United States
                     Environmental Protection
                     Agency
 Industrial Environmental Research
 Laboratory
 Research Triangle Park NC 27711
                     Research and Development
EPA-600/S7-83-035 Oct. 1983
4>EPA          Project  Summary

                     Full-Scale  Utility  FGD  System
                    Adipic  Acid  Demonstration
                     Program

                     0. W. Hargrove, Jr., J. D. Colley, R, L Glover, and M. L Owen
                      This report culminates a series of
                    projects sponsored by the Environmen-
                    tal Protection Agency (EPA), investigat-
                    ing the use of adipic acid as an additive
                    to enhance  sulfur dioxide (SO^) re-
                    moval in aqueous flue gas desulfuriza-
                    tion  (FGD) systems using limestone
                    reagent
                      A 9-month program at the 194 MW
                    Southwest Power Plant (SWPP) of City
                    Utilities, Springfield, MO, demonstrated
                    the effectiveness of adipic acid and
                    dibasic acids, the latter by-products
                    obtained during production of the former.
                    The test program examined the effect of
                    adipic acid addition to a limestone FGD
                    system under natural and forced oxida-
                    tion modes of operation.
                      Major conclusions are: (1) adipic
                    acid addition is a feasible method for
                    improving Sfy removal in systems that
                    are limited by soluble alkalinity in the
                    scrubber slurry feed; (2) the correlation
                    from TCA prototype testing at Shawnee
                    by TVA adequately predicts full-scale
                    system performance for SOj removal
                    as a function of adipic acid concentra-
                    tion and pH;  (3) limestone utilization
                    can be improved with adipic acid addi-
                    tion because a lower scrubber feed can
                    be used to achieve a given SCfe removal;
                    (4) increased limestone utilization from
                    the addition of adipic acid can improve
                    system reliability by reducing the ac-
                    cumulation of precipitating solids, es-
                    pecially in forced oxidation systems;
                    and (5) mixed dibasic acids, an equiv-
                    alent alternative to adipic acid, were
                    the least expensive alternative examined
                    for the SWPP to meet their SO2 removal
                    requirement
  This Pro/at* Summary was developed
by EPA's Industrial Environmental Re-
search Laboratory. Research Triangle
Park. NC, to announce key findings of
the research project that is fully doc-
umented in a separate report of the
same title (see Project Report ordering
information at back).

Introduction
  The U.S. Environmental Protection Agency
(ERA) has focused a portion of its research
activities on developing organic-acid-en-
hanced aqueous flue gas desulfurization
(FGD) technology using limestone reagent
for existing and new facilities. In 1981,
EPA sponsored a full-scale demonstration
at City Utilities of Springfield's Southwest
Power Plant (SWPP); Radian Corporation
was the test contractor.
  Most aqueous FGD systems operated
by the electric utility industry employ lime
or limestone reagent Lime/limestone FGD
systems are also anticipated to be the
most widely used technologies in the
future because of lower capital and operat-
ing costs compared to other aqueous FGD
technologies. Of the two, limestone sys-
tems are generally  more  cost-effective
since limestone is less expensive  than
lime Although limestone systems are
widely used, their performance has been
limited by the  reactivity of the reagent
which produces a lower soluble alkalinity
in the scrubber slurry. This limitation re-
sults in requirements for relatively higher
liquid-to-gas (L/G) ratios and larger reac-
tion tanks than other FGD scrubbing proc-
esses.
  In response to these limitations, EPA's
Industrial Environmental Research Labora-

-------
tory at Research Triangle Park(IERL-RTP)
initiated a research and development pro-
gram to improve the performance of lime/
limestone systems. One of the most prom-
ising aspects of this effort has been the
use of organic acids as buffering agents to
increase soluble alkalinity in  limestone
FGD systems. Adipic acid was identified
as a suitable additive because of its physical
properties, availability, and cost EPA has
sponsored extensive laboratory-, pilot-, and
prototype-scale studies on adipic-acid-en-
hanced limestone systems. Initial testing
indicated that adding adipic acid to  lime-
stone  FGD systems can provide one or
more of the following process improve-
ments: enhanced SOj removal, increased
limestone utilization, lower capital and
operating costs, and greater system flexi-
bility.
  The first three improvements are  self-
explanatory; the fourth requires additional
explanation. Limestone systems with adipic
acid addition can be used successfully in a
wider range of applications than systems
using other additives, and will  result in a
more flexible operating system.
  For example, adipic acid enhances per-
formance on both natural and forced oxida-
tion systems, while inorganic buffers are
less effective in forced oxidation systems.
A forced oxidation limestone system has
been projected as a more attractive option
compared to a  natural oxidation system
because of improved sludge disposal
operations; however, soluble alkalinity can
limit the SO2 removal performance of
limestone systems, and inorganic additives
such as magnesium are ineffective in
forced oxidized operations because of their
low dissolved sulfite concentrations Alter-
natively, adipic acid addition is effective in
increasing soluble alkalinity in forced oxi-
dation systems.
  Another example  is the influence coal
chloride content has on the effectiveness
of additives. Relatively low levels of adipic
acid have been effective in high dissolved
chloride systems while greater amounts of
magnesium are required to counteract the
effects of chlorides. Therefore, adipic acid
could be quite attractive in a plant where
the coal chloride concentration  varies sig-
nificantly.
  These are only two examples of  how
adipic  acid addition can  be  beneficial
Other advantages in flexibility include: (1)
ability to respond to changes in  coal sulfur
content  (2) adaptability to changes in
limestone reactivity; and (3)  ability to
achieve greater than 95 percent removal
to aid in achieving 90 percent removal on a
rolling  30-day average basis.  Reducing
the limitations and increasing the flexibility
of limestone systems will promote the use
of limestone  rather than  lime which is
more reactive but also more expensive.
  The improvements resulting from adipic
acid addition can be realized in new sys-
tems as well as in marginally designed
existing systems experiencing perform-
ance limitations. Available data  indicate
that new  adipic-acid-enhanced  systems
can  be designed to meet New Source
Performance Standards cost-effectively or
to attain more stringent removal  require-
ments  to  comply with local regulations.
Existing systems which are not meeting
design specifications can be easily adapted
to accommodate adipic acid addition to
improve S02 removal and/or limestone
utilization. An excellent potential exists to
use adipic acid as a less expensive alterna-
tive than other remedial actions.
  More recently,  a large  source of by-
product mixed, organic dibasic acids (DBA)
has been identified as a potentially more
cost-effective alternative  to adipic acid.
The DBA selected for testing, a by-product
of adipic acid manufacturing, is a mixture
of glutaric, adipic,  and  succinic acids.
Other by-product organic-acid-containing
streams may also be attractive options.

