.0
0.0
Figure 4.
Total loss of adipic acid',
Loss of adipic acid due
to chemical degradation
Loss of adipic acid due
to coprecipitation or
occlusion in the solids
Loss of adipic acid in liquid
entrained in the filter cake
solids
10 20 30 40 50 60 70 '80
Percent Oxidation of Sulfite to Sulf ate
90
1OO
Estimated total adipic acid losses due to chemical degradation, coprecipitation or
occlusion in the solids, and liquid entrained in the fitter cake versus percent
oxidation.
-------
• The weight percent solids in the fitter
cake, from which the adipic acid loss
in liquid associated with the filter
cake is calculated, was assumed to
be that determined during full-scale
adipic acid testing at Springfield. Ten
percent of the total solids were as-
sumed to be inerts, either from the
fly ash or the limestone.
• The loss of adipic acid copreciprtated
or occluded in the waste solids was
estimated from the bench-scale re-
lationship shown in Figure 3.
• Closed-loop operation was assumed;
i.e., blow-down or miscellaneous
slurry losses were not considered.
Adipic acid losses from actual full-scale
systems may differ somewhat from that
shown in Figure 4 for at least three reasons:
1) Manganese concentrations higher
than those evaluated in this program
may be seen in full-scale systems.
2) Manganese is more soluble under
forced oxidation conditions.
3) Operation under forced oxidation
conditions may result in higher con-
centrations of liquid-phase degrada-
tion products
These three factors tend to reduce the
forced oxidation adipic acid losses relative
to natural oxidation losses.
Implications for Full-Scale
Application
The bench-scale results suggest that
the following variables have the greatest
impact on adipic acid degradation:
• Sulfite oxidation rate (percent solids
oxidation).
• Adipic acid concentration.
• pH (only if manganese is present).
• The manganese concentration in
solution.
• Temperature (hold tank temperature
in forced oxidation systems and
scrubber temperature in natural
oxidation systems).
At a given type of operation (i.e., natural
or forced oxidation) the minimum total
loss of adipic acid must be determined by
optimizing pH and the adipic acid concen-
tration. Naturally, the optimization pro-
cedure must be performed within the
constraints of the desired SQz removal
efficiency. A result of the optimization
procedure would be the equilibrium con-
centration of manganese. The use of
additional manganese, however, is proba-
bly not a realistic option in a full-scale FGD
system.
An additional means of reducing adipic
acid losses is the use of two-tank forced
oxidation, in which sulfite oxidation and
limestone addition take place in separate
tanks: scrubber effluent is routed to the
oxidation tank where sulfite oxidation oc-
curs; and oxidation tank slurry is sent to
the hold tank where limestone is added
and gypsum precipitates. The advantage
of this system over conventional single-
tank forced oxidation is that sulfite oxida-
tion, and the resulting adipic acid degrada-
tion, take place at pH values below that of
the hold tank. The resulting decrease in
the total adipic acid consumption rate can
be illustrated by comparing the estimated
adipic acid consumption of a two-tank
forced oxidation system with that of single-
tank natural and forced oxidation systems.
Data for the total loss rates of adipic acid
from single-tank forced and natural oxida-
tion were given in Figure 4.^ The total
adipic acid loss for the two-tank system
can be estimated by assuming that the
oxidation tank operates at 0.5 pH units
below that of the hold tank This would
decrease the effective adipic acid degrada-
tion rate constant from roughly 0.5 to 0.2
M-1 at 20 ppm manganese (see Figure 2).
Estimates comparing the total adipic acid
consumption rate of a two-tank system
with both natural and forced oxidation
single-tank systems are given in Table 1.
Table 1 shows that the total adipic acid
consumption rate can be minimized using
the two-tank forced oxidation desiga
Two-tank forced oxidation may offer
several other advantages:
• Low pH in the oxidizer tank can
promote sulfite oxidation, while high
pH in the hold tank promotes S02
removal.
• The extra residence time provided by
two tanks allows for more complete
precipitation of calcium sulfate as
well as improved limestone dissolu-
tion.
• The limestone blinding problem,
which is common in adipic acid en-
hanced forced oxidation systems,
can be minimized by oxidizing sulfite
prior to limestone addition.
• Gypsum solids from forced oxidation
systems are easier to dispose of than
solids from natural oxidation systems.
A disadvantage of the two-tank forced
oxidation design is the necessity of two
tanks. Economic factors will center on a
trade-off between higher capital costs for
the two tanks and lower operating costs
due to a decrease in adipic acid consump-
tion. The two-tank forced oxidation design
has been tested at TVA's Shawnee test
facility: no serious operating problems
were encountered.
Table 1. Comparison of Total Adipic Acid Consumption fora Two-Tank forced Oxidation Design with Conventional Single-Tank Natural and Forced
Oxidation Systems
Natural Oxidation System* Forced Oxidation1'
(Single-Tank) (Single-Tank)
Loss of adipic acid through 6.25 15.6
chemical degradation, mg/g
SO? removed
Loss of adipic acid through 0.78 0.27
coprecipitation or occlusion in
the scrubber solid, mg/g SQz
removed
Forced Oxidation1'
(Two-Tank)
6.25
0.27
Loss of adipic acid in liquid
associated with the filtered
solids, mg/g SO2 removed
Total adipic acid consumption
mg/g SOz removed
4.02
0.75
0.75
11.05
16.62
7.27
"40% solids oxidation.
b99+ % solids oxidation.
-------
J. B. Jarvis is wjth Radian Corporation, Austin, TX 78766.
J. David Mobley is the EPA Project Officer (see below).
The complete report, entitled Effect of Trace Metals and Sulfite Oxidation on
Adipic Acid Degradation in FGD Systems," (Order No. PB 83-148 379; Cost:
$ 17.50, subject to change} will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
PS
------- |