Program Objectives
  Results of previous activities were suffi-
ciently encouraging to warrant the dem-
onstration of the use of organic additives
in a full-scale limestone FGD system. Ob-
jectives of this program were to  demon-
strate the  effectiveness of adipic acid to
improve the  performance of limestone
systems, and to demonstrate the validity
of the previously developed data base as a
design tool for full-scale  FGD systems.
DBA addition was also briefly tested for
comparison with the adipic acid data base.
In meeting these objectives, S02  removal
data were collected using monitors that
were certified and operated according to
EPA procedures.
  The test program was influenced by the
characteristics of the full-scale  system
where the demonstration was conducted.
City Utilities  (C.U.) of Springfield  was
willing to participate in the demonstration
program  because  it had been  actively
pursuing methods to improve S02 removal
in the limestone FGD system  at their
SWPP. The system had been  unable to
achieve design S02 removals and,  con-
sequently, was unable to meet permit
requirements of 520 ng/J (1.2 Ib S02/
106Btu). This emission limit typically re-
quires  82 -  85  percent  S02 removal;
therefore,  the test program focused pri-
marily on the use of adipic acid to improve
SO2 removal.
  Other factors also influenced  the test
program. The FGD system at SWPP is a
natural oxidation  system.  However, EPA
also elected to conduct forced oxidation
tests since natural and forced oxidation
designs are feasible options for new and
retrofit systems,  and  since  adipic acid
offers advantages in the forced oxidation
system that inorganic enhancement agents
(eg., magnesium and sodium)  do not
provide As a result of these considerations
the following areas for testing were speci-
fied at SWPP:
  •  forced oxidation tests
     - baseline
     - adipic acid addition for 95 percent
      removal
  •  natural oxidation tests
     - baseline
     - adipic acid addition
      - 85 percent S02  removal
      - 95 percent S02  removal
     - DBA addition
      - 85 percent S02  removal
      - 95 percent S02  removal
  The performance of the FGD system at
95  percent removal was investigated to
determine the potential of organic acid
addition to meet  stringent S02  removal
requirements.  Since the SWPP system is a
natural oxidation process,  85 percent re-
moval was investigated in natural oxidation
tests to assist C. U. in identifying the operat-
ing conditions  necessary to meet its permit
requirements.  DBA was also tested at 85
percent removal as a potentially less expen-
sive option for the SWPP.
  Tests at SWPP lasted 9 months: first 4
months was   spent investigating  adipic
acid addition in a forced oxidation system-
next 3-1/2 months was spent testing adipic
acid addition to a naturally oxidized system
to achieve 85-95 percent SO2 removal;
and, finally, 1 -Vi months was spent testing
DBA in a naturally oxidized system. Results
of these tests are summarized, following a
brief process description.

Process Description
  CU.'s Southwest Unit 1 is a 194 MW
unit designed to burn coal containing 3.5 -
4.0 percent sulfur. After exiting the boiler,
flue gas passes through an air preheater,
electrostatic precipitator, and induced draft
fans prior to entering the FGD system and
subsequently exiting through the stack
  The original FGD system consisted of
two  parallel turbulent contact absorber
(TCA) modules, each sized to handle 60
percent of the design flue gas flow. Three
contact levels, each with 100 mm (4 in.) of
TCA spheres were provided in each module
The L/G ratio at full load was 6 - 7 l/Nm3
(40  - 45  gal./1000  acf). The pressure

-------
drop across the absorber portion of each
module was 25 - 30 cm H20 (10 - 12 in.
H20). Prior to the demonstration program,
CU. had experienced reliability problems
and excessive maintenance expenses with
the TCA modules. During the course of the
forced oxidation tests,  the TCA spheres
were removed from module S-1 and some
testing was conducted without the spheres.
Then, the spheres were replaced for the
remainder of forced oxidation testing. The
module was converted  to a tray absorber
at the conclusion of the forced oxidation
tests. Module S-2 was  not converted to a
tray absorber until the end of the demon-
stration program so that data were collected
for both TCA and tray tower contactors
during natural oxidation testing.
   Figure  1 is a flow schematic for one
module.  Flue gas enters the presaturator
and is cooled by supernatant liquor(thickener
overflow). The gas then passes through
three levels where it is contacted by the
slurry and S02  is  removed.  Entrained
slurry is removed from the flue gas by the
mist eliminator and the cleaned flue gas
exits through the stack The lower bank of
mist eliminator blades are washed  con-
tinuously with supernatant liquor to  con-
trol slurry accumulation and scale formation.
   The dewatering system consists  of a
thickener, a vacuum filter, and a pug mill for
mixing ash with  FGD  sludga The ash/
sludge  mixture is landfilled onsite.  The
thickener overflow is  collected in the
supernatant tank and  returned to the
system. The limestone preparation facility
consists of a wet ball mill producing a rela-
tively coarse slurry, 50-60 percent passing
through 200 mesh. The ground limestone
is collected in a sump,  diluted, and then
pumped to the limestone storage tank.
During the demonstration, adipic acid was
added to the limestone sump. Since the
dewatering and limestone preparation equip-
ment was common to both  modules, the
adipic acid concentration in each module
could not be varied independently.
   Two relatively simple modifications were
made to the FGD system to accommodate
the needs of the demonstration program:
1) an adipic acid feed system was installed
and 2)  forced  oxidation capability  was
provided temporarily. UOPs Air Correction
Division  modified the  equipment  Both
systems were designed for temporary use
during the demonstration program and, as
a  result  suffered from the omission of
some design  features that would be in-
cluded in permanent installations. In par-
ticular, the forced oxidation modifications
were  limited to the  addition of air com-
pressors and  reaction tank spargers. No
changes  were made to the dewatering
                     Thickener
                     Overflow
                                                   Air Sparger
                                              (Forced Oxidation Only)
                                                              Sludge to
                                                               Landfill
                                      Vacuum
                                       Filter
Figure  1.    Southwest Unit 1 scrubber module flow diagram.
system, designed  for natural  oxidation.
The substantial change in sludge properties
created some operating problems during
forced oxidation tests. Also, the existing
reaction tank agitator and gear box were
undersized for operation  in the forced
oxidation mode so that oxidation efficiencies
were affected.  The  types  of  problems
observed, attributable to retrofitting a
natural oxidation system to test in a forced
oxidation mode, do not indicate problems
that would result from the  use of adipic
acid.

Forced Oxidation Test Results
  Forced oxidation tests were conducted
because it is a major design configuration
for limestone FGD systems. Furthermore,
forced oxidation systems produce waste
sludge  containing  less occluded water
than  naturally  oxidized  systems.  Test
data collected prior to the demonstration
program indicated that the small amount
of water in the sludge would minimize the
loss of adipic acid from the system.
  Forced oxidation tests were initially con-
ducted  without  adipic acid addition to
provide baseline performance data Then,
adipic acid was added, and the performance
of both modules was monitored. The strin-
gent S02 removal goal of 95 percent was
established to provide a severe test of the
ability of adipic acid to enhance removal
performanca Since CU. was interested in
eliminating the  TCA spheres from the
system because of repeated  failures.
Module S-1 was operated for a short time
with the TCA spheres removed, and the
surfaces of the TCA cages provided the
only gas/liquid contacting. The rest of the
forced oxidation tests were conducted
with the TCA spheres restored to  S-1.
Table 1  summarizes results of the  forced
oxidation testa
  The results indicate that in tests without
the TCA spheres,  adding  adipic acid to
about 1300  ppm in  the slurry with a
scrubber feed pH of 5.4 increased the S02
removal from less  than  50 percent at
baseline conditions to about 84 percent
These results represent the performance
of a system limited by mass transfer area
In a 1-day test  to further increase  SO2
removal without the TCA spheres, the
adipic acid level and pH were increased to
2000 ppm and 5.6, respectively. Over 90
percent S02  removal was achieved, but
the long-term operabilrty of the system
under these conditions was not demon-
strated.

-------
 Table 1.    Summary of Forced Oxidation Testing
                   Operating
                     Time
                    (days)
Scrubber
  Feed
  pH
                                                  Adipic
                                                  Acid
                                                  fppmj
                           Atoms O/
                           Mole S02
                           Removed
  SO2
Removal
Limestone
Utilization
 Sulfite
Oxidation
Testing Without TCA Spheres (Module S-1)
Baseline              20            5.4
Adipic Acid            6            5.4
Addition
Testing With TCA Spheres
Baseline-
  Module S-2          24
Adipic Acid
Addition -
  Module S-2"         49(13)
  Module S-1          40
                 0            3.1             46
              1340            2.8             84
5.4              0            3.0            58
  5.0(5.0)       2410(3323)      3.0(3.2)         91(96)
  5.1           2340            7.3*            89
                                                                                                89
                                                                                                89
                                                                                                88
                                                                              95
                                                                              90
                                                                              91
                                                                                                 92(92.5)        94(>99)
                                                                                                 O1             7At)
                                                                                                 91

'Results are average: results in parentheses represent performance after start-up and shakedown problems had been resolved.
b Maintenance problems prevented efficient oxidation for Module S-1 compressors.
  In tests with TCA spheres, an average
S02 removal of 91 percent was achieved
in Module S-2 at an adipic acid concentration
of 2410  ppm and pH of 5.0. Similarly, an
average SC>2 removal of 89 percent was
obtained in Module S-1 with2340ppmof
adipic acid and a pH of 5.1. This compares
with a baseline S02 removal of only 58
percent at a higher pH (5.4).  Note that
Module S-1  could not be fully oxidized
throughout the program due to continuing
compressor maintenance problems. There-
fore, the test data from S-1  are more
indicative of a natural  oxidation  system
experiencing high oxidation rates than a
properly designed forced oxidized system.
  Although the 95 percent average S02
removal goal was not achieved during the
entire forced oxidation test period, it  is
significant that 90 percent average S02
removal and 90 percent limestone utiliza-
tion were obtained over the 49 day test
period. These average results included
data collected during start-up and optimiza-
tion periods when process operating dif-
ficulties (e.g., limestone blinding with sul-
fite, and process water discharge from the
dewatering system) were encountered and
solved. These difficulties are primarily at-
tributable to retrofitting the FGD system to
accommodate forced oxidation  as dis-
cussed earlier. During  2 weeks in April,
near the end of the forced oxidation, the
average S02 removal in Module S-2 was
96  percent and 95 percent daily average
S02 removal was achieved on all but 2
days.

  The dramatic increases in S02 removal
with adipic acid  addition  in the  forced
oxidation tests were accomplished without
a decrease in limestone utilization. Baseline
test results with and without TCA spheres
reveal a very good limestone utilization of
88  -  89 percent Limestone  utilization
remained high when greater SO2 removal
       was  attained with adipic acid.  In other
       applications, where an improvement  in
       S02 removal is not the primary concern, it
       is  anticipated  that limestone utilization
       could be  improved by the addition of
       adipic acid while  maintaining constant
       S02  removal. The improvement in lime-
       stone utilization would be due to the ability
       to operate at a lower pH with adipic acid
       addition.
         The performance data  collected from
       Modules S-1   and S-2 during the tests
       with TCA spheres were compared with
       data from a previous test program collected
       in a smaller prototype system. The adipic
       acid prototype testing was sponsored by
       EPA and conducted  at TVA's Shawnee
       test facility between 1978 and 1980. A
       correlation relating S02 removal with adipic
       acid  concentration and pH was subse-
       quently developed by BechteM1' based on
       the Shawnee prototype data base:
              Fractional S02 Removal =
       1 - exp (-0.0047 expfpH + 0.00062 Ad)}
                                         (1)
       where pH and Ad represent the pH of the
       scrubber feed liquor and adipic acid con-
       centration in ppm, respectively.
         The full-scale data collected at SWPP
       were correlated using the form presented
       in Equation 1 for pH values of 4.8 - 5.5 and
       adipic acid concentrations of 0 - 4000
       ppm. The full-scale correlation is shown
       as:
              Fractional S02 Removal =
       1 - expj-0.00415 expfpH + 0.0065 Ad))
                                         (2)
       Figure 2  compares the predicted S02 re-
       moval using Equation 2 with observed re-
       sults. Although the data scatter represents
       both the operational problems discussed
       previously and the fluctuations of a full-scale
       system, 90 percent S02 removal was rou-
       tinely achieved with an adipic acid concen-
       tration of 2000 ppm and a scrubber feed
       pH of 5.0.
                                                                                    The  Shawnee (prototype)  correlation
                                                                                  predicts S02 removals which average 2-5
                                                                                  percent greater than the actual results ob-
                                                                                  served on the full-scale system at SWPP.
                                                                                  Some  differences  in the  Shawnee and
                                                                                  SWPP FGD systems which could contribute
                                                                                  to the difference in correlation include:  1)
                                                                                  a slightly lower TCA sphere charge at SWPP,
                                                                                  2) slightly lower slurry and gas rates on an
                                                                                  equivalent cross-sectional area basis  at
                                                                                  SWPP, and 3) coarser limestone at SWPP.
                                                                                  Significant maldistribution of TCA spheres
                                                                                  observed at the end of the forced oxidation
                                                                                  testing at SWPP would also contribute to
                                                                                  the difference The  conclusion that can be
                                                                                  drawn from  this comparison  is that the
                                                                                  Shawnee data base can be used as an effec-
                                                                                  tive design tool for full-scale systems. Good
                                                                                  engineering  practice would require con-
                                                                                  sideration of important variables affecting
                                                                                  performance that are not included in  Equa-
                                                                                  tion 1  or the application of some conserva-
                                                                                  tism when predicting system performance.

                                                                                  Natural Oxidation Test Results
                                                                                    After a 1-month  scheduled  outage fol-
                                                                                  lowing the forced oxidation testing, investi-
                                                                                  gation of FGD system performance  in the
                                                                                  natural oxidation mode was initiated. During
                                                                                  the outage. Module S-1 was converted to
                                                                                  a tray tower;  however,  the resulting tray
                                                                                  pressure drop initially was too small. The
                                                                                  tray void area was reduced to increase the
                                                                                  gas-side pressure drop to specified  levels
                                                                                  (i.e., 10 -12 in. across the absorber) after
                                                                                  about 6 weeks of natural oxidation testing.
                                                                                    Much of the testing in the natural oxida-
                                                                                  tion mode focused  on defining conditions
                                                                                  required to meet 85 percent S02 removal,
                                                                                  since CD. was interested in gaining experi-
                                                                                  ence in operating the FGD system under
                                                                                  conditions required to meet its S02  emis-
                                                                                  sion limitation. Additionally, some testing
                                                                                  was conducted to achieve 95 percent SO2
                                                                                  removal and to determine the operating
                                                                                  conditions required to meet more stringent

-------
802 emission limits. Table 2 summarizes
the natural oxidation results for both the
TCA and tray towers.
  After baseline testing was conducted,
adipic acid was added to achieve 85 per-
cent S02 removal. Initially, screening tests
were performed by varying the pH and
adipic acid  concentration independently
and observing the resulting S02 removal.
The results of these screening tests were
then  used to select preferred operating
conditions for longer term tests at constant
pH and adipic  acid concentrations. The
longer term steady-state tests were con-
ducted at pHs  of 5.5  (typically used at
SWPP)  5.2 (approached the pH  recom-
mended by EPA for reducing adipic acid
consumption).  Approximate  adipic acid
addition rates for these pHs were estimated
from the screening  testa
  Subsequent test results in the tray tower
produced  an average 90 percent S02 re-
moval at a pH of 5.2  and an adipic acid con-
centration of 760 ppm. An average S02 re-
moval of 91  percent was achieved at a pH
of 5.5 and an adipic acid concentration of
640 ppm. The data, collected by continu-
ous emission monitors, represent average
performance during fluctuating plant loads.
The S02 removal at full load was about 85
percent at each pH level tested. Test results
also show that 95 percent SO2 removal at
full load was achieved with 1 700  - 1800
ppm adipic acid at a pH of 5.4.
  Results for the TCA module show that
system performance deteriorated during
the natural oxidation tests. Note that only
78 percent  removal was achieved in the
TCA compared to 91  percent in the tray
tower during the third set of tests with an
85 percent removal goal. Plugging in the
TCA, at least partially responsible for the
decline in performance, illustrates the type
of problem encountered previously at the
SWPP. A more detailed discussion of these
problems will be presented later in this sec-
tion.
  Results for limestone utilization shown
in Table 2 indicate  that an improvement
was  observed  in the  adipic acid tests
compared to the baseline tests. The primary
reason for the improvement is that adipic
acid  addition allows the desired S02 re-
moval to be obtained at a lower pH which,
in turn, increases limestone utilization.
  Since one module had been converted to
a tray tower and the TCA module suffered
from operating difficulties which reduced
performance efficiency, it  is difficult to
compare the SWPP full-scale results di-
rectly with the  Shawnee TCA prototype
results for natural oxidation. Most of the
correlations reported for prototype natural
oxidation testing were derived for higher
o
V)
 Q)
 O
•5

I
                                                      Line of
                                                   O Perfect
                                                      Agreement
                                                  O =S-2
            35   40   45
50   55   60   65   70   75   SO   85  90   95  100

    Observed SO2 Removal, %
Figure 2.    Fit of Springfield data to revised Bechtel model: Fractional SOx removal = 1 -
            exp{-0.00415 expfpH + 0.00065 Ad)}.
Table 2.    Summary of Natural Oxidation Testing

                                    Operating
                                      Time
                                     (days)
                        Adipic      SO2    Limestone
                       Acid Cone.   Removal  Utilization
                        (ppm)      (%)       (%)
Tray Tower (S-1)
Baseline"
Goal - 85% Removal' (Screening}
Goal - 85% Removal (5.2 pH)
Goal - 85% Removal (5.5 pH)
Goal - 95% Removal
            24
            17
            12
            10
            17
5.6
5.4
5.2
5.5
5.4
   0
1000
 760
 640
1750
68
87
90
91
96
79
81
86
 . c
78
TCA Tower (S-2)
Baseline
Goal - 85% Removal (Screening)
Goal - 85% Removal (5.2 pHj
Goal - 85% Removal *> (5.5 pH)
Goal - 95% Removal11

24
17
12
10
13

5.6
5.3
5.2
5.4
5.2

0
890
700
550
1690

72
91
88
78
94

69
92
82
75
86
'The void area in the tray tower was about 10 percent higher during these tests than in the later tests.
bThese goals were obtained in the tray module but not in the TCA module. The goals for the TCA
 module were not attained due to deteriorating conditions in the TCA module and a common adipic
 acid addition system which prevented independent control of adipic acid levels.
€No data collected.

-------
pH levels than at SWPP. However, spot
checks of full-scale data obtained at higher
pH levels before the TCA module perform-
ance deteriorated indicate that the Shawnee
correlations predict full-scale performance
results reasonably well. As in the forced
oxidation test series, the prototype correla-
tions  for natural oxidation tests predict
slightly higher S02 removals than achieved
in the full-scale system.
  The data collected in the demonstration
program from the full-scale  tray tower
were fit to the following equation:
       Fractional S02 Removal =
         1  -exp(-O.OOl31Ad-
   0.00216exp(pH) - 0.00996L/G} (3)
The L/G ratio is expressed in gallons per
1000 cubic feet  of gas.  Note that the
effects of liquid and gas rates resulting
from load changes could be observed and
correlated  to S02  removal during the
natural oxidation testing. The wider range
of S02  removals  observed during the
natural oxidation tests compared to the
narrow  range resulting from  the forced
oxidation removal goal of 95 percent per-
mitted the effect of I/G to be observed.
  The agreement between actual S02  re-
moval and the removal predicted by Equa-
tion 3 is shown in  Figure 3. This equation
was developed to describe the trends
noted during the program.  Be careful when
using the equation for design purposes,
especially for extrapolation beyond the
conditions tested at the full-scale system.

Adipic Acid Consumption
  Adipic acid which is added to a limestone
FGD system can leave the system several
ways: 1) soluble adipic acid is lost in the
water occluded with the solids of the
scrubber sludge, 2) adipic acid is coprecipi-
tated  with  the solids in  the  sludge,  3)
adipic acid is degraded in the system (eg.,
oxidation), 4) adipic acid is contained  in
fugitive liquid or slurry  losses,  and  5)
adipic acid exits with the  flue gas. An
actual-totheoretical ratio used to describe
adipic acid consumption has been defined
in previous studies to be the actual amount
of adipic acid fed to the system divided by
the amount of adipic acid exiting in the
sludge liquor.
  Adipic acid consumption data for both
natural  and forced oxidation  operating
modes  is summarized in Table 3. The
actual-to-theoretical ratios of approximately
9  for forced oxidation and 5  for natural
oxidation are somewhat greater than those
observed in the Shawnee prototype test-
ing,<2' but are roughly equivalent to values
obtained in bench scale testing under
similar  conditions.'3)  In  spite of lower
   100-

   95-


   90-


   85-
I
I  75-
    70-


    65-^


    60-


    55


    50-1


    45
                                                     .*  ^.':
                                                  <•     s>* •    !
                                                  v»   >*•        •
                                                                •   •
                                                               Line of
                                                               Perfect
                                                               Agreement
      45    50    55    60     65    70    75    80

                               Observed S02 Removal, c,
                                                      85
90
                                                                   95   100
Figure 3.    Fit of Springfield tray tower data to model: Fractional SOz removal = 1 -
            expl-0.00131 Ad - 0.00216 expfpHJ -0.00996 L/G).
Table 3.   Adipic Acid Material Balance Results
Forced Oxidation "
(95% Removal)
Adipic Acid Addition flb/hrj
Adipic Acid Losses in Sludge Liquor
flb/hr)
Actual/Theoretical Ratio b
Adipic Acid Consumption (Ib/ton
SO2 removed)
Cost (C/kWh) c
94
10
9.4
18
0.029
Natural Oxidation
85% Removal
46
JO
4.6
11
0.017
95% Removal
113
21
5.4
29
0.047
"During forced oxidation numerous losses of adipic resulted from water discharges in dewatering
 plant operations. These numbers have been adjusted to reflect more normal operation.
bActual/theoretical ratio = actual adipic acid added/'adipic acid lost in sludge liquor.
cCost of adipic acid assuming 60C/lb.
losses of adipic acid in the sludge liquor,
higher chemical degradation in the forced
oxidation configuration was  responsible
for the  higher actual-to-theoretical ratio
observed.
  The adipic acid consumption in pounds
per ton  S02 removed is a useful ratio in
estimating consumption for proposed sys-
tems. During forced oxidation (95 percent
SO2 removal), 9 g/kg S02 (18  Ib/ton
                                        S02) was consumed. The adipic acid con-
                                        sumption for the combined TCA and tray
                                        tower modules nearly tripled during natural
                                        oxidation testing when the S02 removal
                                        was increased from 85 (6 g/kg S02[11 Ib
                                        /ton S02]) to 95 percent(15 g/kg SO2[29
                                        Ib/ton S02]).  The consumption rate in
                                        forced oxidation tests was less than during
                                        natural oxidation  testing at 95 percent
                                        S02 removal as  well.  This comparison

-------
 illustrates that the losses in the scrubber
 liquid were sufficiently reduced by improv-
 ing dewatering efficiency during forced
 oxidation to overcome the higher degrada-
 tion rates.

 Operating Characteristics
  In addition to the significant increase in
 SO2 removal capability and the ability to
 operate at reduced scrubber feed  pH to
 improve limestone utilization,  improve-
 ments in operating reliability were observed
 during the demonstration program. FGD
 system reliability prior to and during the
 demonstration program is shown in Table
 4. After the demonstration start-up and
 shakedown in January and February, 1981,
 both scrubber modules exceeded an average
 reliability of 85 percent from March through
 August The significant improvement in
 reliability, compared to the  2  previous
 years, can be attributed to several factors.
 First  a number of mechanical  improve-
 ments were implemented by CD. As  a
 result of the demonstration program, some
 process operating procedures were revised,
 and additional process engineering input
 was available. The  use of adipic acid also
 helped improve system reliability by modi-
 fying process chemistry so that less scale
 accumulated on equipment surfaces. While
 the influence of adipic acid  addition on
 reliability cannot be quantified, some of
 the operating characteristics  attributable
 to adipic acid are discussed below.
  During forced oxidation, almost 36 days
 of continuous operation were logged for
 both modules. At the end of  the testing,
 only a  small amount of solids had ac-
 cumulated in  the  TCA  cages and mist
 eliminators. During natural oxidation test-
 ing, reliable operation continued although
 a greater buildup of solids was observed,
 particularly in  the  TCA cages. Increased
 solid accumulation was also observed in
 the mist eliminators during natural oxida-
 tion, but the major cause of the accumula-
 tion was probably plugged mist eliminator
 wash nozzles.
  The  excellent performance observed
 during forced oxidation tests is due in part
 to the relatively high limestone utilization
 which  results  in less calcium carbonate
 dissolution in  the  absorber. In  addition,
 forced oxidation provided beneficial effects;
 e.g., more gypsum  seed crystals available
 as precipitation sites.
  The lower pH and forced oxidation re-
 sulted in higher limestone utilization than
was observed during the natural oxidation
tests.  The increase in calcium carbonate
and sulfite dissolution during natural oxida-
tion can increase the tendency to scale or
 plug. While an increase in solids accumula-
Table 4.    Scrubber Reliability "During Adipic Acid Demonstration Program

                                                Module S-1        Module S-2
Average for 1979

Average for 1980

1981 Demonstration Program
  Forced Oxidation Tests
    January 10-31
    February
    March
    April
    May
          34

          59
          66
          54
          88
          67
          87
42

31
85
35
81
97
66
Natural Oxidation Tests
June
July
August
Average January 10 - August 31
Average March 10 - August 31 b
78
94
90
79
86
74
94
80
86
87
"Defined ashoursoperatedx 100.
          hours required
b Excludes demonstration start-up and shakedown in January and February.
tion in  the  TCA  module during natural
oxidation was observed, the tray module
had only a small amount of solids accumu-
lated on the trays, and the solids could be
removed with a minimum of effort
  It can be  concluded that  adipic acid
addition, while not the sole contributor to
improved reliability at SWPP, can allow an
FGD system to operate at lower pH cor-
respondingly higher limestone utilization
which should reduce accumulation of solids
in the system and improve reliability. Re-
duced solids accumulation was particularly
evident in the forced oxidation testing at
SWPP.
  Although  overall  system performance
was excellent during most of the program,
the air compressor, dewatering equipment,
and adipic acid feed system caused operat-
ing problems which had to be overcome.
Most of these problems, directly attribut-
able to the temporary retrofit of the system
to accommodate the demonstration pro-
gram, do not reflect the general use of
adipic acid in FGD systems.
  A phenomenon that occurred during
natural oxidation tests should be considered
in systems requiring high S02 removals.
Limestone blinding (calcium sulfite precipi-
tation on limestone particles) apparently
occurred during natural oxidation  tests,
particularly during testing to achieve 95
percent S02  removal. As a result  pH
control response was slow. During tests at
a pH of 5.5, the limestone feed could be
turned off or on  for 2 hours without a
noticeable change in pH. Operation under
these conditions  can reduce  limestone
utilization. However, procedures were im-
plemented to reduce the effect of blinding.
Limestone blinding was also observed
during forced oxidation tests, but it was
caused  by air compressor maintenance
problems which reduced oxidation efficien-
cy. Sulfite precipitation on the surface of
limestone particles can reduce the lime-
stone dissolution rate or reactivity,  con-
tributing to reduced limestone utilization.

Dibasic Acid Testing
  Approximately 1 -Vz  months of testing
was conducted to evaluate the performance
of dibasic acid (DBA). DBA is a less expen-
sive buffering agent than adipic acid, and
the feed system for the DBA solution is
simpler than the solids handling system
required for adipic acid DBA was tested at
85 percent removal to assess its feasibility
at SWPP  and at 95 percent removal  to
evaluate its performance under more strin-
gent S02  removal requirements. DBA re-
sults are summarized  in Table 5.
  The 85 percent removal goal was ex-
ceeded in the tray tower at a slightly lower
buffer concentration (500 ppm equivalent
adipic acid) than required in adipic acid
testing (640 ppm adipic acid). However,
the average unit load was lower during the
DBA tests, so the L/G was greater. Lower
S02  removal was observed in the  TCA
module,  reflecting deterioration  due  to
solids accumulation as observed near the
end of the adipic acid natural oxidation
tests.
  A brief test was conducted with DBA to
achieve95 percent S02 removal. This goal

-------
was achieved in the tray tower at 1300 ppm
DBA as adipic acid at a pH of 5.4 compared
to 1750 ppm during adipic acid testing. In
theTCA module, 94 percent S02 removal
was achieved at only 1020 ppm DBA as
adipic acid, but the gas flow through  the
TCA module was less than that passing
through the tray module.
  The concentration of DBA necessary to
achieve a given S02 removal appeared to
be less than for adipic acid on an equivalent
basis, but the amount of DBA consumed
on a weight basis was greater. A feed rate
of about 27 kg DBA/hr (60 Ib/hr) was
required to achieve 85  percent S02  re-
moval at a 75 percent load, compared to
20  kg/hr (45  Ib/hr)  of  adipic acid. This
translates to about 8g DBA/kg S02 (15
Ib/ton) compared to 6 g/kg S02 (II Ib/ton)
with adipic acid.
  The reliability of both modules remained
relatively high during the DBA test period.
The tray tower operated 98 percent of the
time it was required The  TCA module
operated with 89 percent reliability. These
figures compare favorably with the adipic
acid reliability figures.

Solid Waste Testing
  Solid wastes generated during the dem-
onstration program were evaluated for:
1)  H2S generation by fermentation,  2)
bioassay results  including  cytotoxicity,
mutagenicity, and acute toxicity, and 3)
leachate composition in RCRA and ASTM
extraction tests. Four samples were tested
to represent the following operating modes:
natural oxidation without organic acid ad-
dition (baseline), forced and natural oxida-
tion with adipic acid addition, and natural
oxidation with DBA addition. The sludge
samples were mixed with fly ash collected
at SWPP to examine the effects of trace
elements in the fly ash. Small quantities of
river bank soil (less than 1  percent) were
added to the samples to provide a source
of bacteria for the fermentation testa
  Fermentation test results indicate that
anaerobic decomposition of the solid waste
was minimal, both with and without  the
addition of organic acids. Bioassay results
indicate no detectable toxicity or mutagen-
icity from any of the samples except on the
fresh water invertebrate test In the fresh
water invertebrate test  high TDS levels
from the solids extracts may be responsible
for  the  toxic effect seen  for all of  the
samples, including the  baseline  sample
without organic acid addition. None of the
leachates from any of the sludges exceeded
RCRA limits. Likewise, no high values for
elements were observed from the ASTM
extraction other than those normally  as-
sociated with FGD systems (calcium, mag-
Table 5.    DBA Test Results
     Test Period
   Average
Scrubber Feed
     pH
     Average
    DBA Cone.
(ppm as adipic acid)
Average SO2 Removal (%)
85% SO2 Removal
  Module S-1 (tray)         5.4
  Module S-2 (TCA)         5.4

95% SO2 Removal
  Module S-1 (tray)         5.4
  Module S-2 (TCA)         5.4
                      480
                      420
                     1310
                     1020
                              90
                              78
                              96
                              94
nesium, sodium,  chloride,  and sulfate).
Therefore, testing of solid wastes from a
full-scale organic-acid-enhanced FGD sys-
tem did not reveal any significant difference
from an unenhanced system.
  For a savings to be realized from the use
of organic additives in  limestone  FGD
systems, the cost of the organic acid and
the associated feed system must be offset
by reductions in other operating and equip-
ment costs.  Improved S02  removal  cap-
ability and more efficient limestone utiliza-
tion  will allow reductions  in L/G  and
limestone  stoichiometry. Associated  cost
for power, limestone,  pumps, headers,
reagent preparation and solid waste disposal
can result
  Based on studies conducted by WA<4>,
annual costs can be reduced  by about 8
percent by using an adipic-acid-enhanced
limestone  FGD system in a new 500  MW
plant firing high sulfur coal. The savings
are influenced by site specific factors such
as sulfur and chlorine content of the coal
and permit requirements for S02 removal.
While the incremental cost savings  may
not be  a large percentage, the cost  reduc-
tion  can have  a substantial impact on a
plant's operating budget
  For existing facilities, the cost reduction
that can be realized by using organic acids
depends on the specific requirements at
that facility and the cost of other alternatives.
The types of problems at existing facilities
that  can be effectively addressed  by the
use of organic acids include increasing
S02 removal to meet permit  requirements
and fuel switching to accommodate higher
sulfur coals.
  For CU.'s SWPP, incremental cost esti-
mates for three alternatives to  achieve 85
percent S02 removal were considered: 1)
increasing the L/G in the tray tower, 2)
adding adipic acid, and 3)  adding DBA.
The  option to increase L/G  is a more
capital  intensive solution because  new
pumps and  headers  are required.  The
adipic acid and DBA options require lower
incremental capital costs but  higher op-
erating costs.
                       A present worth analysis of the alterna-
                     tives was performed using escalation fac-
                     tors consistent with CU.'s long range plan-
                     ning. The escalation of  adipic  acid cost
                     ($0.60/lb in 1981 dollars [$0.32/kg])
                     and DBA cost ($0.30/lb in 1981 dollars
                     [$0.66/kg])  closely approximated QU.'s
                     assumed escalation rate of fuel oil.  The
                     escalation cost of electricity (production
                     only) closely approximated CU.'s assumed
                     escalation rate. The discount factor for the
                     analysis was assumed to be equal to the
                     general inflation rate.
                       Results of the analysis are  shown  in
                     Figure 4. The adipic acid cost option  is
                     projected to be less expensive than in-
                     creasing the  L/G during  the first 9 years.
                     Beyond that  time, L/G modifications are
                     projected to be less expensive.  While the
                     adipic acid cost estimates are  based on
                     test data at SWPP, the L/G cost estimates
                     are based on experience  with similar sys-
                     tems, and greater uncertainty exists in the
                     resulting L/G analysis.
                       The DBA option is estimated  to be less
                     expensive than either of the  first  two
                     alternatives for the 1 5 years used in the
                     analysis. The estimated DBA cost is based
                     on the limited data base  provided during
                     the demonstration program. Other options
                     for reducing the cost of DBA that have not
                     been fully investigated might result in a
                     further reduction in cost

                    Conclusions
                      The major conclusions  concerning the
                    operation of an adipic-acid-enhanced lime-
                    stone FGD system based on the  results of
                    the demonstration program at SWPP are:
                       •  Adipic acid addition  is a feasible
                         method  for improving S02 removal
                         in systems that are limited by soluble
                         alkalinity in the scrubber slurry feed.
                       •  Correlations from TCA prototype test-
                         ing  at Shawnee adequately predict
                         full-scale system performance for
                         S02 removal as a function of adipic
                         acid concentration and pH.
                       •  Limestone utilization  can be improved
                         with adipic  acid  addition  because

-------
     lower scrubber feed pH is required to
     achieve a given S02 removal.
  •  Increased limestone utilization from
     the addition of adipic acid can improve
     system reliability by  reducing the
     accumulation of precipitating solids,
     especially in forced oxidized systems.
  •  The performance of DBA is sufficient
     to provide an alternative to the use of
     adipic acid. DBA addition is the least
     expensive alternative  examined for
     the SWPP to meet requirements for
     S02 removal.

References
1.  D.A Burbank and S.C. Wang, Test
    Results on Adipic Acid Enhanced Lime/
    Limestone Scrubbing at the EPA
    Shawnee Test Facility - Second Re-
   port In:  Proceedings:  Fifth Industry
    Briefing on IERL-RTP Lime/Limestone
    Wet Scrubbing Test Programs, EPA-
    600/9-80-032 (NTIS PB 80-199813),
    July (1980).
2.  S.C Wang and D.A Burbank, Adipic
   Acid Enhanced Lime/Limestone Test
    Results at the EPA Alkali Scrubbing
    Test Facility,  EPA-625/2-82-029,
    April (1982).
3.  J. B. Jarvis, et al.. Effect of Trace Metals
    and Sulfite Oxidation on Adipic Acid
    Degradation in FGD Systems,  EPA-
    600/7-82-067 (NTIS PB 83-148379),
    December (1982).
4.  R.L Torstrick, C.D. Stephenson, J.D.
    Veitch, Economics of Limestone FGD
    Systems Using Adipic Acid, In:  Pro-
    ceedings: EPA's Industry Briefing on
    the Adipic Acid Enhanced Limestone
    FGD  Process,  EPA-600/9-82-012
    (NTIS PB82-231853), June(1982).
*>
o
   2.0
                                                                  Adipic Acid
                                                                   Addition
                    Increased
                   Recycle Rate
  DBA
Addition
                             5                      10
                                   Time (years)


 Figure 4.    Comparison of alternatives for meeting 85% removal at Springfield.
                                                                            15
  O. Hargrove, J. Co/ley, R. Glover, and M.  Owen are with Radian Corporation,
    Austin, JX 78759.
  J. David Mobley is the EPA Project Officer (see below).
  The complete report  consists of two  volumes, entitled "Full-Scale Utility FGD
    System Adipic Acid Demonstration Program:"
      "Volume I.  Process  Results," (Order  No. PB 83-238 683; Cost: $29.50.
      subject to change)
      "Volume II. Continuous Emissions Monitoring Results," (Order No. PB 83-
      238 691; Cost: $41.50, subject to change)
  The above reports will be available only from:
          National Technical Information Service
          5285 Port Royal Road
          Springfield, MA 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

-------
                                                                                                                         n
                                                                                                                         Q) O
> m

-------